Dual-Licensing Linux Kernel with GPL V2 and GPL V3

Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Hi,

As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
and Linux Kernel is GPL V2 only.
So, another point is, which is previously mentioned by Linus and others, 
that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
it is needed the permission of all the maintainers permission who 
contributed to the Linux Kernel and there are a lot of lost or dead 
maintainers. Which makes it impossible to get all the maintainers' 
permission.
But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it 
will allow us the both worlds' fruits like code exchanging from other 
Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3 
and not with GPL V2 and of course the opposite is applicable,too.

So;at this situation, what is possibility to make the Linux Kernel 
Dual-Licensed as I mentioned above and what is your opinions and 
suggestions about this idea ?

Regards,

Tarkan Erimer

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Neil Brown]

On Saturday June 9, tarkan@netone.net.tr wrote:
> Hi,
> 
> As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
> and Linux Kernel is GPL V2 only.
> So, another point is, which is previously mentioned by Linus and others, 
> that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
> it is needed the permission of all the maintainers permission who 
> contributed to the Linux Kernel and there are a lot of lost or dead 
> maintainers. Which makes it impossible to get all the maintainers' 
> permission.

You don't need the permission of maintainers.  You need the permission
of copyright owners.  The two groups overlap, but are not the same.
Dead people cannot own anything, even copyright.  Their estate
probably can.  I don't think it is theoretically impossible to get
everyone's permission, though it may be quite close to practically
impossible. 

> But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it 
> will allow us the both worlds' fruits like code exchanging from other 
> Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3 
> and not with GPL V2 and of course the opposite is applicable,too.
> 
> So;at this situation, what is possibility to make the Linux Kernel 
> Dual-Licensed as I mentioned above and what is your opinions and 
> suggestions about this idea ?

Dual licensing is no easier.  It means it is licensed to be used under
either license.  You already have permission to use it under GPLv2.
So to get a dual license, you precisely need to get access under GPLv3
i.e. to convince copyright owners to make that license grant.  A thing
that we have already agreed is at least "hard".

NeilBrown

RE: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[David Schwartz]

> But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it
> will allow us the both worlds' fruits like code exchanging from other
> Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3
> and not with GPL V2 and of course the opposite is applicable,too.

That is a misleading claim. While being dual-licensed would make it either
for other projects to adopt Linux code, it would have three downsides:

1) If Linux code were adopted into other projects that were not
dual-licensed, changes could not be imported back into Linux unless the
changes were dual-licensed which is not likely when the contributions are
made to a project that's not dual-licensed.

2) Linux could no longer take code from other projects that are GPL v2
licensed unless it could obtain them under a dual license.

And, last and probably most serious:

3) Linux derivatives could be available with just a GPL v3 license and no
GPL v2. license if the derivers wanted things that way.

DS

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jan-Benedict Glaw]

[-- Attachment #1: Type: text/plain, Size: 1508 bytes --]

On Sat, 2007-06-09 15:57:55 +1000, Neil Brown <neilb@suse.de> wrote:
> On Saturday June 9, tarkan@netone.net.tr wrote:
> > As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
> > and Linux Kernel is GPL V2 only.
> > So, another point is, which is previously mentioned by Linus and others, 
> > that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
> > it is needed the permission of all the maintainers permission who 
> > contributed to the Linux Kernel and there are a lot of lost or dead 
> > maintainers. Which makes it impossible to get all the maintainers' 
> > permission.
> 
> You don't need the permission of maintainers.  You need the permission
> of copyright owners.  The two groups overlap, but are not the same.
> Dead people cannot own anything, even copyright.  Their estate
> probably can.  I don't think it is theoretically impossible to get
> everyone's permission, though it may be quite close to practically
> impossible. 

And the next question is: How much copyright does a copyright owner
own?  For example, think of drivers written by one person, but a small
number of lines changed here and there by others to adopt the code to
new APIs.  Ask them all, I think?

MfG, JBG

-- 
      Jan-Benedict Glaw      jbglaw@lug-owl.de              +49-172-7608481
Signature of:             God put me on earth to accomplish a certain number of
the second  :            things. Right now I am so far behind I will never die.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Hi David,

David Schwartz wrote:
>> But; if the Linux kernel should Dual-Licensed (GPL V2 and GPL V3), it
>> will allow us the both worlds' fruits like code exchanging from other
>> Open Source Projects (OpenSolaris etc.) that is compatible with GPL V3
>> and not with GPL V2 and of course the opposite is applicable,too.
>>     
>
> That is a misleading claim. While being dual-licensed would make it either
> for other projects to adopt Linux code, it would have three downsides:
>
> 1) If Linux code were adopted into other projects that were not
> dual-licensed, changes could not be imported back into Linux unless the
> changes were dual-licensed which is not likely when the contributions are
> made to a project that's not dual-licensed.
>
> 2) Linux could no longer take code from other projects that are GPL v2
> licensed unless it could obtain them under a dual license.
>
> And, last and probably most serious:
>
> 3) Linux derivatives could be available with just a GPL v3 license and no
> GPL v2. license if the derivers wanted things that way.
>
>   

Thanks for the corrections ;-) The whole picture is more clear now for 
me :-)
BTW,I found a really interesting blog entry about which code in Linux 
Kernel is using which version of GPL :

http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html

The work done on a Linux 2.6.20. The result is quite interesting. 
Because almost half  (Around %60 of the code licensed under "GPLv2 Only" 
and the rest is "GPLv2 or above","GPL-Version not specified,others that 
have not stated which and what version of License has been used) of the 
code is "GPLv2 or above" licensed. And also stated in the article that 
some of the codes should be "Dual Licensed" not the whole Linux kernel 
needed to be "Dual Licensed". So,if it is really like this, maybe we can 
make,for example: "File system related Codes", "Dual Licensed" and it 
will allow us to port ZFS from OpenSolaris requested by a lot of people 
or other things maybe ?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Jan-Benedict Glaw wrote:
> On Sat, 2007-06-09 15:57:55 +1000, Neil Brown <neilb@suse.de> wrote:
>   
>> On Saturday June 9, tarkan@netone.net.tr wrote:
>>     
>>> As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
>>> and Linux Kernel is GPL V2 only.
>>> So, another point is, which is previously mentioned by Linus and others, 
>>> that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
>>> it is needed the permission of all the maintainers permission who 
>>> contributed to the Linux Kernel and there are a lot of lost or dead 
>>> maintainers. Which makes it impossible to get all the maintainers' 
>>> permission.
>>>       
>> You don't need the permission of maintainers.  You need the permission
>> of copyright owners.  The two groups overlap, but are not the same.
>> Dead people cannot own anything, even copyright.  Their estate
>> probably can.  I don't think it is theoretically impossible to get
>> everyone's permission, though it may be quite close to practically
>> impossible. 
>>     
>
> And the next question is: How much copyright does a copyright owner
> own?  For example, think of drivers written by one person, but a small
> number of lines changed here and there by others to adopt the code to
> new APIs.  Ask them all, I think?
>
> MfG, JBG
>
>   

And maybe another questions should be : How long a copyright owner can 
hold the copyright, if died or lost for sometime ? if died, the 
copyright still should be valid or not ? If lost, what the law orders at 
this point for copyright holding ?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Hi Neil,

Neil Brown wrote:
> On Saturday June 9, tarkan@netone.net.tr wrote:
>   
>> Hi,
>>
>> As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
>> and Linux Kernel is GPL V2 only.
>> So, another point is, which is previously mentioned by Linus and others, 
>> that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
>> it is needed the permission of all the maintainers permission who 
>> contributed to the Linux Kernel and there are a lot of lost or dead 
>> maintainers. Which makes it impossible to get all the maintainers' 
>> permission.
>>     
>
> You don't need the permission of maintainers.  You need the permission
> of copyright owners.  The two groups overlap, but are not the same.
> Dead people cannot own anything, even copyright.  Their estate
> probably can.  I don't think it is theoretically impossible to get
> everyone's permission, though it may be quite close to practically
> impossible. 
>
>   
So, does it mean we can change the license of the dead people's code ?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jan Engelhardt]

On Jun 10 2007 11:37, Tarkan Erimer wrote:
>
> Thanks for the corrections ;-) The whole picture is more clear now for me :-)
> BTW,I found a really interesting blog entry about which code in Linux Kernel is
> using which version of GPL :
>
> http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html

You've got to take MODULE_LICENSE() into account. There is

	MODULE_LICENSE("GPL");
	MODULE_LICENSE("GPL v2");
	MODULE_LICENSE("GPL and additional rights");
	MODULE_LICENSE("Dual BSD/GPL");
	MODULE_LICENSE("Dual MIT/GPL");
	MODULE_LICENSE("Dual MPL/GPL");

I think it's time to set things right, making

	* MODULE_LICENSE the authoritative place for the license
	  (also makes it easier to grep for)

	* sync up license into MODULE_LICENSE



	Jan
-- 

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Simon Arlott]

On 10/06/07 09:37, Tarkan Erimer wrote:
> BTW,I found a really interesting blog entry about which code in Linux
> Kernel is using which version of GPL :
> 
> http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
> 
> 
> The work done on a Linux 2.6.20. The result is quite interesting.
> Because almost half  (Around %60 of the code licensed under "GPLv2 Only"
> and the rest is "GPLv2 or above","GPL-Version not specified,others that
> have not stated which and what version of License has been used) of the
> code is "GPLv2 or above" licensed. And also stated in the article that
> some of the codes should be "Dual Licensed" not the whole Linux kernel
> needed to be "Dual Licensed". So,if it is really like this, maybe we can
> make,for example: "File system related Codes", "Dual Licensed" and it
> will allow us to port ZFS from OpenSolaris requested by a lot of people
> or other things maybe ?

Once code obtained under the GPLv2 only licence that the kernel is released 
under is modified and submitted back to Linus for inclusion, that code 
would become GPLv2 only - only the original would be BSD/LGPL/GPLv2+ and 
only separate changes to the original could continue to be available under 
dual licence. Since most files will have been modified at various stages 
in Linux's development when major internal changes occur, surely practically 
everything is now GPLv2 only?


> So, does it mean we can change the license of the dead people's code ? 

If you can contact whoever currently owns the copyright, they can release 
it under another licence... however this is no good because the derivative 
work would be GPLv2 only. Perhaps if you got *everyone* at all stages of 
development (including code that has been removed if existing code is a 
derivative work of it) to agree - then it could work.

It only takes one person's code, uncooperative or not contactable, to 
prevent a change to the licence, so there's not much point in trying 
unless you intend to start replacing such code.


On 10/06/07 10:03, Jan Engelhardt wrote:
> You've got to take MODULE_LICENSE() into account. There is
> 
> 	MODULE_LICENSE("GPL");
> 	MODULE_LICENSE("GPL v2");
> 	MODULE_LICENSE("GPL and additional rights");
> 	MODULE_LICENSE("Dual BSD/GPL");
> 	MODULE_LICENSE("Dual MIT/GPL");
> 	MODULE_LICENSE("Dual MPL/GPL");

Surely that doesn't work since the entire Linux kernel is (and can only be) 
released as GPLv2? Wouldn't anyone making changes to those files need to 
obtain a copy under the other licence and explicitly release it under both 
licenses in order to maintain that?

-- 
Simon Arlott

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Neil Brown]

On Sunday June 10, tarkan@netone.net.tr wrote:
> Hi Neil,
> 
> Neil Brown wrote:
> > On Saturday June 9, tarkan@netone.net.tr wrote:
> >   
> >> Hi,
> >>
> >> As we know the forthcoming GPL V3 will be not compatible with the GPL V2 
> >> and Linux Kernel is GPL V2 only.
> >> So, another point is, which is previously mentioned by Linus and others, 
> >> that if it is decided to upgrade the Linux Kernel's License to GPL V3, 
> >> it is needed the permission of all the maintainers permission who 
> >> contributed to the Linux Kernel and there are a lot of lost or dead 
> >> maintainers. Which makes it impossible to get all the maintainers' 
> >> permission.
> >>     
> >
> > You don't need the permission of maintainers.  You need the permission
> > of copyright owners.  The two groups overlap, but are not the same.
> > Dead people cannot own anything, even copyright.  Their estate
> > probably can.  I don't think it is theoretically impossible to get
> > everyone's permission, though it may be quite close to practically
> > impossible. 
> >
> >   
> So, does it mean we can change the license of the dead people's code ?
> 

I presume the heirs of the dead people could change the license.  And
if they have no heir, then there is no-one to sue for breach of
copyright, so I assume the copyright lapses.

And I wouldn't be surprised if there were some legal precedent that
allowed for some process whereby we could make a "best effort" to
contact copyright holders (including registered paper letters and
entries in the "Public Notices" section of major newspapers) and if
no-one stepped forward to claim copyright in a reasonable period of
time we could assume that the copyright had lapsed.  But you would
need to ask a lawyer, and it would be different in different
countries.

But I think this is largely academic.  You only need a fairly small
number of fairly significant contributors to say "no" and the rest of
the process would be pointless.  And at last count, the number of
kernel people who were not keen on GPLv3 was fairly high.  Of course
no-one knows for certain yet what the final GPLv3 will be, and maybe
lots of people would change their mind when it comes out.

There would certainly be value in a straw-pole once GPLv3 was out and
had been discussed for a while - to see if a license change to GPLv3
would be accepted by a majority of current developers.  Doing that
would at least provide a clear statistic to point people at.

NeilBrown

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

david@lang.hm wrote:
> On Sun, 10 Jun 2007, Tarkan Erimer wrote:
>
>> Date: Sun, 10 Jun 2007 11:43:28 +0300
>> From: Tarkan Erimer <tarkan@netone.net.tr>
>> To: linux-kernel@vger.kernel.org
>> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
>>
>> Jan-Benedict Glaw wrote:
>>>  On Sat, 2007-06-09 15:57:55 +1000, Neil Brown <neilb@suse.de> wrote:
>>>
>>> >  On Saturday June 9, tarkan@netone.net.tr wrote:
>>> > > >  As we know the forthcoming GPL V3 will be not compatible with 
>>> the GPL > >  V2 and Linux Kernel is GPL V2 only.
>>> > >  So, another point is, which is previously mentioned by Linus 
>>> and > >  others, that if it is decided to upgrade the Linux Kernel's 
>>> License to > >  GPL V3, it is needed the permission of all the 
>>> maintainers permission > >  who contributed to the Linux Kernel and 
>>> there are a lot of lost or > >  dead maintainers. Which makes it 
>>> impossible to get all the > >  maintainers' permission.
>>> > > >  You don't need the permission of maintainers.  You need the 
>>> permission
>>> >  of copyright owners.  The two groups overlap, but are not the same.
>>> >  Dead people cannot own anything, even copyright.  Their estate
>>> >  probably can.  I don't think it is theoretically impossible to get
>>> >  everyone's permission, though it may be quite close to practically
>>> >  impossible. >
>>>  And the next question is: How much copyright does a copyright owner
>>>  own?  For example, think of drivers written by one person, but a small
>>>  number of lines changed here and there by others to adopt the code to
>>>  new APIs.  Ask them all, I think?
>>>
>>>  MfG, JBG
>>>
>>>
>>
>> And maybe another questions should be : How long a copyright owner 
>> can hold the copyright, if died or lost for sometime ? if died, the 
>> copyright still should be valid or not ? If lost, what the law orders 
>> at this point for copyright holding ?
>
> I believe that in the US it's life + 90 years.
>
> David Lang
Hmm... Really,it is damn too much time to wait! It's really better idea 
to replace the code of this person as said before instead of waiting 
such  90+ years!

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[david]

On Sun, 10 Jun 2007, Tarkan Erimer wrote:

>> >  And maybe another questions should be : How long a copyright owner can 
>> >  hold the copyright, if died or lost for sometime ? if died, the 
>> >  copyright still should be valid or not ? If lost, what the law orders at 
>> >  this point for copyright holding ?
>>
>>  I believe that in the US it's life + 90 years.
>>
>>  David Lang
> Hmm... Really,it is damn too much time to wait! It's really better idea to 
> replace the code of this person as said before instead of waiting such  90+ 
> years!

exactly, however as others are pointing out, there are a lot of active 
developers who do not agree with some of the key points of the GPLv3 
(including Linus), so until you convince them that the GPLv3 is better it 
really doesn't matter how hard it is to deal with the people who you can't 
contact.

David Lang

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[debian developer]

On 6/10/07, david@lang.hm <david@lang.hm> wrote:
> On Sun, 10 Jun 2007, Tarkan Erimer wrote:
>
> >> >  And maybe another questions should be : How long a copyright owner can
> >> >  hold the copyright, if died or lost for sometime ? if died, the
> >> >  copyright still should be valid or not ? If lost, what the law orders at
> >> >  this point for copyright holding ?
> >>
> >>  I believe that in the US it's life + 90 years.
> >>
> >>  David Lang
> > Hmm... Really,it is damn too much time to wait! It's really better idea to
> > replace the code of this person as said before instead of waiting such  90+
> > years!
>
> exactly, however as others are pointing out, there are a lot of active
> developers who do not agree with some of the key points of the GPLv3
> (including Linus), so until you convince them that the GPLv3 is better it

Last heard, Linus was quite impressed with the toned down version of
the final draft of GPLv3. I think Linus, and other major developers
should make their stand on this issue clear so that the kernel
community can discuss the future steps.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jan Engelhardt]

On Jun 10 2007 10:17, Simon Arlott wrote:
>On 10/06/07 09:37, Tarkan Erimer wrote:
>> BTW,I found a really interesting blog entry about which code in Linux
>> Kernel is using which version of GPL :
>> 
>> http://6thsenseless.blogspot.com/2007/02/how-much-linux-kernel-code-is-gpl-2.html
>> 
>> You've got to take MODULE_LICENSE() into account. There is
>> 
>> 	MODULE_LICENSE("GPL");
>> 	MODULE_LICENSE("GPL v2");
>> 	MODULE_LICENSE("GPL and additional rights");
>> 	MODULE_LICENSE("Dual BSD/GPL");
>> 	MODULE_LICENSE("Dual MIT/GPL");
>> 	MODULE_LICENSE("Dual MPL/GPL");
>
>Surely that doesn't work since the entire Linux kernel is (and can only be) 
>released as GPLv2? Wouldn't anyone making changes to those files need to 
>obtain a copy under the other licence and explicitly release it under both 
>licenses in order to maintain that?

http://www.eweek.com/article2/0,1759,1915720,00.asp
has the answer. Quoting Linus:

"If you want to license a program under any later version of the GPL, you have
to state so explicitly. Linux never did."

Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files
are GPL2 only.


	Jan
-- 

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jiri Kosina]

On Sun, 10 Jun 2007, Neil Brown wrote:

> I presume the heirs of the dead people could change the license.  And if 
> they have no heir, then there is no-one to sue for breach of copyright, 
> so I assume the copyright lapses.

In most of the law systems out there the copyright stays valid for 70 
years (or so) after the holder's death.

-- 
Jiri Kosina

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Sunday 10 June 2007 08:45:41 Jiri Kosina wrote:
> On Sun, 10 Jun 2007, Neil Brown wrote:
> > I presume the heirs of the dead people could change the license.  And if
> > they have no heir, then there is no-one to sue for breach of copyright,
> > so I assume the copyright lapses.
>
> In most of the law systems out there the copyright stays valid for 70
> years (or so) after the holder's death.

I'm almost certain that it is the same in the US, not the death+90 previously 
stated. (I've read the copyright laws a number of times to deal with some 
involved conversations) In some of the writings tied the change that made it 
death+70 its stated that said change was made to "bring US laws in line with 
Europe and most of the rest of the world" (paraphrase - I didn't bother going 
and digging out the page again). It's a relatively common belief - and, IIRC, 
was even brought before the US Supreme Court - that the copyrights length was 
changed to give Disney longer protection, if just because a lot of Disney's 
copyrights were going to expire and the change was applied retroactively. 
(And, unsurprisingly, the suit was shot down - 70 years is still a "limited" 
period. However, IIRC, there was some noted concern by the Supreme Court that 
the US Congress would exploit the legal loophole and just keep extending the 
copyright period retroactively to make it, effectively, never-ending. That 
scenario, while not *technically* in violation of the language of the US 
Constitution (which grants the US Congress the power to set the length of 
copyrights) would violate the spirit of it)

DRH

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Sun, Jun 10, 2007 at 07:36:39PM +1000, Neil Brown wrote:
>...
> And I wouldn't be surprised if there were some legal precedent that
> allowed for some process whereby we could make a "best effort" to
> contact copyright holders (including registered paper letters and
> entries in the "Public Notices" section of major newspapers) and if
> no-one stepped forward to claim copyright in a reasonable period of
> time we could assume that the copyright had lapsed.  But you would
> need to ask a lawyer, and it would be different in different
> countries.
>...

A legal precedent valid in all jurisdictions?

Harald suceessfully takes legal actions against people violating his 
copyright on the Linux kernel under the terms of the GPL in Germany at 
German courts based on German laws.

If someone finds any legal precedent in Finland or the USA or Russia 
that some copyright would have lapsed for some reason, would this have 
any legal effect in Germany?

> NeilBrown

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> But I think this is largely academic.  You only need a fairly small
> number of fairly significant contributors to say "no" and the rest of
> the process would be pointless.  And at last count, the number of
> kernel people who were not keen on GPLv3 was fairly high.  Of course
> no-one knows for certain yet what the final GPLv3 will be, and maybe
> lots of people would change their mind when it comes out.

You can take a fair bet someone will say no, or much more likely they or
whoever inherited their copyright will say $50,000

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Sunday 10 June 2007 09:40:23 Alan Cox wrote:
> > But I think this is largely academic.  You only need a fairly small
> > number of fairly significant contributors to say "no" and the rest of
> > the process would be pointless.  And at last count, the number of
> > kernel people who were not keen on GPLv3 was fairly high.  Of course
> > no-one knows for certain yet what the final GPLv3 will be, and maybe
> > lots of people would change their mind when it comes out.
>
> You can take a fair bet someone will say no, or much more likely they or
> whoever inherited their copyright will say $50,000

I seeds shades of Merkey there :P

Seriously, though, this was all settled a long time ago. Linus said "While 
individual parts of the kernel *could* be licensed [with another license] the 
kernel as a whole is strictly GPLv2" (I've tried to get it right, but my 
memory isn't as good as it used to be when it comes to useful quotes like 
that)

DRH

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

debian developer wrote:
> On 6/10/07, david@lang.hm <david@lang.hm> wrote:
>> On Sun, 10 Jun 2007, Tarkan Erimer wrote:
>>
>> >> >  And maybe another questions should be : How long a copyright 
>> owner can
>> >> >  hold the copyright, if died or lost for sometime ? if died, the
>> >> >  copyright still should be valid or not ? If lost, what the law 
>> orders at
>> >> >  this point for copyright holding ?
>> >>
>> >>  I believe that in the US it's life + 90 years.
>> >>
>> >>  David Lang
>> > Hmm... Really,it is damn too much time to wait! It's really better 
>> idea to
>> > replace the code of this person as said before instead of waiting 
>> such  90+
>> > years!
>>
>> exactly, however as others are pointing out, there are a lot of active
>> developers who do not agree with some of the key points of the GPLv3
>> (including Linus), so until you convince them that the GPLv3 is 
>> better it
>
> Last heard, Linus was quite impressed with the toned down version of
> the final draft of GPLv3. I think Linus, and other major developers
> should make their stand on this issue clear so that the kernel
> community can discuss the future steps.
Yep, the GPLv3 probably will release around July time. So;luckily, we 
had very little time to see the final decision about it :-) I hope we 
should upgrade to GPLv3 and Sun should "Dual License" the OpenSolaris 
via GPLv3 (or at least,GPLv3 should be CDDL compatible.). So,we should 
have more fruits (like ZFS,DTrace etc.) ;-)

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 1369 bytes --]

> In most of the law systems out there the copyright stays valid for 70 
> years (or so) after the holder's death.

[I don't want to appear picky but IMO it is important to be as precise
as one could possibly be, therefor...]

That is not quite correct. At least for the german law system and
presumably for most if not all similar systems the corrrect stmts is:

In most of the law systems out there the copyright stays valid for 70 
years (or so) after initial publication of the copyrighted work. For
work whose publication date can't be determined (*) (e.g. unpublished work)
the creators death is assumed to be the publication date for the sake
of determining the  aforementioned 70 years protection period.

(*) "can't be determined" does not apply for published work with an
insecure initial publication date. In those cases the earliest documented
publication (i.e. prooveable in court) is considered the initial publication
.date

Note that the original phrase could last indefinitely by simply moving
the copyright from person to person.

Best,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 1334 bytes --]

[legal precedence to force waiving copyright> 

> A legal precedent valid in all jurisdictions?
> 
> Harald suceessfully takes legal actions against people violating his 
> copyright on the Linux kernel under the terms of the GPL in Germany at 
> German courts based on German laws.
> 
> If someone finds any legal precedent in Finland or the USA or Russia 
> that some copyright would have lapsed for some reason, would this have 
> any legal effect in Germany?

That is difficult but AFAIK it work in the "increase protection"
direction but not the other way around, at least in germany.

For example even though the copyright on Micky Mouse is no longer
valid by german law it is still enforcible under german law because
the german jurisdiction does acknowledge the (recently extended)
longer such period in the US.

Apart from that I'm sure you'd find some state "legally" waiving a
copyright on whatever. If that would start an avalanche of similar
terminations everywhere else I'd be utterly surprised.

Best,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Sun, Jun 10, 2007 at 04:23:20PM +0200, Michael Gerdau wrote:
> > In most of the law systems out there the copyright stays valid for 70 
> > years (or so) after the holder's death.
> 
> [I don't want to appear picky but IMO it is important to be as precise
> as one could possibly be, therefor...]
> 
> That is not quite correct. At least for the german law system and
> presumably for most if not all similar systems the corrrect stmts is:
> 
> In most of the law systems out there the copyright stays valid for 70 
> years (or so) after initial publication of the copyrighted work.
>...

Please read the German "Gesetz über Urheberrecht und verwandte 
Schutzrechte" before making completely false claims - in Germany it's
70 years after the death of the author.

> Best,
> Michael

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 830 bytes --]

> Please read the German "Gesetz über Urheberrecht und verwandte 
> Schutzrechte" before making completely false claims - in Germany it's
> 70 years after the death of the author.

Oups, I had been mixing up named published works and anonymously
published works.

My previous claim is valid for anonymously published work only and
thus holds no relevance for the situation discussed. I apologize for
having raised a false claim.

The copyright for works with a known is valid for 70 years after
the authors death.

Sorry,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tim Post]

On Sun, 2007-06-10 at 19:36 +1000, Neil Brown wrote:

> I presume the heirs of the dead people could change the license.  And
> if they have no heir, then there is no-one to sue for breach of
> copyright, so I assume the copyright lapses.
> 
> And I wouldn't be surprised if there were some legal precedent that
> allowed for some process whereby we could make a "best effort" to
> contact copyright holders (including registered paper letters and
> entries in the "Public Notices" section of major newspapers) and if
> no-one stepped forward to claim copyright in a reasonable period of
> time we could assume that the copyright had lapsed.  But you would
> need to ask a lawyer, and it would be different in different
> countries.

I've done some research on this and from what I can tell you are
correct. There is some sort of "Due Diligence" law that you have to
satisfy that is slightly different from country to country. 

>From what I can tell the process is similar to changing your name, an ad
in the paper would be o.k. in most places. Since the intent of the
copyright holder to make their work free is clear, its pretty clear cut.

It was hard to find specifics, because not many people worry about it if
they know the copyright holder to be existentially challenged. I didn't
find many cases of people even bothering with the formality.

I say existentially challenged because you don't need to be dead to
vanish. I also read it as ".. a person of questionable existence." as
well as "A null (or moot) party".

> But I think this is largely academic.  You only need a fairly small
> number of fairly significant contributors to say "no" and the rest of
> the process would be pointless.  And at last count, the number of
> kernel people who were not keen on GPLv3 was fairly high.  Of course
> no-one knows for certain yet what the final GPLv3 will be, and maybe
> lots of people would change their mind when it comes out.

I think its good that people have their own viewpoints instead of just
watching the Linus - RMS tennis match as the points of the license get
debated and assuming the views of whoever 'wins' in their eyes. Neither
person is going to be around forever. 

> There would certainly be value in a straw-pole once GPLv3 was out and
> had been discussed for a while - to see if a license change to GPLv3
> would be accepted by a majority of current developers.  Doing that
> would at least provide a clear statistic to point people at.

Unless you are distributing the Linux kernel in a whole or parts of it
in some way you don't even HAVE to accept the terms of the GPL2, or GPL3
for that matter. This doesn't really concern me as an end user, but it
should interest me (and does).

I'm more worried about someone finding a way to leverage existing or
future patents against users of the Linux kernel, or people who provide
the use of a computer using the Linux kernel as a service.

I'm not *overly* concerned about that. Anyone trying to get people to
pay for the right to use the Linux kernel would put themselves in clear
and certain danger of becoming ' a person of questionable existence '
themselves.

Best,
--Tim

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Sun, Jun 10, 2007 at 04:25:55PM +0530, debian developer wrote:
>  On 6/10/07, david@lang.hm <david@lang.hm> wrote:
> > On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> >
> > >> >  And maybe another questions should be : How long a copyright owner 
> > can
> > >> >  hold the copyright, if died or lost for sometime ? if died, the
> > >> >  copyright still should be valid or not ? If lost, what the law orders 
> > at
> > >> >  this point for copyright holding ?
> > >>
> > >>  I believe that in the US it's life + 90 years.
> > >>
> > >>  David Lang
> > > Hmm... Really,it is damn too much time to wait! It's really better idea 
> > to
> > > replace the code of this person as said before instead of waiting such  
> > 90+
> > > years!
> >
> > exactly, however as others are pointing out, there are a lot of active
> > developers who do not agree with some of the key points of the GPLv3
> > (including Linus), so until you convince them that the GPLv3 is better it
> 
>  Last heard, Linus was quite impressed with the toned down version of
>  the final draft of GPLv3. I think Linus, and other major developers
>  should make their stand on this issue clear so that the kernel
>  community can discuss the future steps.

"future steps"?  Hah.

My code is going to stay GPLv2 as the v3 license is horrible for kernel
code for all of the reasons I have said in the past, plus a few more
(what, I can make an "industrial" product but not a commercial one?
That's horrible...)

thanks,

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Sun, Jun 10, 2007 at 05:21:53PM +0300, Tarkan Erimer wrote:
>  I hope we should upgrade to GPLv3 and Sun should "Dual License" the
>  OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL compatible.).

The OpenSolaris community has already stated that they do not want to
accept GPLv3, why not discuss this with them if you want to try to
change their minds?

>  So,we should have more fruits (like ZFS,DTrace etc.) ;-)

I think the transfer would be more the other way, we have a zillion more
things that they do not than the other way around.

thanks,

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> > 
> > Last heard, Linus was quite impressed with the toned down version of
> > the final draft of GPLv3.

I was impressed in the sense that it was a hell of a lot better than the 
disaster that were the earlier drafts.

I still think GPLv2 is simply the better license.

I consider dual-licensing unlikely (and technically quite hard), but at 
least _possible_ in theory. I have yet to see any actual *reasons* for 
licensing under the GPLv3, though. All I've heard are shrill voices about 
"tivoization" (which I expressly think is ok) and panicked worries about 
Novell-MS (which seems way overblown, and quite frankly, the argument 
seems to not so much be about the Novell deal, as about an excuse to push 
the GPLv3).

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Sun, 10 Jun 2007, Linus Torvalds wrote:
>
> I have yet to see any actual *reasons* for licensing under the GPLv3, 
> though.

Btw, if Sun really _is_ going to release OpenSolaris under GPLv3, that 
_may_ be a good reason. I don't think the GPLv3 is as good a license as 
v2, but on the other hand, I'm pragmatic, and if we can avoid having two 
kernels with two different licenses and the friction that causes, I at 
least see the _reason_ for GPLv3. As it is, I don't really see a reason at 
all.

I personally doubt it will happen, but hey, I didn't really expect them to 
open-source Java either(*), so it's not like I'm infallible in my 
predictions.

				Linus

(*) And I've been pushing for that since before they even released it - I 
walked out on Bill Joy at a private event where they discussed their 
horrible previous Java license.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> licensing under the GPLv3, though. All I've heard are shrill voices about 
> "tivoization" (which I expressly think is ok) and panicked worries about 

GPLv2 probably forbids Tivoisation anyway. Which is good IMHO even if not
yours 8)

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jeff Garzik]

Linus Torvalds wrote:
> I still think GPLv2 is simply the better license.

Ditto.

	Jeff

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[debian developer]

On 6/10/07, Tarkan Erimer <tarkan@netone.net.tr> wrote:
> debian developer wrote:
> > On 6/10/07, david@lang.hm <david@lang.hm> wrote:
> >> On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> >>
> >> >> >  And maybe another questions should be : How long a copyright
> >> owner can
> >> >> >  hold the copyright, if died or lost for sometime ? if died, the
> >> >> >  copyright still should be valid or not ? If lost, what the law
> >> orders at
> >> >> >  this point for copyright holding ?
> >> >>
> >> >>  I believe that in the US it's life + 90 years.
> >> >>
> >> >>  David Lang
> >> > Hmm... Really,it is damn too much time to wait! It's really better
> >> idea to
> >> > replace the code of this person as said before instead of waiting
> >> such  90+
> >> > years!
> >>
> >> exactly, however as others are pointing out, there are a lot of active
> >> developers who do not agree with some of the key points of the GPLv3
> >> (including Linus), so until you convince them that the GPLv3 is
> >> better it
> >
> > Last heard, Linus was quite impressed with the toned down version of
> > the final draft of GPLv3. I think Linus, and other major developers
> > should make their stand on this issue clear so that the kernel
> > community can discuss the future steps.
> Yep, the GPLv3 probably will release around July time. So;luckily, we
> had very little time to see the final decision about it :-) I hope we
> should upgrade to GPLv3 and Sun should "Dual License" the OpenSolaris
> via GPLv3 (or at least,GPLv3 should be CDDL compatible.). So,we should
> have more fruits (like ZFS,DTrace etc.) ;-)
>
>
>
I don't think that upgrading to GPLv3 just for the sake of tools
present in some other software should be the reason. We are capable
enough of developing our own tools, and many experienced people are
working on equivalent(etx4 etc.,) and much sophisticated tools for the
linux kernel.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[debian developer]

On 6/10/07, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
> > licensing under the GPLv3, though. All I've heard are shrill voices about
> > "tivoization" (which I expressly think is ok) and panicked worries about
>
> GPLv2 probably forbids Tivoisation anyway. Which is good IMHO even if not
              ^^^^^^^^

Now that is a bit waving in the air. GPLv2 forbids Tivoisation
theoretically but practically it didnt stop them doing it practically.
I agree with Linus that software licenses should have their influence
only on the software part and leave the freedom of the hardware on
which the software runs to the hardware manufacturers.

But was it the goal of GPLv2??

And what does Andrew Morton think of all this? I really want to know
his opinions....

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Andrew Morton]

On Mon, 11 Jun 2007 01:02:42 +0530 "debian developer" <debiandev@gmail.com> wrote:

> And what does Andrew Morton think of all this? I really want to know
> his opinions....

I have yet to see Linus make a statement on these matters with which
I didn't agree.

RE: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[David Schwartz]

> http://www.eweek.com/article2/0,1759,1915720,00.asp
> has the answer. Quoting Linus:
>
> "If you want to license a program under any later version of the
> GPL, you have
> to state so explicitly. Linux never did."
>
> Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files
> are GPL2 only.

The GPL states the default position:

"If the Program does not specify a version number of this License, you may
choose any version ever published by the Free Software Foundation."

Leaving the question of whether Linus's comment at the top of the license
changes the default:

"Also note that the only valid version of the GPL as far as the kernel is
concerned is _this_ particular version of the license (ie v2, not v2.2 or
v3.x or whatever), unless explicitly otherwise stated."

So we have dueling defaults. The GPL says the default is any version. Linus'
statement at the top of the GPL says the default is v2 only. It's not clear,
at least to me, that there is any clear reason why one should win out over
the other.

DS

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> > GPLv2 probably forbids Tivoisation anyway. Which is good IMHO even if not
>               ^^^^^^^^
> 
> Now that is a bit waving in the air. GPLv2 forbids Tivoisation
> theoretically but practically it didnt stop them doing it practically.

They've never been given permission and there is no caselaw yet, doesn't
mean they are allowed to.

GPL2 actually in some ways was saner than GPL3 on this - you could
sensibly argue the key was part of the source/build environment but it
didn't then get muddled in with questions like ROMs which the new GPL3
wording is a bit messy about still.

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[James Bruce]

Linus Torvalds wrote:
[ snip ]
> I consider dual-licensing unlikely (and technically quite hard), but at 
> least _possible_ in theory. I have yet to see any actual *reasons* for 
> licensing under the GPLv3, though.
[ snip ]

One thing that would make that easier in the future is if contributers
at least started to dual-license their submissions.  I.e. if instead
of "GPL version 2", one could say "GPL version 2 or GPL version 3".
It isn't the same thing as the problematic "GPL version 2 or later",
because the developer is not agreeing to an unseen license (GPLv4,
etc).  What it does do is make it easier to move to GPLv3 a few years
from now, if that is decided then, as a significant fraction of the
code will already be GPLv3 compatible.  That way, if a reason is ever
found to move to v3, at least some of the work will already be done.

  - Jim Bruce

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jesper Juhl]

On 10/06/07, James Bruce <bruce@andrew.cmu.edu> wrote:
> Linus Torvalds wrote:
> [ snip ]
> > I consider dual-licensing unlikely (and technically quite hard), but at
> > least _possible_ in theory. I have yet to see any actual *reasons* for
> > licensing under the GPLv3, though.
> [ snip ]
>
> One thing that would make that easier in the future is if contributers
> at least started to dual-license their submissions.  I.e. if instead
> of "GPL version 2", one could say "GPL version 2 or GPL version 3".
> It isn't the same thing as the problematic "GPL version 2 or later",
> because the developer is not agreeing to an unseen license (GPLv4,
> etc).  What it does do is make it easier to move to GPLv3 a few years
> from now, if that is decided then, as a significant fraction of the
> code will already be GPLv3 compatible.  That way, if a reason is ever
> found to move to v3, at least some of the work will already be done.
>
Good luck convincing all contributors to do that.

Personally I'm happy with GPL v2, and I for one won't be
dual-licensing anything I contribute until I see a clear benefit of
doing so (and I don't yet).

In any case, this whole debate is still a bit premature since GPL v3
has not even arrived in its final form yet.

-- 
Jesper Juhl <jesper.juhl@gmail.com>
Don't top-post  http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please      http://www.expita.com/nomime.html

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Mon, Jun 11, 2007 at 12:52:41AM +0530, debian developer wrote:

> I don't think that upgrading to GPLv3 just for the sake of tools
> present in some other software should be the reason. We are capable
> enough of developing our own tools, and many experienced people are
> working on equivalent(etx4 etc.,) and much sophisticated tools for the
> linux kernel.

I don't think that switch to GPLv3 can be described as upgrade.  I certainly
have no intention to do that to my code; some of it I might release under BSD
license, and that can be used in any project.  The rest of the kernel stuff
I've done (and that's the majority of my contributions) is under GPLv2 *only*.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[James Bruce]

Jesper Juhl wrote:
>> One thing that would make that easier in the future is if contributers
>> at least started to dual-license their submissions.  I.e. if instead
>> of "GPL version 2", one could say "GPL version 2 or GPL version 3".
>> It isn't the same thing as the problematic "GPL version 2 or later",
>> because the developer is not agreeing to an unseen license (GPLv4,
>> etc).  What it does do is make it easier to move to GPLv3 a few years
>> from now, if that is decided then, as a significant fraction of the
>> code will already be GPLv3 compatible.  That way, if a reason is ever
>> found to move to v3, at least some of the work will already be done.
>>
> Good luck convincing all contributors to do that.

Well, it's something that pro-GPLv3 people can do right now, instead of 
just lobbying/complaining.  Given 1000 developers, if 400 start dual 
licensing now, and down the road some compelling reason for GPLv3 does 
arise (read: a lawsuit with teeth), that's 600 people you need to 
contact/convince to change, not 1000.  This is made more interesting by 
that fact that 40% of the kernel code is already "GPLv2 or later", as 
someone else pointed out.

> Personally I'm happy with GPL v2, and I for one won't be
> dual-licensing anything I contribute until I see a clear benefit of
> doing so (and I don't yet).

Well, all my personal (non-kernel) stuff is still GPLv2 only right now 
(Linus' opinion is what convinced me that "or later" is dumb), and like 
many I disliked the original GPLv3 draft.  I'm willing to wait until the 
final one is out though, and I think my libraries will end up being 
dual-licensed, with contributions required to be dual-licensed.  I want 
to avoid v3 lock-in, but I don't want to cripple v3 projects either.

> In any case, this whole debate is still a bit premature since GPL v3
> has not even arrived in its final form yet.

Agreed.

  - Jim Bruce

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Sun, Jun 10, 2007 at 09:54:58PM +0100, Alan Cox wrote:
> > > GPLv2 probably forbids Tivoisation anyway. Which is good IMHO even if not
> >               ^^^^^^^^
> > 
> > Now that is a bit waving in the air. GPLv2 forbids Tivoisation
> > theoretically but practically it didnt stop them doing it practically.
> 
> They've never been given permission and there is no caselaw yet, doesn't
> mean they are allowed to.

Are you sure?  Tivo went and got a FSF "verification" of their system a
number of years ago and got their blessing that what they were doing was
just fine with regards to the GPLv2.

This is one reason Tivo's lawyers were so perplexed when the FSF then
turned around and made their company's name into a term to describe DRM
stuff and started preaching how it was so bad.  It seemed to be in
direct crontridiction from what they had previously been told by the
very same people.

Now yes, they didn't consult with the individual owners of the kernel,
who might hold different views as to if v2 covers keys like you have
stated in the past, but the FSF's position in this area does hold some
ammount of weight, especially in court if it were to come to that.

thanks,

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[H. Peter Anvin]

Tarkan Erimer wrote:
> And maybe another questions should be : How long a copyright owner can
> hold the copyright, if died or lost for sometime ? if died, the
> copyright still should be valid or not ? If lost, what the law orders at
> this point for copyright holding ?

In most countries, copyright lasts for 70 years after the death of the
author.  Either way, it is most definitely valid.

	-hpa

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Greg KH wrote:
> On Sun, Jun 10, 2007 at 05:21:53PM +0300, Tarkan Erimer wrote:
>   
>>  I hope we should upgrade to GPLv3 and Sun should "Dual License" the
>>  OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL compatible.).
>>     
>
> The OpenSolaris community has already stated that they do not want to
> accept GPLv3, why not discuss this with them if you want to try to
> change their minds?
>   
It was just an example came to my mind at first when thinking about 
"Dual Licensing" or upgrading Linux Kernel to the GPLv3. Yeah maybe,the 
"OpenSolaris Community" do not want GPLv3. But; IMHO, it is in the hands 
of "Sun" not the "OpenSolaris Community".

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

debian developer wrote:
> On 6/10/07, Tarkan Erimer <tarkan@netone.net.tr> wrote:
>> debian developer wrote:
>> > On 6/10/07, david@lang.hm <david@lang.hm> wrote:
>> >> On Sun, 10 Jun 2007, Tarkan Erimer wrote:
>> >>
>> >> >> >  And maybe another questions should be : How long a copyright
>> >> owner can
>> >> >> >  hold the copyright, if died or lost for sometime ? if died, the
>> >> >> >  copyright still should be valid or not ? If lost, what the law
>> >> orders at
>> >> >> >  this point for copyright holding ?
>> >> >>
>> >> >>  I believe that in the US it's life + 90 years.
>> >> >>
>> >> >>  David Lang
>> >> > Hmm... Really,it is damn too much time to wait! It's really better
>> >> idea to
>> >> > replace the code of this person as said before instead of waiting
>> >> such  90+
>> >> > years!
>> >>
>> >> exactly, however as others are pointing out, there are a lot of 
>> active
>> >> developers who do not agree with some of the key points of the GPLv3
>> >> (including Linus), so until you convince them that the GPLv3 is
>> >> better it
>> >
>> > Last heard, Linus was quite impressed with the toned down version of
>> > the final draft of GPLv3. I think Linus, and other major developers
>> > should make their stand on this issue clear so that the kernel
>> > community can discuss the future steps.
>> Yep, the GPLv3 probably will release around July time. So;luckily, we
>> had very little time to see the final decision about it :-) I hope we
>> should upgrade to GPLv3 and Sun should "Dual License" the OpenSolaris
>> via GPLv3 (or at least,GPLv3 should be CDDL compatible.). So,we should
>> have more fruits (like ZFS,DTrace etc.) ;-)
>>
>>
>>
> I don't think that upgrading to GPLv3 just for the sake of tools
> present in some other software should be the reason. We are capable
> enough of developing our own tools, and many experienced people are
> working on equivalent(etx4 etc.,) and much sophisticated tools for the
> linux kernel.
It is not because of the sake of the tools and we have no capable enough 
developers. It's just about an example that came to my mind, as I 
mentioned before and also,it is the same thing as we, all the time, did. 
I mean getting and sharing codes from many different open source 
projects like BSD and countless others. So, OpenSolaris makes no 
difference at this.

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Mon, Jun 11, 2007 at 09:46:18AM +0300, Tarkan Erimer wrote:
> Greg KH wrote:
> >On Sun, Jun 10, 2007 at 05:21:53PM +0300, Tarkan Erimer wrote:
> >  
> >> I hope we should upgrade to GPLv3 and Sun should "Dual License" the
> >> OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL compatible.).
> >>    
> >
> >The OpenSolaris community has already stated that they do not want to
> >accept GPLv3, why not discuss this with them if you want to try to
> >change their minds?
> >  
> It was just an example came to my mind at first when thinking about 
> "Dual Licensing" or upgrading Linux Kernel to the GPLv3. Yeah maybe,the 
> "OpenSolaris Community" do not want GPLv3. But; IMHO, it is in the hands 
> of "Sun" not the "OpenSolaris Community".

Perhaps.  However, since the only thing in hands of your kind of advocates
is best not mentioned on a family-friendly maillist, may I suggest taking
that exciting thread to more appropriate place?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Al Viro wrote:
> Perhaps.  However, since the only thing in hands of your kind of advocates
> is best not mentioned on a family-friendly maillist, may I suggest taking
> that exciting thread to more appropriate place?
>   
I don't think that this thread is going unfriendly or harmfully. 
However, what is your suggestion ?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Tarkan Erimer <tarkan@netone.net.tr> wrote:

> > > >> I hope we should upgrade to GPLv3 and Sun should "Dual License" 
> > > >> the OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL 
> > > >> compatible.).
> > > >
> > > > The OpenSolaris community has already stated that they do not 
> > > > want to accept GPLv3, why not discuss this with them if you want 
> > > > to try to change their minds?
> > >
> > > It was just an example came to my mind at first when thinking 
> > > about "Dual Licensing" or upgrading Linux Kernel to the GPLv3. 
> > > Yeah maybe,the "OpenSolaris Community" do not want GPLv3. But; 
> > > IMHO, it is in the hands of "Sun" not the "OpenSolaris Community".
> >
> > Perhaps.  However, since the only thing in hands of your kind of 
> > advocates is best not mentioned on a family-friendly maillist, may I 
> > suggest taking that exciting thread to more appropriate place?
>
> I don't think that this thread is going unfriendly or harmfully. 
> However, what is your suggestion ?

if you want to change the minds of the OpenSolaris community, i'd 
proffer that it's perhaps more efficient to talk to them, not to the 
linux-kernel mailing list. Thanks,

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Ingo Molnar wrote:
> if you want to change the minds of the OpenSolaris community, i'd 
> proffer that it's perhaps more efficient to talk to them, not to the 
> linux-kernel mailing list. Thanks,
>
> 	Ingo
>   

I do not want to and try to change anyone's mind: nor the Open Solaris 
Community nor the Linux Community. Just, I asked simple question and 
included a simple example in it. Son, including an example related to 
OpenSolaris does not mean that I want to push OpenSolaris things. That's 
all.

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Tarkan Erimer <tarkan@netone.net.tr> wrote:

> [...] Just, I asked simple question and included a simple example in 
> it. [...]

actually, what you said was this:

" I hope we should upgrade to GPLv3 and Sun should "Dual License"
  the OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL
  compatible.). "

and to that the answer was:

" The OpenSolaris community has already stated that they do not want to
  accept GPLv3 [...] "

in other words: your hypothetical is false today. You called us to do a 
specific action, but why did you then include a factually false 
'example' to underline that point of yours? Or if you simply did not 
know about the OpenSolaris community's position beforehand, why dont you 
just admit that and withdraw from that line of argument gracefully?

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Ingo Molnar wrote:
> * Tarkan Erimer <tarkan@netone.net.tr> wrote:
>
>   
>> [...] Just, I asked simple question and included a simple example in 
>> it. [...]
>>     
>
> actually, what you said was this:
>
> " I hope we should upgrade to GPLv3 and Sun should "Dual License"
>   the OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL
>   compatible.). "
>   
Why don't you include the last sentence I wrote: "So,we should have more 
fruits (like ZFS,DTrace etc.) ;-) "
So, that's why I said it. Because, as all the time, we did it: Importing 
and exporting codes to/from different open source projects.

> and to that the answer was:
>
> " The OpenSolaris community has already stated that they do not want to
>   accept GPLv3 [...] "
>
> in other words: your hypothetical is false today. You called us to do a 
> specific action, but why did you then include a factually false 
> 'example' to underline that point of yours? Or if you simply did not 
> know about the OpenSolaris community's position beforehand, why dont you 
> just admit that and withdraw from that line of argument gracefully?
>
> 	Ingo
>   

As I mentioned in my previous posts: This is **not** in the hands of the 
"OpenSolaris Community" to make and apply such decision. Sun itself 
**will decide** it. Also, there are strong indications that Sun is very 
interested to make "OpenSolaris" at least "Dual-Licensed" with GPLv3.

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Linus Torvalds wrote:
> On Sun, 10 Jun 2007, Linus Torvalds wrote:
>   
>> I have yet to see any actual *reasons* for licensing under the GPLv3, 
>> though.
>>     
>
> Btw, if Sun really _is_ going to release OpenSolaris under GPLv3, that 
> _may_ be a good reason. I don't think the GPLv3 is as good a license as 
> v2, but on the other hand, I'm pragmatic, and if we can avoid having two 
> kernels with two different licenses and the friction that causes, I at 
> least see the _reason_ for GPLv3. As it is, I don't really see a reason at 
> all.
>
> I personally doubt it will happen, but hey, I didn't really expect them to 
> open-source Java either(*), so it's not like I'm infallible in my 
> predictions.
>
> 				Linus
>
> (*) And I've been pushing for that since before they even released it - I 
> walked out on Bill Joy at a private event where they discussed their 
> horrible previous Java license.
>   

Thanks for making things more clear :-) Some really strong indications 
that Sun is very willing to,at least, "Dual-License" the OpenSolaris 
with GPLv3. I think; in a very short time; we will see when the GPLv3 
finalized and released.

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Tarkan Erimer <tarkan@netone.net.tr> wrote:

> Ingo Molnar wrote:
> >* Tarkan Erimer <tarkan@netone.net.tr> wrote:
> >
> >  
> >>[...] Just, I asked simple question and included a simple example in 
> >>it. [...]
> >>    
> >
> >actually, what you said was this:
> >
> >" I hope we should upgrade to GPLv3 and Sun should "Dual License"
> >  the OpenSolaris via GPLv3 (or at least,GPLv3 should be CDDL
> >  compatible.). "
> >  
> Why don't you include the last sentence I wrote: "So,we should have more 
> fruits (like ZFS,DTrace etc.) ;-) "

You might as well have said "the moon is made out of cheese" and i'd not 
have quoted it either. Why? Because it's irrelevant to the fundamental 
point that was raised and which you keep ignoring: that the only 
"example" you cited is a hypothetical that is currently false. In any 
case, speculation about what Sun might or might not do, up until the 
point it actually does it, is not something i feel compelled to do 
anything over, so please stop wasting my time by Cc:-ing me. Thanks,

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Ingo Molnar wrote:
> You might as well have said "the moon is made out of cheese" and i'd not 
> have quoted it either. Why? Because it's irrelevant to the fundamental 
> point that was raised and which you keep ignoring: that the only 
> "example" you cited is a hypothetical that is currently false. In any 
> case, speculation about what Sun might or might not do, up until the 
> point it actually does it, is not something i feel compelled to do 
> anything over, so please stop wasting my time by Cc:-ing me. Thanks,
>
> 	Ingo
>   

I think, you do not want to understand what I really  mean. OK,I 
stopping here. Because, you already wasted a lot of my time via always 
not understanding what I really mean.

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Tarkan Erimer <tarkan@netone.net.tr> wrote:

> > (*) And I've been pushing for that since before they even released 
> > it - I walked out on Bill Joy at a private event where they 
> > discussed their horrible previous Java license.
> 
> Thanks for making things more clear :-) Some really strong indications 
> that Sun is very willing to,at least, "Dual-License" the OpenSolaris 
> with GPLv3. I think; in a very short time; we will see when the GPLv3 
> finalized and released.

that would certainly be a good and productive move from them. Note the 
issue that others have pointed out to you: OpenSolaris is probably more 
interested in picking up code from Linux than the other way around! :-) 
You mentioned "dtrace" and "ZFS". Firstly, Linux already has a "dtrace" 
equivalent. Secondly, ZFS might be interesting in theory, although our 
prior experience of having compatibly-licensed filesystems ported over 
to Linux has been pretty negative: XFS ended up being an integration 
nightmare - and that doesnt have to do anything with the qualities of 
XFS (it's one of the cleanest Linux filesystems, if not the cleanest), 
the problem is that components within a kernel are very tightly 
integrated and rarely does it make sense to port over more than just 
drivers or maybe libraries. And that's i guess what OpenSolaris lacks 
and which i suspect it is mostly interested in: lots of nice Linux 
drivers ;-) XFS, the largest Linux filesystem is 100K lines of code - 
and ZFS (i've never seen it) is very likely smaller than that. Linux 
drivers on the other hand, as of today, are _3.7 million_ lines of code 
and enable Linux to run on 99% of the hardware that is produced today. 
Guess which one has the larger strategic significance? ;-)

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> Now yes, they didn't consult with the individual owners of the kernel,
> who might hold different views as to if v2 covers keys like you have
> stated in the past, but the FSF's position in this area does hold some
> ammount of weight, especially in court if it were to come to that.

The authors position does have rather a lot of weight too. Especially as
they have been made plain to Tivo and various other relevant parties.

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tim Post]

On Sun, 2007-06-10 at 11:49 +0300, Tarkan Erimer wrote:
>  
> So, does it mean we can change the license of the dead people's code ?
> 

Please realize that one doesn't need to be dead to become
uncommunicative incapacitated or vanish. The only need to be somewhere
other than where they were without updating anyone. 

Here is a very humorous, but sort of scary theoretical :

"Linus was so disturbed by the code in a submitted patch that he had a
nervous breakdown and spent the next 30 years in a padded room.
Unfortunately, no provisions were left to determine what happens to his
copyrights should he become incapacitated."

That is a LOT different than 

"Linus passed away at age 397 today, and left an appropriate will so his
work remains useful for those enjoying it."

If you are a substantial contributor, address this in a will should you
die and limited power of attorney to someone that you trust should you
fall off the face of the planet. Just like you would see to it that your
assets were properly dispersed if something happened to you.

If your unfortunate enough to not have anyone in your life that you can
trust, you can grant power of attorney to any reputable charitable
organization that you feel mirrors your beliefs and ideals. 

Prevention is so much better than a cure and quite cheap. 

Best,
--Tim

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bron Gondwana]

On Mon, Jun 11, 2007 at 11:03:48AM +0200, Ingo Molnar wrote:
> 
> * Tarkan Erimer <tarkan@netone.net.tr> wrote:
> 
> > > (*) And I've been pushing for that since before they even released 
> > > it - I walked out on Bill Joy at a private event where they 
> > > discussed their horrible previous Java license.
> > 
> > Thanks for making things more clear :-) Some really strong indications 
> > that Sun is very willing to,at least, "Dual-License" the OpenSolaris 
> > with GPLv3. I think; in a very short time; we will see when the GPLv3 
> > finalized and released.
> 
> that would certainly be a good and productive move from them. Note the 
> issue that others have pointed out to you: OpenSolaris is probably more 
> interested in picking up code from Linux than the other way around! :-) 
> You mentioned "dtrace" and "ZFS". Firstly, Linux already has a "dtrace" 
> equivalent. Secondly, ZFS might be interesting in theory
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Having just got my first Sun box (one of those 48 disk Thumper machines)
I can say it's very interesting in practice too.  The admin tools for
zfs are a dream to use (unlike just about everything else on Solaris
which is a bugwards compatible nightmare to use) and the attitude of
checksumming everything on to disk and checking that those checksums
match on the way back out presses my "do your own safety checking and
don't trust the hardware" buttons very much the right way - especially
with that much hardware in there.

I would love to see Sun GPL3 OpenSolaris, not so much for the code
itself (maybe not portable into Linux) but for the clarity it would
give to the patent position.  The patent peace would be portable back
to GPL2.

Bron.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Ingo Molnar wrote:
> * Tarkan Erimer <tarkan@netone.net.tr> wrote:
>
>   
>>> (*) And I've been pushing for that since before they even released 
>>> it - I walked out on Bill Joy at a private event where they 
>>> discussed their horrible previous Java license.
>>>       
>> Thanks for making things more clear :-) Some really strong indications 
>> that Sun is very willing to,at least, "Dual-License" the OpenSolaris 
>> with GPLv3. I think; in a very short time; we will see when the GPLv3 
>> finalized and released.
>>     
>
> that would certainly be a good and productive move from them. Note the 
> issue that others have pointed out to you: OpenSolaris is probably more 
> interested in picking up code from Linux than the other way around! :-) 
>   
Totally agreed :-)
> You mentioned "dtrace" and "ZFS". Firstly, Linux already has a "dtrace" 
> equivalent. Secondly, ZFS might be interesting in theory, although our 
> prior experience of having compatibly-licensed filesystems ported over 
> to Linux has been pretty negative: XFS ended up being an integration 
> nightmare - and that doesnt have to do anything with the qualities of 
> XFS (it's one of the cleanest Linux filesystems, if not the cleanest), 
> the problem is that components within a kernel are very tightly 
> integrated and rarely does it make sense to port over more than just 
> drivers or maybe libraries. And that's i guess what OpenSolaris lacks 
> and which i suspect it is mostly interested in: lots of nice Linux 
> drivers ;-) XFS, the largest Linux filesystem is 100K lines of code - 
> and ZFS (i've never seen it) is very likely smaller than that. Linux 
> drivers on the other hand, as of today, are _3.7 million_ lines of code 
> and enable Linux to run on 99% of the hardware that is produced today. 
> Guess which one has the larger strategic significance? ;-)
>
> 	Ingo
>   
Yep, it is clear that sun needs more things like drivers etc. to make 
OpenSolaris more usable and user friendly. Here is an article about this 
subject and some thoughts of Ian (Murdock) about it ;-)

http://www.zdnet.com.au/news/software/soa/Sun-hopes-for-Linux-like-Solaris/0,130061733,339276057,00.htm

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 11, 2007, Ingo Molnar <mingo@elte.hu> wrote:

> And that's i guess what OpenSolaris lacks and which i suspect it is
> mostly interested in: lots of nice Linux drivers ;-) XFS, the
> largest Linux filesystem is 100K lines of code - and ZFS (i've never
> seen it) is very likely smaller than that. Linux drivers on the
> other hand, as of today, are _3.7 million_ lines of code and enable
> Linux to run on 99% of the hardware that is produced today.  Guess
> which one has the larger strategic significance? ;-)

Per this reasoning, Sun wouldn't be waiting for GPLv3, and it would
have already released the OpenSolaris kernel under GPLv2, would it
not? ;-)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Tue, 12 Jun 2007, Alexandre Oliva wrote:
> 
> Per this reasoning, Sun wouldn't be waiting for GPLv3, and it would
> have already released the OpenSolaris kernel under GPLv2, would it
> not? ;-)

Umm. You are making the fundamental mistake of thinking that Sun is in 
this to actually further some open-source agenda.

Here's a cynical prediction (but backed up by past behaviour of Sun):

 - first off: they may be talking a lot more than they are or ever will
   be doing. How many announcements about Sun and Linux have you seen over
   the years? And how much of that has actually happened?

 - They may like open source, but Linux _has_ hurt them in the 
   marketplace. A lot.

   They almost used to own the chip design market, and it took quite a 
   long time before the big EDA vendors ported to Linux (and x86-64 in 
   particular). But when they did, their chip design market just basically 
   disappeared: sparc performance is so horribly bad (especially on a 
   workstation kind of setup), that to do chip design on them is just 
   idiotic. Which is not to say that there aren't holdouts, but let's face 
   it, for a lot of things, Solaris is simply the wrong choice these days.

   Ergo: they sure as hell don't want to help Linux. Which is fine. 
   Competition is good.

 - So they want to use Linux resources (_especially_ drivers), but they do 
   *not* want to give anything back (especially ZFS, which seems to be one 
   of their very very few bright spots).

 - Ergo: they'll not be releasing ZFS and the other things that people are 
   drooling about in a way that lets Linux use them on an equal footing. I 
   can pretty much guarantee that. They don't like competition on that 
   level. They'd *much* rather take our drivers and _not_ give anythign 
   back, or give back the stuff that doesn't matter (like core Solaris: 
   who are you kidding - Linux code is _better_).

End result:

 - they'll talk about it. They not only drool after our drivers, they 
   drool after all the _people_ who write drivers. They'd love to get 
   kernel developers from Linux, they see that we have a huge amount of 
   really talented people. So they want to talk things up, and the more 
   "open source" they can position themselves, the better.

 - They may release the uninteresting parts under some fine license. See 
   the OpenSolaris stuff - instead of being blinded by the code they _did_ 
   release under an open source license, ask yourself what they did *not* 
   end up releasing. Ask yourself why the open source parts are not ready 
   to bootstrap a competitive system, or why they are released under 
   licenses that Sun can make sure they control.

So the _last_ thing they want to do is to release the interesting stuff 
under GPLv2 (quite frankly, I think the only really interesting thing they 
have is ZFS, and even there, I suspect we'd be better off talking to 
NetApp, and seeing if they are interested in releasing WAFL for Linux).

Yes, they finally released Java under GPLv2, and they should be commended 
for that. But you should also ask yourself why, and why it took so long. 
Maybe it had something to do with the fact that other Java implementations 
started being more and more relevant?

Am I cynical? Yes. Do I expect people to act in their own interests? Hell 
yes! That's how things are _supposed_ to happen. I'm not at all berating 
Sun, what I'm trying to do here is to wake people up who seem to be living 
in some dream-world where Sun wants to help people. 

So to Sun, a GPLv3-only release would actually let them look good, and 
still keep Linux from taking their interesting parts, and would allow them 
to take at least parts of Linux without giving anything back (ahh, the 
joys of license fragmentation). 

Of course, they know that. And yes, maybe ZFS is worthwhile enough that 
I'm willing to go to the effort of trying to relicense the kernel. But 
quite frankly, I can almost guarantee that Sun won't release ZFS under the 
GPLv3 even if they release other parts. Because if they did, they'd lose 
the patent protection.

And yes, I'm cynical, and yes, I hope I'm wrong. And if I'm wrong, I'll 
very happily retract anything cynical I said about Sun. They _have_ done 
great things, and maybe I'm just too pessimistic about all the history 
I've seen of Sun with open source.

The _good_ news is that Jonathan Schwartz actually does seem to have made 
a difference, and I hope to God he is really as serious about 
open-sourcing things as he says he is. And don't get me wrong: I think a 
truly open-source GPLv3 Solaris would be a really really _good_ thing, 
even if it does end up being a one-way street as far as code is concerned!

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[debian developer]

On 6/10/07, Greg KH <greg@kroah.com> wrote:
> On Sun, Jun 10, 2007 at 04:25:55PM +0530, debian developer wrote:
> >  On 6/10/07, david@lang.hm <david@lang.hm> wrote:
> > > On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> > >
> > > >> >  And maybe another questions should be : How long a copyright owner
> > > can
> > > >> >  hold the copyright, if died or lost for sometime ? if died, the
> > > >> >  copyright still should be valid or not ? If lost, what the law orders
> > > at
> > > >> >  this point for copyright holding ?
> > > >>
> > > >>  I believe that in the US it's life + 90 years.
> > > >>
> > > >>  David Lang
> > > > Hmm... Really,it is damn too much time to wait! It's really better idea
> > > to
> > > > replace the code of this person as said before instead of waiting such
> > > 90+
> > > > years!
> > >
> > > exactly, however as others are pointing out, there are a lot of active
> > > developers who do not agree with some of the key points of the GPLv3
> > > (including Linus), so until you convince them that the GPLv3 is better it
> >
> >  Last heard, Linus was quite impressed with the toned down version of
> >  the final draft of GPLv3. I think Linus, and other major developers
> >  should make their stand on this issue clear so that the kernel
> >  community can discuss the future steps.
>
> "future steps"?  Hah.
>
> My code is going to stay GPLv2 as the v3 license is horrible for kernel
> code for all of the reasons I have said in the past, plus a few more
> (what, I can make an "industrial" product but not a commercial one?

     ^^^^^^^^^^^^^
What exactly in GPLv3 forbids you from making a commercial product?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Tue, Jun 12, 2007 at 11:37:11PM +0530, debian developer wrote:
>  On 6/10/07, Greg KH <greg@kroah.com> wrote:
> > On Sun, Jun 10, 2007 at 04:25:55PM +0530, debian developer wrote:
> > >  On 6/10/07, david@lang.hm <david@lang.hm> wrote:
> > > > On Sun, 10 Jun 2007, Tarkan Erimer wrote:
> > > >
> > > > >> >  And maybe another questions should be : How long a copyright 
> > owner
> > > > can
> > > > >> >  hold the copyright, if died or lost for sometime ? if died, the
> > > > >> >  copyright still should be valid or not ? If lost, what the law 
> > orders
> > > > at
> > > > >> >  this point for copyright holding ?
> > > > >>
> > > > >>  I believe that in the US it's life + 90 years.
> > > > >>
> > > > >>  David Lang
> > > > > Hmm... Really,it is damn too much time to wait! It's really better 
> > idea
> > > > to
> > > > > replace the code of this person as said before instead of waiting 
> > such
> > > > 90+
> > > > > years!
> > > >
> > > > exactly, however as others are pointing out, there are a lot of active
> > > > developers who do not agree with some of the key points of the GPLv3
> > > > (including Linus), so until you convince them that the GPLv3 is better 
> > it
> > >
> > >  Last heard, Linus was quite impressed with the toned down version of
> > >  the final draft of GPLv3. I think Linus, and other major developers
> > >  should make their stand on this issue clear so that the kernel
> > >  community can discuss the future steps.
> >
> > "future steps"?  Hah.
> >
> > My code is going to stay GPLv2 as the v3 license is horrible for kernel
> > code for all of the reasons I have said in the past, plus a few more
> > (what, I can make an "industrial" product but not a commercial one?
> 
>      ^^^^^^^^^^^^^
>  What exactly in GPLv3 forbids you from making a commercial product?

Nothing "forbids" me, it's just the artifical distinstion of the two is,
in my opinion, stupid and foolish.  You are trying to define use-cases
to justify their notion that you must give up the hardware keys for one
type of device, yet not for another.

Even the people that feel that v2 says you need to give up the keys
think this is dumb.  But we've been through all of that before (see
previous long thread about v3 and why the kernel developers hate it, it
all still applys to the final draft.)

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 12, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Tue, 12 Jun 2007, Alexandre Oliva wrote:
>> 
>> Per this reasoning, Sun wouldn't be waiting for GPLv3, and it would
>> have already released the OpenSolaris kernel under GPLv2, would it
>> not? ;-)

> Umm. You are making the fundamental mistake of thinking that Sun is in 
> this to actually further some open-source agenda.

Err, no.  I was merely questioning Ingo's reasoning that Sun wanted
Linux's drivers as badly as he made it seem.  All the fuss about
waiting for and going to GPLv3 wouldn't get them that.  Moving to
GPLv2 would, and still, they aren't doing it.  That was my point.

FWIW, I share most of your assessment and wait-and-see attitude about
Sun's situation.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 12, 2007, Greg KH <greg@kroah.com> wrote:

> (see previous long thread about v3 and why the kernel developers
> hate it, it all still applys to the final draft.)

You mean all the misunderstandings? ;-)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Florian Weimer]

* Linus Torvalds:

> I consider dual-licensing unlikely (and technically quite hard), but at 
> least _possible_ in theory. I have yet to see any actual *reasons* for 
> licensing under the GPLv3, though. All I've heard are shrill voices about 
> "tivoization" (which I expressly think is ok)

In a strange twist, the last-call draft contains a clause that
expressly permits some forms of "tivoization", provided a suitable
contractual arrangement exists ("Basic Permissions", second
paragraph).

Now a lot of the free software market follows this "sell yourself into
slavery" model (and even the FSF recommends to make money this way),
but I'm not sure if it's a good idea to state it so plainly in the
license.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Matthias Kaehlcke]

El Tue, Jun 12, 2007 at 08:45:46AM -0700 Linus Torvalds ha dit:

> 
> 
> On Tue, 12 Jun 2007, Alexandre Oliva wrote:
> > 
> > Per this reasoning, Sun wouldn't be waiting for GPLv3, and it would
> > have already released the OpenSolaris kernel under GPLv2, would it
> > not? ;-)
> 
> Umm. You are making the fundamental mistake of thinking that Sun is in 
> this to actually further some open-source agenda.
> 
> Here's a cynical prediction (but backed up by past behaviour of Sun):
> 
>  - first off: they may be talking a lot more than they are or ever will
>    be doing. How many announcements about Sun and Linux have you seen over
>    the years? And how much of that has actually happened?
> 
>  - They may like open source, but Linux _has_ hurt them in the 
>    marketplace. A lot.
> 
>    They almost used to own the chip design market, and it took quite a 
>    long time before the big EDA vendors ported to Linux (and x86-64 in 
>    particular). But when they did, their chip design market just basically 
>    disappeared: sparc performance is so horribly bad (especially on a 
>    workstation kind of setup), that to do chip design on them is just 
>    idiotic. Which is not to say that there aren't holdouts, but let's face 
>    it, for a lot of things, Solaris is simply the wrong choice these days.
> 
>    Ergo: they sure as hell don't want to help Linux. Which is fine. 
		 >    Competition is good.
> 
>  - So they want to use Linux resources (_especially_ drivers), but they do 
>    *not* want to give anything back (especially ZFS, which seems to be one 
>    of their very very few bright spots).
> 
>  - Ergo: they'll not be releasing ZFS and the other things that people are 
>    drooling about in a way that lets Linux use them on an equal footing. I 
>    can pretty much guarantee that. They don't like competition on that 
>    level. They'd *much* rather take our drivers and _not_ give anythign 
>    back, or give back the stuff that doesn't matter (like core Solaris: 
>    who are you kidding - Linux code is _better_).
> 
> End result:
> 
>  - they'll talk about it. They not only drool after our drivers, they 
>    drool after all the _people_ who write drivers. They'd love to get 
>    kernel developers from Linux, they see that we have a huge amount of 
>    really talented people. So they want to talk things up, and the more 
>    "open source" they can position themselves, the better.
> 
>  - They may release the uninteresting parts under some fine license. See 
>    the OpenSolaris stuff - instead of being blinded by the code they _did_ 
>    release under an open source license, ask yourself what they did *not* 
>    end up releasing. Ask yourself why the open source parts are not ready 
>    to bootstrap a competitive system, or why they are released under 
>    licenses that Sun can make sure they control.
> 
> So the _last_ thing they want to do is to release the interesting stuff 
> under GPLv2 (quite frankly, I think the only really interesting thing they 
> have is ZFS, and even there, I suspect we'd be better off talking to 
> NetApp, and seeing if they are interested in releasing WAFL for Linux).
> 
> Yes, they finally released Java under GPLv2, and they should be commended 
> for that. But you should also ask yourself why, and why it took so long. 
> Maybe it had something to do with the fact that other Java implementations 
> started being more and more relevant?
> 
> Am I cynical? Yes. Do I expect people to act in their own interests? Hell 
> yes! That's how things are _supposed_ to happen. I'm not at all berating 
> Sun, what I'm trying to do here is to wake people up who seem to be living 
> in some dream-world where Sun wants to help people. 
> 
> So to Sun, a GPLv3-only release would actually let them look good, and 
> still keep Linux from taking their interesting parts, and would allow them 
> to take at least parts of Linux without giving anything back (ahh, the 
> joys of license fragmentation). 
> 
> Of course, they know that. And yes, maybe ZFS is worthwhile enough that 
> I'm willing to go to the effort of trying to relicense the kernel. But 
> quite frankly, I can almost guarantee that Sun won't release ZFS under the 
> GPLv3 even if they release other parts. Because if they did, they'd lose 
> the patent protection.
> 
> And yes, I'm cynical, and yes, I hope I'm wrong. And if I'm wrong, I'll 
> very happily retract anything cynical I said about Sun. They _have_ done 
> great things, and maybe I'm just too pessimistic about all the history 
> I've seen of Sun with open source.
> 
> The _good_ news is that Jonathan Schwartz actually does seem to have made 
> a difference, and I hope to God he is really as serious about 
> open-sourcing things as he says he is. And don't get me wrong: I think a 
> truly open-source GPLv3 Solaris would be a really really _good_ thing, 
> even if it does end up being a one-way street as far as code is
> concerned!

FYI, Jonathan Schwartz' response:

Linus,

First, I'm glad you give credit to Sun for the contributions we've
made to the open source world, and Linux specifically - we take the
commitment seriously. It's why we freed OpenOffice, elements of Gnome,
Mozilla, delivered Java, and a long list of other contributions that
show up in almost every distro. Individuals will always define
communities, but Sun as a company has done its part to grow the market
- for others as much as ourselves.

But I disagree with a few of your points. Did the Linux community hurt
Sun? No, not a bit. It was the companies that leveraged their work. I
draw a very sharp distinction - even if our competition is
conveniently reckless. They like to paint the battle as Sun vs. the
community, and it's not. Companies compete, communities simply
fracture.

And OpenSolaris has come a very long way since you last looked. It and
its community are growing, as a result of more than ZFS (although we
seem to be generating a lot of interest there, not all intentional) -
OpenSolaris scales on any hardware, has built in virtualization, great
web service infrastucture, fault management, diagnosability, and tons
more. Feel free to try for yourself (and yes, we're fixing
installability, no fair knocking us for that.)

Now despite what you suggest, we love where the FSF's GPL3 is
headed. For a variety of mechanical reasons, GPL2 is harder for us
with OpenSolaris - but not impossible, or even out of the
question. This has nothing to do with being afraid of the community
(if it was, we wouldn't be so interested in seeing ZFS everywhere,
including Linux, with full patent indemnity). Why does open sourcing
take so long? Because we're starting from products that exist, in
which a diversity of contributors and licensors/licensees have rights
we have to negotiate. Indulge me when I say It's different than
starting from scratch. I would love to go faster, and we are all doing
everything under our control to accelerate progress. (Remember, we
can't even pick GPL3 yet - it doesn't officially exist.) It's also a
delicate dance to manage this transition while growing a corporation.

But most of all, from where I sit, we should put the swords down -
you're not the enemy for us, we're not the enemy for you. Most of the
world doesn't have access to the internet - that's the enemy to slay,
the divide that separates us. By joining our communities, we can bring
transparency and opportunity to the whole planet. Are we after your
drivers? No more than you're after ZFS or Crossbow or dtrace - it's
not predation, it's prudence. Let's stop wasting time recreating
wheels we both need to roll forward.

I wanted you to hear this from me directly. We want to work together,
we want to join hands and communities - we have no intention of
holding anything back, or pulling patent nonsense. And to prove the
sincerity of the offer, I invite you to my house for dinner. I'll
cook, you bring the wine. A mashup in the truest sense.

Best,
Jonathan

President, Chief Executive Officer,
Sun Microsystems, Inc. 

http://blogs.sun.com/jonathan/entry/one_plus_one_is_fifty

-- 
Matthias Kaehlcke
Linux Application Developer
Barcelona

    Yo soy como soy y tú eres como eres, construyamos un mundo donde yo
    pueda ser sin dejar de ser yo, donde tú puedas ser sin dejar de ser
    tú, y donde ni yo ni tú obliguemos al otro a ser como yo o como tú
                                                                 .''`.
    using free software / Debian GNU/Linux | http://debian.org  : :'  :
                                                                `. `'`
gpg --keyserver pgp.mit.edu --recv-keys 47D8E5D4                  `-

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

On Wednesday 13 June 2007 06:53, Alexandre Oliva wrote:
> You mean all the misunderstandings? ;-)

My impression as well is that there are many misunderstandings, even 
concerning the status of Linux itself. Linus is much better at kernel 
hacking than at license issues, and that's true for most other kernel 
hackers, too - that's why we have Eben Moglen to hack the license.

I want to add my two cents on what I think the legal status of the 
individual contributions to Linux are. The thing in question is not the 
GPLv2 itself (which is pretty clear that code without explicit statements 
is under "any", and if you make an explicit statement, it should read "GPL 
version two, or (at your option) any later version"), it's this text on the 
top of /usr/src/linux/COPYING:

"Also note that the only valid version of the GPL as far as the kernel
 is concerned is _this_ particular version of the license (ie v2, not
 v2.2 or v3.x or whatever), unless explicitly otherwise stated.

                        Linus Torvalds"

This text was added in or around 2.4.0-test9, but without asking for 
permission (neither from the FSF, which has the copyright of the GPL, nor 
from the other authors of the Linux kernel), and with some controversion 
afterwards.

This particular comment to how the GPL is applied to the Linux kernel 
therefore doesn't change the GPL as such (it can't without breaking 
copyright), neither does it change the licensing conditions the original 
authors put on their contribution (it can't without breaking copyright, 
either), but may only provide interpretations downstream (for the user). 
Linus is also entitled to make clairifications there, which the first 
paragraph obviously does (i.e. the text Linus added is not a change of the 
license, but a comment on it).

Again: What Linus is entitled to do is to *select* the license under which 
he redistributes the code downstream. What he can't do is to *change* the 
intention of the original author. So if you can choose what this somewhat 
ambiguous message means, and restrict yourself to reasoning that doesn't go 
into nonsense or copyright infringement, you'll pretty much come to the 
conclusion that the only thing Linus could have done back then without 
asking for permission is the license condition how he *redistributes* the 
compiled work called "Linux kernel" (it's a "compiled work", because it 
consists of a compilation of individual files from many authors).

The GPLv2 however is very clear how the end user gets the license: from the 
original author. Not from the man in the middle, from a distributor or 
kernel maintainer, who can neither add nor drop restrictions/permissions 
(and thus the special rights of a compilation editor are void). The author 
can only speak for himself, not by behalf of somebody else, as well as the 
compilation editor. That's why the FSF is so strict about having each 
author stating copyright and the license conditions on the top of the 
file - nobody else can.

So my conclusion is: If you, as contributor to the Linux kernel, want to 
make clear that your work really is GPLv2 only, you have to do that 
yourself, you have to add a notice like above to files where you 
exclusively own copyright. Very few have done that in the past, most people 
who *did* explicitely declare what versions of the GPL they want their work 
under, did choose the default text from the GPLv2, which sais "GPLv2 or 
later" (most use the GPL text template). The rest (the majority) did not 
choose to say anything, which under the GPL regime means "any"; and nobody 
but the author himself can change that (by adding a specific version). 
Linus can't change the GPL regime, because he can't change the GPL.

So, IMHO and IANAL, technically, there are only a few files in Linux which 
really can't work in a GPLv2+GPLv3 compiled Linux, and a few files wouldn't 
be a problem.

>From a practical point of view, I fully agree with Linus that there's no 
point in switching over to the GPLv3 next month unless there's some 
valueable contribution out there that's only available under GPLv3 (maybe 
from OpenSolaris), or the Linux kernel developers understand the GPLv3 
better. I don't see this point in the near future.

But what I want to say: The route to GPLv3 is not as blocked as it appears. 
And the GPLv2 is even better than you think: It paves that road as well: In 
practice, only projects that have a thight authorship control can really 
make their project GPLv2 only (like MySQL), and those projects have no 
problem to change their mind later.

What I don't understand about the GPLv3 with keys is why that depends on the 
use case. As user of commercial devices like company routers, firewalls and 
such, which often are Linux based, I don't want them sealed by the vendor, 
as well. An explicit statement is even worse than an implicit one (as in 
the GPLv2, which has been tested in a German court by Harald Welte - 
Siemens had to turn in the keys).

And now flame me to death ;-).

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Krzysztof Halasa]

Bernd Paysan <bernd.paysan@gmx.de> writes:

> I want to add my two cents on what I think the legal status of the 
> individual contributions to Linux are. The thing in question is not the 
> GPLv2 itself (which is pretty clear that code without explicit statements 
> is under "any",

That's not exactly true. A work without explicit statements is not
licenced at all.

> This particular comment to how the GPL is applied to the Linux kernel 
> therefore doesn't change the GPL as such

Of course.

> Again: What Linus is entitled to do is to *select* the license under which 
> he redistributes the code downstream.

Sure.

> The GPLv2 however is very clear how the end user gets the license: from the 
> original author.

I'd be surprised if it's for GPL to decide.

> Not from the man in the middle, from a distributor or 
> kernel maintainer, who can neither add nor drop restrictions/permissions 
> (and thus the special rights of a compilation editor are void). The author 
> can only speak for himself, not by behalf of somebody else, as well as the 
> compilation editor.

How about derived works?
Am I free to get BSD source, incorporate it in GPL project, and release
the whole under GPL?
Sure, the original source stays BSD but I don't distribute it.

> That's why the FSF is so strict about having each 
> author stating copyright and the license conditions on the top of the 
> file - nobody else can.

I'm not sure the copyright laws define "files".

> So my conclusion is: If you, as contributor to the Linux kernel, want to 
> make clear that your work really is GPLv2 only, you have to do that 
> yourself, you have to add a notice like above to files where you 
> exclusively own copyright.

I don't think the law works like that.
By default you have no rights to someone's work (file or project).
The only licence I can find with Linux is GPL v2, isn't it? And even
that wasn't stated explicite until that 2.4.0something (though there
is a consensus that the COPYING file was indeed a licence for the
whole kernel).

Then you may have additional rights, such as those given in various
source files.

> The rest (the majority) did not 
> choose to say anything, which under the GPL regime means "any";

What exactly is the "GPL regime" and how is it defined by copyright
law and/or the GPL licence itself (or will of copyright holders etc.)?
-- 
Krzysztof Halasa

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Pavel Machek]

Hi!

> > So, does it mean we can change the license of the dead people's code ?
> > 
> 
> Please realize that one doesn't need to be dead to become
> uncommunicative incapacitated or vanish. The only need to be somewhere
> other than where they were without updating anyone. 
> 
> Here is a very humorous, but sort of scary theoretical :
> 
> "Linus was so disturbed by the code in a submitted patch that he had a
> nervous breakdown and spent the next 30 years in a padded room.
> Unfortunately, no provisions were left to determine what happens to his
> copyrights should he become incapacitated."

License is only promise not to sue. If Linus is safely in a padded
room, I'll happily relicense Linux under GPLv17, knowing he is not
going to sue me :-).

(And actually, if he's in a padded room, there's probably someone who
can act in his name. At least czech law works like that.)
							Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 6225 bytes --]

On Wednesday 13 June 2007 15:11, Krzysztof Halasa wrote:
> > The GPLv2 however is very clear how the end user gets the license: from
> > the original author.
>
> I'd be surprised if it's for GPL to decide.

If you choose the GPL as license, the text of the GPL are the conditions. 
Otherwise, the GPL would be pure nonsense (as would be any other license).

> > Not from the man in the middle, from a distributor or
> > kernel maintainer, who can neither add nor drop
> > restrictions/permissions (and thus the special rights of a compilation
> > editor are void). The author can only speak for himself, not by behalf
> > of somebody else, as well as the compilation editor.
>
> How about derived works?
> Am I free to get BSD source, incorporate it in GPL project, and release
> the whole under GPL?
> Sure, the original source stays BSD but I don't distribute it.

Derivated work is a product of several authors, therefore each author may 
put different conditions on his part of the work - as long as they are 
compatible, it's ok. A derivated work originally under BSD, now with a 
patch under GPL can only be distributed under GPL, but not under BSD 
(because GPL requires redistribution under GPL, whereas BSD doesn't care). 
If you take out the patch, and revert the work to the BSD one, you are free 
to redistribute it under BSD.

There's no point of discussing that the Linux kernel *as a whole* (as 
compilation) currently is under GPLv2 only, since it sais so, and a few 
files also explicitely say so. The whole combination is GPLv2 only, but 
most parts aren't.

> > That's why the FSF is so strict about having each
> > author stating copyright and the license conditions on the top of the
> > file - nobody else can.
>
> I'm not sure the copyright laws define "files".

Copyright law defines "work", and in terms of computer programs, source 
code "work" goes into files. Or patches, but patches end up distributed 
over several files.

The nice thing about this is that you can make automatic checks about the 
license you actually have to fulfill. E.g. if you compile a hypothetical 
Linux 2.8.15.3 without ZFS and dtrace in 2009, you may end up with 
compiling only GPLv2-compatible code, and therefore can tivoize your system 
(unless you sell it to Germany, where the GPLv2 outlaws tivoizing by 
intent), but if you add either ZFS or dtrace, you can't.

> > So my conclusion is: If you, as contributor to the Linux kernel, want
> > to make clear that your work really is GPLv2 only, you have to do that
> > yourself, you have to add a notice like above to files where you
> > exclusively own copyright.
>
> I don't think the law works like that.
> By default you have no rights to someone's work (file or project).
> The only licence I can find with Linux is GPL v2, isn't it?

Yes, and the GPLv2 sais "if the FSF releases a new version of the GPL, you 
may update" (section 9):

"  9. The Free Software Foundation may publish revised and/or new versions
of the General Public License from time to time.  Such new versions will
be similar in spirit to the present version, but may differ in detail to
address new problems or concerns.

Each version is given a distinguishing version number.  If the Program
specifies a version number of this License which applies to it and "any
later version", you have the option of following the terms and conditions
either of that version or of any later version published by the Free
Software Foundation.  If the Program does not specify a version number of
this License, you may choose any version ever published by the Free Software
Foundation."

This is normal contract law (you have to say "yes" to a new M$ EULA every 
few months or so if you are an unlucky Windows user, and like to use their 
patches), contracts can change over time. The FSF is rather nice here, they 
say users and contributors can make choices to which contract they use.

Given that the license you find with Linux is GPLv2, anyway, the comment 
from Linus seems to be superfluous. The license already says which version 
it is. But it has this upgrade option, and one possible interpretation of 
Linus' comment is "no, it doesn't have this update option".

> And even 
> that wasn't stated explicite until that 2.4.0something (though there
> is a consensus that the COPYING file was indeed a licence for the
> whole kernel).
>
> Then you may have additional rights, such as those given in various
> source files.
>
> > The rest (the majority) did not
> > choose to say anything, which under the GPL regime means "any";
>
> What exactly is the "GPL regime" and how is it defined by copyright
> law and/or the GPL licence itself (or will of copyright holders etc.)?

If I use GPL as license, I'm under "GPL regime", i.e. the terms of the GPL 
apply. The GPL requires that I need to speak up explicitely if I want to 
limit the choice of licenses - if I don't say anything, it defaults to "any 
GPL". This is a restriction that goes from author to user, sinde the GPL 
cuts away all middle-men (restrictions applied by middle-men can be 
reverted).

If I decide to build my own compilation of Linux kernel patches* (i.e. a -bp 
kernel, like the -mm kernel is a different compilation of Linux kernel 
patches as the mainline kernel), I'm free to choose under which conditions 
I redistribute it, given that it's compatible with the conditions the 
original authors have chosen. Most of them have said nothing (other than 
implicitely that it's ok for them to put it under GPL, as they haven't 
opposed to inclusion into the Linux kernel), some have said GPLv2 or later, 
some say GPLv1.1 or later (e.g. the parport driver) and a few have 
said "GPLv2 only". Now, I may rewrite those few "GPLv2 only" files, and 
then I have a GPLv2-or later compatible linux-some.version-bp kernel. And 
into this kernel, I can add code under GPLv3 (once the GPLv3 is ready and 
there's code worth to add under GPLv3), which limits me to redistribute the 
whole thing under GPLv3.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Hi Linus,

Linus Torvalds wrote:
> On Tue, 12 Jun 2007, Alexandre Oliva wrote:
>   
>> Per this reasoning, Sun wouldn't be waiting for GPLv3, and it would
>> have already released the OpenSolaris kernel under GPLv2, would it
>> not? ;-)
>>     
>
> Umm. You are making the fundamental mistake of thinking that Sun is in 
> this to actually further some open-source agenda.
>
> Here's a cynical prediction (but backed up by past behaviour of Sun):
>
>  - first off: they may be talking a lot more than they are or ever will
>    be doing. How many announcements about Sun and Linux have you seen over
>    the years? And how much of that has actually happened?
>
>  - They may like open source, but Linux _has_ hurt them in the 
>    marketplace. A lot.
>
>    They almost used to own the chip design market, and it took quite a 
>    long time before the big EDA vendors ported to Linux (and x86-64 in 
>    particular). But when they did, their chip design market just basically 
>    disappeared: sparc performance is so horribly bad (especially on a 
>    workstation kind of setup), that to do chip design on them is just 
>    idiotic. Which is not to say that there aren't holdouts, but let's face 
>    it, for a lot of things, Solaris is simply the wrong choice these days.
>
>    Ergo: they sure as hell don't want to help Linux. Which is fine. 
>    Competition is good.
>
>  - So they want to use Linux resources (_especially_ drivers), but they do 
>    *not* want to give anything back (especially ZFS, which seems to be one 
>    of their very very few bright spots).
>
>  - Ergo: they'll not be releasing ZFS and the other things that people are 
>    drooling about in a way that lets Linux use them on an equal footing. I 
>    can pretty much guarantee that. They don't like competition on that 
>    level. They'd *much* rather take our drivers and _not_ give anythign 
>    back, or give back the stuff that doesn't matter (like core Solaris: 
>    who are you kidding - Linux code is _better_).
>
>   
Completely agreed :-)

> End result:
>
>  - they'll talk about it. They not only drool after our drivers, they 
>    drool after all the _people_ who write drivers. They'd love to get 
>    kernel developers from Linux, they see that we have a huge amount of 
>    really talented people. So they want to talk things up, and the more 
>    "open source" they can position themselves, the better.
>
>   
Definitely. They already began to pull some people like Ian Murdock. And 
I'm really very disappointed of this move,Ian did. Especially, such a 
person who has very good reputation and high profile in the Linux 
Community. He immediately shut down his company (also leaved 
Linux-Foundation) and joined to sun. After joining, he made statements 
like "How to make Solaris more like Linux ?" etc. Like a 40 years 
employee at Sun. Another interesting thing is the timing of this hiring. 
So, this situation is a good example of it.

>  - They may release the uninteresting parts under some fine license. See 
>    the OpenSolaris stuff - instead of being blinded by the code they _did_ 
>    release under an open source license, ask yourself what they did *not* 
>    end up releasing. Ask yourself why the open source parts are not ready 
>    to bootstrap a competitive system, or why they are released under 
>    licenses that Sun can make sure they control.
>
> So the _last_ thing they want to do is to release the interesting stuff 
> under GPLv2 (quite frankly, I think the only really interesting thing they 
> have is ZFS, and even there, I suspect we'd be better off talking to 
> NetApp, and seeing if they are interested in releasing WAFL for Linux).
>
> Yes, they finally released Java under GPLv2, and they should be commended 
> for that. But you should also ask yourself why, and why it took so long. 
> Maybe it had something to do with the fact that other Java implementations 
> started being more and more relevant?
>
> Am I cynical? Yes. Do I expect people to act in their own interests? Hell 
> yes! That's how things are _supposed_ to happen. I'm not at all berating 
> Sun, what I'm trying to do here is to wake people up who seem to be living 
> in some dream-world where Sun wants to help people. 
>
> So to Sun, a GPLv3-only release would actually let them look good, and 
> still keep Linux from taking their interesting parts, and would allow them 
> to take at least parts of Linux without giving anything back (ahh, the 
> joys of license fragmentation). 
>
> Of course, they know that. And yes, maybe ZFS is worthwhile enough that 
> I'm willing to go to the effort of trying to relicense the kernel. But 
> quite frankly, I can almost guarantee that Sun won't release ZFS under the 
> GPLv3 even if they release other parts. Because if they did, they'd lose 
> the patent protection.
>
> And yes, I'm cynical, and yes, I hope I'm wrong. And if I'm wrong, I'll 
> very happily retract anything cynical I said about Sun. They _have_ done 
> great things, and maybe I'm just too pessimistic about all the history 
> I've seen of Sun with open source.
>
> The _good_ news is that Jonathan Schwartz actually does seem to have made 
> a difference, and I hope to God he is really as serious about 
> open-sourcing things as he says he is. And don't get me wrong: I think a 
> truly open-source GPLv3 Solaris would be a really really _good_ thing, 
> even if it does end up being a one-way street as far as code is concerned!
>
> 			Linus
>   
One more time,agreed ;-)

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Tarkan Erimer]

Matthias Kaehlcke wrote:
> FYI, Jonathan Schwartz' response:
>
> I wanted you to hear this from me directly. We want to work together,
> we want to join hands and communities - we have no intention of
> holding anything back, or pulling patent nonsense. And to prove the
> sincerity of the offer, I invite you to my house for dinner. I'll
> cook, you bring the wine. A mashup in the truest sense.
>
> Best,
> Jonathan
>
> President, Chief Executive Officer,
> Sun Microsystems, Inc. 
>
> http://blogs.sun.com/jonathan/entry/one_plus_one_is_fifty
>
>   

Really very very interesting! This words reminded me the same dialogues  
and affair, between "Linus" and "Steve Jobs", that have been happened 
several years ago :-)

Regards,

Tarkan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jan Harkes]

On Wed, Jun 13, 2007 at 04:24:36PM +0200, Bernd Paysan wrote:
> There's no point of discussing that the Linux kernel *as a whole* (as 
> compilation) currently is under GPLv2 only, since it sais so, and a few 
> files also explicitely say so. The whole combination is GPLv2 only, but 
> most parts aren't.

You claim that any source files without a notices are 'any version of
the GPL'. But I read the license and you are totally wrong about that.

The GPL applies to "the Program" which in this case is the Linux kernel
as a whole and it in fact does indicate a specific version. All code
submitted and included in this program has has been submitted with the
understanding that the work as a whole is specifically licensed as
GPLv2. Some authors have granted additional rights, such as dual BSD/GPL
or GPLv2 and later and explicitly added such a notice.

All other code is simply copyrighted, and the only available license is
the GPLv2. Take for example fs/inode.c. Notice how it doesn't have GPL
boilerplate, but it is clearly indicating that it is copyrighted. So
taking that file by itself out of the context of the kernel and then
distributing it would clearly be a copyright violation. The only one
reason you can distribute that code is because of the GPLv2 that covers
the Linux kernel (i.e. "the Program").

> > > So my conclusion is: If you, as contributor to the Linux kernel, want
> > > to make clear that your work really is GPLv2 only, you have to do that
> > > yourself, you have to add a notice like above to files where you
> > > exclusively own copyright.

The kernel is explicitly licensed as GPLv2, any contributions (source
files/parts of the work) that wish to grant additional rights have to
specify so explicitly, and not the other way around however much you'd
like that.

> patches as the mainline kernel), I'm free to choose under which conditions 
> I redistribute it, given that it's compatible with the conditions the 
> original authors have chosen. Most of them have said nothing (other than 
> implicitely that it's ok for them to put it under GPL, as they haven't 
> opposed to inclusion into the Linux kernel), some have said GPLv2 or later, 

Reread section 9 and consider that "the Program" is the Linux kernel,
which does explicitly state a version and does not include the "and any
later" option. Any source that does not explicitly specify additional
rights is GPLv2.

Jan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alexandre Oliva wrote:

> On Jun 12, 2007, Greg KH <greg@kroah.com> wrote:
> 
> > (see previous long thread about v3 and why the kernel developers
> > hate it, it all still applys to the final draft.)
> 
> You mean all the misunderstandings? ;-)

I see the smiley, but I hate it how the FSF thinks others are morons and 
cannot read or think for themselves.

Any time you disagree with the FSF, you "misunderstand" (insert 
condescending voice) the issue. 

_Please_ don't continue that idiocy. Disagreement and thinking that the 
FSF is controlling and putting its fingers where they don't belong is 
_not_ misunderstanding. It's just not "blind and unquestioning obedience".

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Rob Landley]

On Sunday 10 June 2007 16:42:35 David Schwartz wrote:
> > http://www.eweek.com/article2/0,1759,1915720,00.asp
> > has the answer. Quoting Linus:
> >
> > "If you want to license a program under any later version of the
> > GPL, you have
> > to state so explicitly. Linux never did."
> >
> > Hence, unless there's a "GPL 2 or later", all the "unspecified GPL" files
> > are GPL2 only.
>
> The GPL states the default position:
>
> "If the Program does not specify a version number of this License, you may
> choose any version ever published by the Free Software Foundation."
>
> Leaving the question of whether Linus's comment at the top of the license
> changes the default:
>
> "Also note that the only valid version of the GPL as far as the kernel is
> concerned is _this_ particular version of the license (ie v2, not v2.2 or
> v3.x or whatever), unless explicitly otherwise stated."
>
> So we have dueling defaults. The GPL says the default is any version.
> Linus' statement at the top of the GPL says the default is v2 only. It's
> not clear, at least to me, that there is any clear reason why one should
> win out over the other.

Except that Linux included a copy of the license it was distributed under in 
the tarball, which was GPLv2.  There's a case to be made that this DOES count 
as "selection of the license version".  It's hard to get more explicit than 
including the complete license text.

On top of that, Linus clarified his position back in 2000:
http://www.uwsg.iu.edu/hypermail/linux/kernel/0009.1/0096.html

This means that not only are his on contributions so licensed, but there's a 
strong case to be made that everything he merged since then is explicitly 
GPLv2 only when sourced from the Linux kernel.  As maintainer there's a good 
argument that he has a compliation copyright on the kernels he releases.  He 
was certainly accepting patches under GPLv2 only, and merging them into a 
GPLv2 only work.  Sourcing ANY code from that work and declaring it non-GPLv2 
is really fishy; much better to find a clean upstream source, such as the 
original authors.

Of course the FSF would very, very much like this not to have been the case, 
and they've been trying to wish it away ever since...

Rob

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Wed, 13 Jun 2007, Alexandre Oliva wrote:

>> On Jun 12, 2007, Greg KH <greg@kroah.com> wrote:
>> 
>> > (see previous long thread about v3 and why the kernel developers
>> > hate it, it all still applys to the final draft.)
>> 
>> You mean all the misunderstandings? ;-)

> I see the smiley, but I hate it how the FSF thinks others are morons and 
> cannot read or think for themselves.

Look, there was room for misunderstandings in earlier drafts of the
license.  Based on the public comments, the wording was improved.  I'd
like to think the issues that arose from misunderstandings of the
earlier drafts are no longer an issue.  Is it not so?

Keeping on making false claims about the license drafts can be one of
two things: misunderstandings, out of ambiguity in the text or
preconceptions, or ill intentions.  I'd rather believe it's the
former.

Now, of course you can look at the licenses and decide that you never
agreed with the spirit of the GPL in the first place, and that GPLv2
models better your intentions than GPLv3.

Your assessment about sharing of code between Linux and OpenSolaris
very much makes it seem like that the spirit of sharing, of letting
others run, study, modify and share the code as long as they respect
others' freedoms, has never been what moved you.  Rather, you seem to
perceive the GPL as demanding some form of payback, of contribution,
rather than the respect for others' freedoms that it requires.  In
fact, you said something along these lines yourself many months ago.

With this different frame of mind, it is not surprising at all that
you don't find GPLv3 a better license.  With different goals in mind,
reasonable people can reach different conclusions.  But claiming that
GPLv3 is changing the spirit of the license, or that it prohibits
certain kinds of software, is plain false.  In fact, the spirit has
always been described in its preamble, and it didn't change at all:
it's all about respecting others' freedoms.

Sure, this evokes a number of other nice behaviors in various players,
and it's clear to me that it's in these other nice behaviors that you
seek when you choose GPLv2.  There's nothing inherently wrong in that.

However, it seems to me that GPLv3 would do an even better job at
serving these goals than GPLv2, even if the holes v3 plugs that
enabled players to disrespect others' freedoms might steer away the
participants who are not willing to contribute, to really be part of
your community.  It's not like you lose much.

But the new defenses against disrespect for freedoms introduced in
GPLv3 may turn out to be very helpful, not only in protecting your
community from external threats, but also in strengthening
participation, as the benefits of participation outweight the
perceived costs of respecting others' freedoms.

It sure seems to me that trading some threats and non-contributors for
some more-committed participants is a good idea.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Krzysztof Halasa]

Bernd Paysan <bernd.paysan@gmx.de> writes:

> If you choose the GPL as license, the text of the GPL are the conditions. 
> Otherwise, the GPL would be pure nonsense (as would be any other license).

The licence can't redefine the copyright laws. It doesn't make it pure
nonsense BTW.

> Each version is given a distinguishing version number.  If the Program
> specifies a version number of this License which applies to it and "any
> later version", you have the option of following the terms and conditions
> either of that version or of any later version published by the Free
> Software Foundation.  If the Program does not specify a version number of
> this License, you may choose any version ever published by the Free Software
> Foundation."

That would be the case if "the Program" (the whole or individual file(s))
contained something like:
"you can redistribute it and/or modify it under the terms of the GNU
General Public License as published by the Free Software Foundation".

Obviously, you could then choose any version including the first one.

> This is normal contract law (you have to say "yes" to a new M$ EULA every 
> few months or so if you are an unlucky Windows user, and like to use their 
> patches),

This is very different (though unrelated) - patches are new work and
I hope you're free to use your MS Windows under old conditions if you
don't need the new work.

> contracts can change over time. The FSF is rather nice here, they 
> say users and contributors can make choices to which contract they use.

FSF has exactly nothing to say here (except that they've created
a useful licence). The author can choose whatever conditions he/she
likes.

> Given that the license you find with Linux is GPLv2, anyway, the comment 
> from Linus seems to be superfluous. The license already says which version 
> it is.

It seems so.

> But it has this upgrade option, and one possible interpretation of 
      ^^
> Linus' comment is "no, it doesn't have this update option".

It? What "it"?
I don't get it. If you say the licence is v2 only, then how can it have
options?

> If I use GPL as license, I'm under "GPL regime", i.e. the terms of the GPL 
> apply.

First, the local and international laws apply. It's not like selling your
soul to the devil.

> The GPL requires that I need to speak up explicitely if I want to 
> limit the choice of licenses - if I don't say anything, it defaults to "any 
> GPL". This is a restriction that goes from author to user, sinde the GPL 
> cuts away all middle-men (restrictions applied by middle-men can be 
> reverted).

... theoretically, by removing their work, perhaps. Back to reality...

> If I decide to build my own compilation of Linux kernel patches* (i.e. a -bp 
> kernel, like the -mm kernel is a different compilation of Linux kernel 
> patches as the mainline kernel), I'm free to choose under which conditions 
> I redistribute it, given that it's compatible with the conditions the 
> original authors have chosen. Most of them have said nothing (other than 
> implicitely that it's ok for them to put it under GPL, as they haven't 
> opposed to inclusion into the Linux kernel),

There is no assumption of "GPL", you can only assume GPL v2 as the kernel
is v2. And it's not left for assumptions anymore, see "signed-off-by" and
licence tags (though the tags often specify "GPL" when the actual
licence, as indicated by the author, is GPL v2).

> Now, I may rewrite those few "GPLv2 only" files, and 
> then I have a GPLv2-or later compatible linux-some.version-bp kernel.

Sure, you can rewrite all non "GPLv2 or later" code and have v3 Linux.
The problem is you think only "few" files are v2.
-- 
Krzysztof Halasa

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Rob Landley]

On Sunday 10 June 2007 15:32:42 debian developer wrote:
> On 6/10/07, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
> > > licensing under the GPLv3, though. All I've heard are shrill voices
> > > about "tivoization" (which I expressly think is ok) and panicked
> > > worries about
> >
> > GPLv2 probably forbids Tivoisation anyway. Which is good IMHO even if not
>
>               ^^^^^^^^
>
> Now that is a bit waving in the air. GPLv2 forbids Tivoisation
> theoretically but practically it didnt stop them doing it practically.

A law never stops anybody from doing anything.  Enforcing the law does.

First of all, let's not confuse civil with criminal law:

http://nielsenhayden.com/makinglight/archives/009034.html
>  The difference between tort—breach of private rights—and crime—commission
> of an offence designated as such by the State—is one of the key legal
> concepts which the pro lawyer understands and the fan lawyer does not

Most variants of copyright infringement are a civil, not criminal matter.  
This means the state has no interest in enforcing the, it's your job to sue 
for damages and a restraining order if you want to exercise these rights 
(some people choose not to), and you have to have standing (I.E. be a holder 
of one of the infringed copyrights, or a designated legal representative 
thereof) in order to sue.  If none of the copyright holders sue to stop it, 
then it doesn't get stopped no matter how blatantly infringing it is.  Did 
anybody even bother to send Tivo a cease and desist?

Erik Andersen burned himself out trying to enforce the copyrights on BusyBox 
before Pamela Jones hooked that project (and uClibc) up with the SFLC.  
Harald Welte's been burning the candle at both ends with gpl-violations.org, 
but he's focusing on stuff sold in Germany.

As for anti-tivoisation, you can make a case that your signed binary is a 
derived work of the GPL source code just like a non-signed binary is, 
therefore the signing key is part of the source code used to create that 
binary, therefore GPLv2 requires the signing key be handed over on request.  
(I don't know if you'd WIN, I just know you could reasonably argue it in 
court and probably get past the inevitable initial motions to dismiss.)

GPLv2 is a nice, elegant license.  It's not perfect but it's very simple for 
what it does.

GPLv3 is not simple, not elegant, and contains numerous of special cases.  
Lots of the programmers here have an instinctive aversion to it because it 
reads like bad code.  We don't necessarily have to program in legalese to 
sense bad code in that language, at least compared to a "good code" example 
we've been using for some time.

Rob

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Lennart Sorensen]

On Wed, Jun 13, 2007 at 05:11:16PM -0300, Alexandre Oliva wrote:
> On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> Now, of course you can look at the licenses and decide that you never
> agreed with the spirit of the GPL in the first place, and that GPLv2
> models better your intentions than GPLv3.

I believe a number of people don't think the GPL v3 is in the same
spirit as the GPL v2.  I guess it comes down to what people thought the
spirit of the GPL v2 was.  There certainly seems to be a variety of
opinions on that, and I am not sure the FSF's opinion on it agrees with
what most others believe, but that would be rather difficult to
determine.

--
Len Sorensen

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alexandre Oliva wrote:
> 
> Look, there was room for misunderstandings in earlier drafts of the
> license.  Based on the public comments, the wording was improved.  I'd
> like to think the issues that arose from misunderstandings of the
> earlier drafts are no longer an issue.  Is it not so?

No. The anti-DRM language is still there, and no, it was never a 
misunderstanding. Now it's been limited to "consumer devices" (after I 
pointed out some of the _obvious_ problems with the original language), 
and the only people who called anything a "misunderstanding" were the ones 
that tried to point to *other* points in the license altogether (ie there 
was also a "drm section", which didn't really seem to say anything much at 
all).

Rms calls it "tivoization", but that's a word he has made up, and a term I 
find offensive, so I don't choose to use it. It's offensive because Tivo 
never did anything wrong, and the FSF even acknowledged that. The fact 
that they do their hardware and have some DRM issues with the content 
producers and thus want to protect the integrity of that hardware.

The kernel license covers the *kernel*. It does not cover boot loaders and 
hardware, and as far as I'm concerned, people who make their own hardware 
can design them any which way they want. Whether that means "booting only 
a specific kernel" or "sharks with lasers", I don't care.

> Keeping on making false claims about the license drafts can be one of
> two things: misunderstandings, out of ambiguity in the text or
> preconceptions, or ill intentions.  I'd rather believe it's the
> former.

No, it was not the former. And I think the whole "the kernel developers 
misunderstand the license" crap that the FSF was saying (several times) 
was very trying to confuse the issue: the FSF knew damn well which part of 
the license was obnoxious, they just tried to confuse the issue by 
pointing to *another* part of the license.

And you're just parrotting their idiotic line.

> Now, of course you can look at the licenses and decide that you never
> agreed with the spirit of the GPL in the first place, and that GPLv2
> models better your intentions than GPLv3.

And this is again the same *disease*. You claim that I "misunderstood" the 
"spirit of the GPL".

Dammit, the GPL is a license. I understand it quite well. Probably better 
than most. The fact that the FSF then noticed that there were *other* 
things that they wanted to do, and that were *not* covered by the GPLv2, 
does *not* mean that they can claim that others "misunderstood" the 
license.

I understood it perfectly fine, and it fit my needs. So tell me: who is 
the more confused one: the one who chose the license fifteen years ago, 
and realized what it means legally, and still stands behind it? I don't 
think so.

> Your assessment about sharing of code between Linux and OpenSolaris
> very much makes it seem like that the spirit of sharing, of letting
> others run, study, modify and share the code as long as they respect
> others' freedoms, has never been what moved you.  Rather, you seem to
> perceive the GPL as demanding some form of payback, of contribution,
> rather than the respect for others' freedoms that it requires.  In
> fact, you said something along these lines yourself many months ago.

I have said *exactly* that many many times.

The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and 
you can use it without having to drink the kool-aid.

I've said that over and over again. It's the "spirit of the GPLv2". It's 
what has made it such a great license, that lots of people (and companies) 
can use, is very fundamentally that it's fair.

The fact that the FSF sees *another* spirit to it is absolutely not a 
reason to say that I'm "confused". Quite frankly, apparently I'm _less_ 
confused than they are, since I saw the GPLv2 for what it was, and they 
did not - and as a result they felt they needed to extend upon it, because 
the license didn't actually match what they thought it would do.

> In fact, the spirit has always been described in its preamble, and it 
> didn't change at all: it's all about respecting others' freedoms.

That's a lot of bullshit. You are apparently the grand poobah, and can 
decide _which_ freedoms and for _what_ others' that matter.

I respect peoples freedoms too. I just disagree with the FSF on what that
slippery word means.

The fact that you are unable to even apparently fathom this fundamental 
issue, and that the FSF thinks that they own the definition of "freedom" 
is _your_ problem.

You're acting like some Alice-in-Wonderland character, saying that your 
definition of words is the only one that matter. And that others are 
"confused". Read up on your humpty-dumpty some day. 

I'm damn fed up with the FSF being the "protector of freedoms", and also 
feeling that they can define what those freedoms mean.

The GPLv2 is a *legal*license*. And no, the FSF doesn't get to define what 
the words mean to suit their agenda.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> find offensive, so I don't choose to use it. It's offensive because Tivo 
> never did anything wrong, and the FSF even acknowledged that. The fact 

Not all of us agree with this for the benefit of future legal
interpretation.

> The GPLv2 is a *legal*license*. And no, the FSF doesn't get to define what 
> the words mean to suit their agenda.

Agreed - everyone contributed to the kernel based upon the GPLv2. Lots of
different reasons, lots of different viewpoints about GPL2 v  GPL3, DRM ,
Treacherous Computing, etc. The commonality is not political, not a
grand plan, not a grand unified social agenda but a bunch of people for
whom the GPLv2 was an acceptable license for furthering their intentions
whether that is education for all, a shared commons or just making a
quick buck

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alan Cox wrote:
>
> > find offensive, so I don't choose to use it. It's offensive because Tivo 
> > never did anything wrong, and the FSF even acknowledged that. The fact 
> 
> Not all of us agree with this for the benefit of future legal
> interpretation.

Well, even the FSF lawyers did, but one of the reasons I never wanted to 
do the copyright assignments(*) is exactly because I think people need to 
make their own judgments on what the GPLv2 means. In the end, the only 
thing that really matters is what a judge says (after appeals etc), and 
the fact is, any license will always have gray areas where people disagree 
about interpretation.

And I actually am of the very firm opinion that a world with gray areas 
(and purple, and pink, and green) is a hell of a lot better than one where 
everything is black-and-white. Only lawyers want a black-and-white world.

So I would actually *encourage* other people to sue over their GPLv2 
interpretations, as they have done in Germany (and as IBM has done in the 
US). I'd sue based on _my_ reading of it, but hey, while my opinion is 
obviously always correct, I recognize that I live in a world where not 
everybody else always sees that.

[ (*) Obviously, the *biggest* reason not to do copyright assignments is 
  that they are just a total pain in the ass to do, and cause tons of 
  totally pointless paperwork. So "Linus is lazy and not interested in 
  being a lawyer" is obviously the primary reason for the lack of 
  assignments. I'm just much happier with people owning their own code 
  outright. ]

Of course, I also realize that suing people over license violations is a 
big pain in the ass, and in that sense while I "encourage" people to 
assert their own copyrights, I would obviously also say that it's almost 
certainly not worth doing if it's in a "gray" area. But that, in the end, 
has to be the copyright owners own decision!

> > The GPLv2 is a *legal*license*. And no, the FSF doesn't get to define what 
> > the words mean to suit their agenda.
> 
> Agreed - everyone contributed to the kernel based upon the GPLv2. Lots of
> different reasons, lots of different viewpoints about GPL2 v  GPL3, DRM ,
> Treacherous Computing, etc. The commonality is not political, not a
> grand plan, not a grand unified social agenda but a bunch of people for
> whom the GPLv2 was an acceptable license for furthering their intentions
> whether that is education for all, a shared commons or just making a
> quick buck

Indeed. And it's _fine_ to even be in it "just to make a quick buck". We 
do want all kinds of input. I think the community is much healthier having 
lots of different reasons for people wanting to be involved, rather than 
concentrating on just some specific reason.

For some it's the technology. For some it's the license. For some it's 
just a thing to pass boredom. Others like to learn. Whatever. It's all 
good!

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Bernd Paysan <bernd.paysan@gmx.de> wrote:

> What I don't understand about the GPLv3 with keys is why that depends on the 
> use case. As user of commercial devices like company routers, firewalls and 
> such, which often are Linux based, I don't want them sealed by the vendor, 
> as well. An explicit statement is even worse than an implicit one (as in 
> the GPLv2, which has been tested in a German court by Harald Welte - 
> Siemens had to turn in the keys).

My personal guess as to the reasoning behind this decision is that
consumer devices are the ones that require most attention, mainly
because the home users are the ones with least (individual) power to
demand respect for their freedoms.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> I believe a number of people don't think the GPL v3 is in the same
> spirit as the GPL v2.  I guess it comes down to what people thought the
> spirit of the GPL v2 was.

So let's go back to the preamble, that provides motivations and some
guidance as to the interpretation of the legal text (i.e., the spirit
of the license):

  [...] the GNU General Public License is intended to guarantee your
  freedom to share and change free software--to make sure the software
  is free for all its users. [...]

  [...] Our General Public Licenses are designed to make sure that you have
  the freedom to distribute copies of free software (and charge for
  this service if you wish), that you receive source code or can get
  it if you want it, that you can change the software or use pieces of
  it in new free programs; and that you know you can do these things.

  To protect your rights, we need to make restrictions that forbid
  anyone to deny you these rights or to ask you to surrender the
  rights.  These restrictions translate to certain responsibilities
  for you if you distribute copies of the software, or if you modify
  it.

  [...] if you distribute copies of such a program, whether gratis or
  for a fee, you must give the recipients all the rights that you have


Can anyone show me how any of the provisions of GPLv3 fails to meet
this spirit?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alexandre Oliva wrote:
> 
>   [...] Our General Public Licenses are designed to make sure that you have
>   the freedom to distribute copies of free software (and charge for
>   this service if you wish), that you receive source code or can get
>   it if you want it, that you can change the software or use pieces of
>   it in new free programs; and that you know you can do these things.
> 
>   To protect your rights, we need to make restrictions that forbid
>   anyone to deny you these rights or to ask you to surrender the
>   rights.  These restrictions translate to certain responsibilities
>   for you if you distribute copies of the software, or if you modify
>   it.
> 
>   [...] if you distribute copies of such a program, whether gratis or
>   for a fee, you must give the recipients all the rights that you have
> 
> 
> Can anyone show me how any of the provisions of GPLv3 fails to meet
> this spirit?

What kind of logic is that? It sounds like "Can you prove that God doesn't 
exist?"

The fact is, Tivo didn't take those rights away from you, yet the FSF says 
that what Tivo did was "against the spirit". That's *bullshit*.

So the whole "to protect these rights, we take away other rigths" argument 
hinges on the false premise that the new language in GPLv3 is somehow 
needed. It's not. You still had the right to distribute the software (and 
modify it), even if the *hardware* is limited to only one version.

In other words, GPLv3 restricts rights that do not need to be restricted, 
and yes, I think that violates the spirit of the GPLv2 preamble!

Think of it this way: what if the GPLv3 had an addition saying "You can 
not use this software to make a weapon". Do you see the problem? It 
restricts peoples rights, would you agree? Would you _also_ agree that it 
doesn't actually follow that "To protect your rights" logic AT ALL?

And this is exactly where the GPLv3 *diverges* from the above logic. If I 
build hardware, and sell it with software installed, you can still copy 
and modify the software. You may not do so within the confines of the 
hardware I built, but the hardware was never under the license in the 
first place.

In other words, GPLv3 *restricts* peoples freedoms more than it protects 
them. It does *not* cause any additional stated freedoms - quite the 
reverse. It tries to free up stuff that was never mentioned in the first 
place.

And then the FSF has the gall to call themselves the "protector of 
freedoms", and claim that everybody else is evil. What a crock. 

In other words, if you want to argue for the changes in GPLv3, you need to 
CHANGE THE PREAMBLE TOO! You should change:

	When we speak of free software, we are referring to freedom, not
	price.  Our General Public Licenses are designed to make sure that you
	have the freedom to distribute copies of free software (and charge for
	this service if you wish), that you receive source code or can get it
	if you want it, that you can change the software or use pieces of it
	in new free programs, that you can do so in place on your devices, 
        even if those devices weren't licensed under the GPL;  and that 
        you know you can do these things.

where I added the "that you can do so in place on your devices, even if 
those devices weren't licensed under the GPL".

That wasn't there in the original. Yet it's what the GPLv3 tries to shove 
down our throats in the name of "freedom".

I don't know if you've followed US politics very much over the last six 
years, but there's been a lot of "protecting our freedoms" going on. And 
it's been ugly. Maybe you could realize that sometimes "protecting your 
freedom" is *anything*but*!

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Wed, 13 Jun 2007, Alexandre Oliva wrote:
>> 
>> Look, there was room for misunderstandings in earlier drafts of the
>> license.  Based on the public comments, the wording was improved.  I'd
>> like to think the issues that arose from misunderstandings of the
>> earlier drafts are no longer an issue.  Is it not so?

> No. The anti-DRM language is still there, and no, it was never a 
> misunderstanding.

It was claimed that GPLv3 would forbid implementations of DRM.  That's
just plain false.  If you don't think so, please show what terms in
the latest draft prohibit DRM (as opposed to merely making it
ineffective, a necessary consequence of abiding by the spirit of all
GNU GPLs)

> It's offensive because Tivo never did anything wrong, and the FSF
> even acknowledged that.

Another misunderstanding.  The FSF never said TiVo didn't do anything
wrong.  It only said it didn't think there was a license violation.  I
personally disagree with that assessment, but IANAL.

Anyhow, deciding whether it's right or wrong is not the same as
deciding whether it's legal or illegal.  Law doesn't define what's
right or wrong.  That's what morals and ethics do.

> want to protect the integrity of that hardware.

> The kernel license covers the *kernel*.

When they choose to include a copy of the kernel in their hardware
that they can modify but others can't, they're failing to comply with
the spirit of the license.  For brevity, I won't repeat the quotes
from the GPLv2 preamble, that I just included in the message I sent to
Lennart Sorensen in this same thread.  Can you justify how you came to
the conclusion (if you did) that TiVo is abiding by the spirit of the
license?

>> Keeping on making false claims about the license drafts can be one of
>> two things: misunderstandings, out of ambiguity in the text or
>> preconceptions, or ill intentions.  I'd rather believe it's the
>> former.

> No, it was not the former.

Wow, I didn't see that coming.  Public admission of ill intentions?
;-) :-D :-P

> And I think the whole "the kernel developers misunderstand the
> license" crap that the FSF was saying (several times) was very
> trying to confuse the issue: the FSF knew damn well which part of
> the license was obnoxious, they just tried to confuse the issue by
> pointing to *another* part of the license.

Let me see if I got this right.  There was a section entitled
"3. Digital Restrictions Management" in GPLv3dd1.  Are you saying
that, when people complained about the DRM clause, they actually meant
the provisions in "1. Source Code", that established a requirement to
include the source code corresponding to functional signatures, namely
the signing keys, as part of the corresponding source code?

> And you're just parrotting their idiotic line.

Please watch your tone.  If you find offense at the allegedly
condescending tone in which the FSF says "misunderstanding", how do
you expect me and the FSF to take this?

It is also odd that you claim the right to be entitled to your own
opinion and reading about stuff, while denying myself the same right.
Please don't do that.  I have a mind of my own, and the fact that I
reach similar conclusions doesn't make me a parrot.  Even more so when
I actually have some influence on those conclusions.

>> Now, of course you can look at the licenses and decide that you never
>> agreed with the spirit of the GPL in the first place, and that GPLv2
>> models better your intentions than GPLv3.

> And this is again the same *disease*. You claim that I "misunderstood" the 
> "spirit of the GPL".

> Dammit, the GPL is a license. I understand it quite well. Probably better 
> than most. The fact that the FSF then noticed that there were *other* 
> things that they wanted to do, and that were *not* covered by the GPLv2, 
> does *not* mean that they can claim that others "misunderstood" the 
> license.

> I understood it perfectly fine, and it fit my needs. So tell me: who is 
> the more confused one: the one who chose the license fifteen years ago, 
> and realized what it means legally, and still stands behind it? I don't 
> think so.

You are definitely confused.  You're talking about the legal terms,
while I'm talking about the spirit.  The legal terms tried to reflect
the spirit as best as they could, but they left some holes.  Some
people found them and started exploiting them.

Sure, if you want to leave those holes unplugged in your code, that's
your decision.  I don't doubt that the GPLv2 legal terms fit the bill
for you.  I think GPLv3 would do even better in this regard.  But none
of this is about the spirit of the GPL.  Claiming GPLv3 changes the
spirit is totally missing the point of what the spirit amounts to.
The spirit is described in the preamble, it's not the legal terms.

> The beauty of the GPLv2 is exactly that it's a "tit-for-tat"
> license,

Ok, let's explore this argument.  In what sense is it tit-for-tat?
What is tit-for-tat about it?  What is the payback an author who
releases software under the GPL can legitimately expect to get?

>> In fact, the spirit has always been described in its preamble, and it 
>> didn't change at all: it's all about respecting others' freedoms.

> That's a lot of bullshit. You are apparently the grand poobah, and can 
> decide _which_ freedoms and for _what_ others' that matter.

The freedoms I'm talking about are very clearly described in the
spirit (preamble) of the license you chose for your project.  Go look
at the preamble one more time, "grand poobah" ;-)

  [...] the GNU General Public License is intended to guarantee your
  freedom to share and change free software--to make sure the software
  is free for all its users [...]

  [...] Our General Public Licenses are designed to make sure that you
  have the freedom to distribute copies of free software (and charge
  for this service if you wish), that you receive source code or can
  get it if you want it, that you can change the software or use
  pieces of it in new free programs; and that you know you can do
  these things.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Wed, 13 Jun 2007, Alan Cox wrote:

>> > find offensive, so I don't choose to use it. It's offensive because Tivo 
>> > never did anything wrong, and the FSF even acknowledged that. The fact 

>> Not all of us agree with this for the benefit of future legal
>> interpretation.

> Well, even the FSF lawyers did,

Or rather they didn't think an attempt to enforce that in the US would
prevail (or so I'm told).  That's not saying what TiVo did was right,
and that's not saying that what TiVo did was permitted by the license.
Only courts of law can do that.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Jörn Engel]

On Wed, 13 June 2007 14:33:07 -0700, Linus Torvalds wrote:
> 
> The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and 
> you can use it without having to drink the kool-aid.

One could even add that "tit-for-tat" appears to be the best strategy
in game theory for continuous runs of the prisoners dilemma.  At times I
wonder why game theory isn't taught in schools yet - it might shorten
discussions like these.

Jörn

-- 
All art is but imitation of nature.
-- Lucius Annaeus Seneca

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > On Wed, 13 Jun 2007, Alan Cox wrote:
> >> > find offensive, so I don't choose to use it. It's offensive because
> >> > Tivo never did anything wrong, and the FSF even acknowledged that. The
> >> > fact
> >>
> >> Not all of us agree with this for the benefit of future legal
> >> interpretation.
> >
> > Well, even the FSF lawyers did,
>
> Or rather they didn't think an attempt to enforce that in the US would
> prevail (or so I'm told).  That's not saying what TiVo did was right,
> and that's not saying that what TiVo did was permitted by the license.
> Only courts of law can do that.

Wrong! Anyone with half a brain can make the distinction. What TiVO did is 
entirely legal - they fully complied with the GPLv2. Note that what they 
*DON'T* allow people to do is run whatever version of whatever software they 
want on their hardware. They have that right - its the "Free Software 
Foundation" and the GPL - regardless of version - is a *SOFTWARE* license. 
TiVO never stopped people from copying, modifying or distributing the code - 
what they did was say "The code is GPL'd, the hardware is restricted" - 
ie: "You can do what you want with the code, but you can only run compiled 
version of it that we provide on our hardware". Why is that legal? Because 
TiVO produces the hardware and sells it to you with a certain *LICENSE* - 
because it does contain hardware covered under any number of patents. That 
license grants you the right to use the patents - in this case algorithms - 
provided you comply with the terms of the license. (Just like the GPL gives 
you the right to copy, modify and distribute GPL'd code as long as you comply 
with its terms)

If you believe otherwise then you are sadly mistaken. Now stop parroting the 
FSF's worn and tired tripe.

DRH
PS: Looking at your .sig I guess maybe you can't do that without getting 
kicked out of the FSF-LA

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Wed, 13 Jun 2007, Alexandre Oliva wrote:
>> 
>> [...] Our General Public Licenses are designed to make sure that you have
>> the freedom to distribute copies of free software (and charge for
>> this service if you wish), that you receive source code or can get
>> it if you want it, that you can change the software or use pieces of
>> it in new free programs; and that you know you can do these things.
>> 
>> To protect your rights, we need to make restrictions that forbid
>> anyone to deny you these rights or to ask you to surrender the
>> rights.  These restrictions translate to certain responsibilities
>> for you if you distribute copies of the software, or if you modify
>> it.
>> 
>> [...] if you distribute copies of such a program, whether gratis or
>> for a fee, you must give the recipients all the rights that you have
>> 
>> 
>> Can anyone show me how any of the provisions of GPLv3 fails to meet
>> this spirit?

> What kind of logic is that? It sounds like "Can you prove that God doesn't 
> exist?"

By this reasoning, it sounds like you've been claiming that "God does
exist", even though you can't prove it.

It shouldn't be anywhere that difficult to show that the GPLv3 fails
to meet the spirit of the GPLs.  You just have to show a single
counter-example.  Since there are so many objections to the changes,
it shouldn't be that hard.  Can you at least try?

> The fact is, Tivo didn't take those rights away from you, yet the FSF says 
> that what Tivo did was "against the spirit". That's *bullshit*.

Oh, good, let's take this one.

  if you distribute copies of such a program, [...]
  you must give the recipients all the rights that you have

So, TiVo includes a copy of Linux in its DVR.  

TiVo retains the right to modify that copy of Linux as it sees fit.

It doesn't give the recipients the same right.

Oops.

Sounds like a violation of the spirit to me.

Sounds like plugging this hole would retain the same spirit.

> In other words, GPLv3 restricts rights that do not need to be restricted, 

That's correct.  They don't need to be restricted.  The whole idea of
copyleft, implemented through the GPL, is not based on needs, but
rather on the wish to defend the freedoms established in the preamble
from those who would rather not respect them.

Do you deny that TiVo prevents you (or at least a random customer)
from modifying the copy of Linux that they ship in their DVR?

Do you deny that they can still do it themselves?

> Think of it this way: what if the GPLv3 had an addition saying "You can 
> not use this software to make a weapon".

This would make GPLv3 a non-Free Software license.

But the GPLv3 last call draft doesn't say anything along these lines.

You can use the software as much as you like, even in DVRs, and even
to implement DRM in them, as long as you respect the users' freedoms
to change and share the software.  Per the GPLv3 (paraphrased), if it
is possible to install modified versions of the covered program in the
device, you must tell the recipient how to do it.  Otherwise, the
freedom to modify the program is being too severely limited.

And, in the particular case of TiVo, it's a case of distributing
incomplete source code, of refraining from including functional
portions of the source code.

> In other words, GPLv3 *restricts* peoples freedoms more than it
> protects them.

While you look at it from the point of view of TiVo, who wants to be
free to prohibit people from modifying the workings of the device it
sells while it can still modify it itself, and it does that in order
to prohibit people from removing locks that stop them from doing
things they're legally entitled to do, I see a lot more prohibitions
than freedoms in what TiVo does.  I don't understand why you'd stand
up for it.  Is it more important that a single company be allowed to
impose prohibitions on others in order for its business model to work,
than to maintain the spirit of hacking and sharing that enabled Free
Software and Linux to flourish?

Do you expect Linux would have flourished if computers had locks that
stopped people from modifying Linux in them?

> where I added the "that you can do so in place on your devices, even if 
> those devices weren't licensed under the GPL".

You're mistakenly focusing on the device.  As you say, the device is
not under the license.

What's under the license is the software in it.  And that license
spirit requires the distributor to pass on the right to modify the
software.

> I don't know if you've followed US politics very much over the last
> six years, but there's been a lot of "protecting our freedoms" going
> on. And it's been ugly. Maybe you could realize that sometimes
> "protecting your freedom" is *anything*but*!

Is this why you're overreacting?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 19:35:41 Jörn Engel wrote:
> On Wed, 13 June 2007 14:33:07 -0700, Linus Torvalds wrote:
> > The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and
> > you can use it without having to drink the kool-aid.
>
> One could even add that "tit-for-tat" appears to be the best strategy
> in game theory for continuous runs of the prisoners dilemma.  At times I
> wonder why game theory isn't taught in schools yet - it might shorten
> discussions like these.
>
> Jörn

With the sheer amount of sheeple[1] in the world (and on this list), I doubt 
anything could make these discussions any shorter.

(While I hate thinking that sheeple are on this list, it is an unavoidable 
fact. (I had hoped I wouldn't find any sheeple here, as my favorite theory is 
that they are all "Fundamentalist Christians" like the "Creationist" fools))

DRH
1: Sheeple (n): People that act like sheep - ie: they cannot think or form 
opinions for themselves and always look to someone else for their thoughts 
and parrot the opinions of some "trusted" figure.

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Neil Brown]

On Wednesday June 13, dhazelton@enter.net wrote:
> 1: Sheeple (n): People that act like sheep - ie: they cannot think or form 
> opinions for themselves and always look to someone else for their thoughts 
> and parrot the opinions of some "trusted" figure.

I recommend that you avoid definitions like this.  Using them simply
makes you appear to have a very poor understanding of your fellow
humans.

"cannot" and "always" are absolutes that never apply (well, hardly
ever).

In my experience, most people do think for themselves and do form
opinions about areas where they have an interest/ability, and tend to
follow trusted others in areas where they have less interest or
ability.

The problem that I think you see is particularly the "parrot the
opinions" bit.  When a person tries to argue a case based largely on
the opinion of someone else with little personal understanding, they
are in very risky territory.   This is akin to 'fundamentalism' that
you also mentioned in your post.

Accusing people of arguing opinions that are not their own may well be
appropriate. Accusing them of not be able to think is not.

NeilBrown

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bongani Hlope]

On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:
> On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > The fact is, Tivo didn't take those rights away from you, yet the FSF
> > says that what Tivo did was "against the spirit". That's *bullshit*.
>
> Oh, good, let's take this one.
>
>   if you distribute copies of such a program, [...]
>   you must give the recipients all the rights that you have
>
> So, TiVo includes a copy of Linux in its DVR.
>

And they give you the same right that they had, which is obtain free software 
that you can modify and redistribute. There's nothing in there that says they 
should give you the tools they used after they received the software, which 
is what you seem to be looking for.

> TiVo retains the right to modify that copy of Linux as it sees fit.
>
> It doesn't give the recipients the same right.
>

It does, can't you modify their kernel source? Where does it say you should be 
able to run you modifications on the same hardware?

> Oops.
>
> Sounds like a violation of the spirit to me.
>
> Sounds like plugging this hole would retain the same spirit.

The only fear that I have with the whole Tivo saga, is that companies like 
Dell can use the same thing to say: "Our hardware will only run Company's X 
distribution of Linux". 

Do we just hope users won't buy those Dell machines, or do you modify your 
software license to force Dell to allow custom distributions to run on their 
machines? Then where do we draw the line of "Software Licenses".

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:
<snip>
> > The fact is, Tivo didn't take those rights away from you, yet the FSF
> > says that what Tivo did was "against the spirit". That's *bullshit*.
>
> Oh, good, let's take this one.
>
>   if you distribute copies of such a program, [...]
>   you must give the recipients all the rights that you have
>
> So, TiVo includes a copy of Linux in its DVR.
>
> TiVo retains the right to modify that copy of Linux as it sees fit.
>
> It doesn't give the recipients the same right.
>
> Oops.
>
> Sounds like a violation of the spirit to me.
>
> Sounds like plugging this hole would retain the same spirit.

Are you an idiot, or do you just choose to ignore all proof that doesn't fit 
your preconceived beliefs? TiVO gives you every right to the Linux kernel 
that they recieved. What they don't give you the right to do is use modified 
versions on their *HARDWARE* - which they have *NEVER* given you any rights 
to, except for "normal use". (And no, it isn't legal to put those 200G hard 
drives in your TiVO no matter what you think) 

> > In other words, GPLv3 restricts rights that do not need to be restricted,
>
> That's correct.  They don't need to be restricted.  The whole idea of
> copyleft, implemented through the GPL, is not based on needs, but
> rather on the wish to defend the freedoms established in the preamble
> from those who would rather not respect them.
>
> Do you deny that TiVo prevents you (or at least a random customer)
> from modifying the copy of Linux that they ship in their DVR?

Exactly. They don't. What TiVO prevents is using that modified version on 
their hardware. And they have that right, because the Hardware *ISN'T* 
covered by the GPL.

Do you understand that, or do I need to break out the finger-puppets next ?

> Do you deny that they can still do it themselves?
>
> > Think of it this way: what if the GPLv3 had an addition saying "You can
> > not use this software to make a weapon".
>
> This would make GPLv3 a non-Free Software license.
>
> But the GPLv3 last call draft doesn't say anything along these lines.
>
> You can use the software as much as you like, even in DVRs, and even
> to implement DRM in them, as long as you respect the users' freedoms
> to change and share the software.  Per the GPLv3 (paraphrased), if it
> is possible to install modified versions of the covered program in the
> device, you must tell the recipient how to do it.  Otherwise, the
> freedom to modify the program is being too severely limited.

And this unnaturally restricts the freedom of hardware manufacturers. If they 
add a custom, internal connector so a repair shop can restore the hardware to 
its *FACTORY* state then it is "possible to install modified versions", 
provided the person doing it has the specialized hardware needed.

And this is what the FSF, RMS and yes, *YOU*, Alexandre, fail to realize - the 
GPL covers *ONLY* the software. It has *ZERO* legal standing when applied to 
hardware. Not even the most draconian of MS EULA's tries to apply itself to 
the hardware.

In the case of 99% of the hardware targeted by the clause of the GPLv3 you 
elucidate on, the "ability to install modified versions of the software" was 
*NOT* intended for that use, nor was it intended for *ANYONE* *EXCEPT* 
trained service personell to have *ACCESS* to that functionality. Arguing 
otherwise is just idiotic - I have never found a piece of "high tech" 
hardware (like a TiVO) that was designed for the end-user to modify. (yes, 
installing a new version of the linux kernel is "modifying" the system)

> And, in the particular case of TiVo, it's a case of distributing
> incomplete source code, of refraining from including functional
> portions of the source code.

And? They distribute the kernel source - as they recieved it - in compliance 
with the GPL. Their additions - whether they be "modules" or just the UI - do 
not, necessarily, fall under the GPL. (Yes, there have been discussions about 
whether a kernel module is a derived work, but most of the time those 
discussions ended "Legally they aren't, even though I feel they should be")

> > In other words, GPLv3 *restricts* peoples freedoms more than it
> > protects them.
>
> While you look at it from the point of view of TiVo, who wants to be
> free to prohibit people from modifying the workings of the device it
> sells while it can still modify it itself, and it does that in order
> to prohibit people from removing locks that stop them from doing
> things they're legally entitled to do, I see a lot more prohibitions
> than freedoms in what TiVo does.  I don't understand why you'd stand
> up for it.  Is it more important that a single company be allowed to
> impose prohibitions on others in order for its business model to work,
> than to maintain the spirit of hacking and sharing that enabled Free
> Software and Linux to flourish?

What "Legally Entitled" things?

And... You do realize that almost every difference between the GPLv2 and the 
GPLv3 is going to cause a hell of a lot of problems? The fact that the GPLv3 
is designed to prevent things that RMS *PERSONALLY* finds distasteful - DRM 
and the like - is a big turn-off for a *LOT* of people. (Personally I don't 
like *ANY* version of the GPL, because there are chunks I have problems with)

> Do you expect Linux would have flourished if computers had locks that
> stopped people from modifying Linux in them?

But you aren't talking about a "computer" here. You're talking about a 
mass-market device that must comply with both US and International copyright 
law - and that's just a TiVO. Other devices have other laws they have to 
comply with - in the US the FCC's regulations control all radio devices, so 
if you upload a modified linux kernel to your wireless router that gives it a 
2000 foot range, you've just broken the law *AND* violated the license on the 
hardware which states that you "won't modify it or the controlling 
software" - in most cases "the controlling software" is just the firmware, 
but with modern wireless hardware, the firmware is loaded by the OS.

> > where I added the "that you can do so in place on your devices, even if
> > those devices weren't licensed under the GPL".
>
> You're mistakenly focusing on the device.  As you say, the device is
> not under the license.

But he isn't. The GPLv3 says, and I'll quote you here - "Per the GPLv3 
(paraphrased), if it is possible to install modified versions of the covered 
program in the device, you must tell the recipient how to do it."

>From the latest version of the GPLv3:
"Installation Information" for a User Product means any methods, procedures, 
authorization keys, or other information required to install and execute 
modified versions of a covered work in that User Product from a modified 
version of its Corresponding Source. The information must suffice to ensure 
that the continued functioning of the modified object code is in no case 
prevented or interfered with solely because modification has been made.

and:

If you convey an object code work under this section in, or with, or 
specifically for use in, a User Product, and the conveying occurs as part of 
a transaction in which the right of possession and use of the User Product is 
transferred to the recipient in perpetuity or for a fixed term (regardless of 
how the transaction is characterized), the Corresponding Source conveyed 
under this section must be accompanied by the Installation Information. But 
this requirement does not apply if neither you nor any third party retains 
the ability to install modified object code on the User Product (for example, 
the work has been installed in ROM).

So it's not just a designed in ability to run modified code - ie: running of 
modified code is a feature meant for the user to take advantage of - but even 
things like the connectors used to upload the operating software at the 
factory that people now cannot have in a device that runs GPL(v3) covered 
software unless they ship the related "Installation Information". In other 
words, companies are no longer allowed to have a completely separate license  
for hardware and software.

That, to me, reads like RMS got mad about TiVO and said "I don't like it, lets 
add a clause making it wrong to the next GPL". Hell, that *IS* what happened, 
and nothing the FSF or Eben Moglen says will convince me otherwise. It's the 
same for the bits that were added after Novell signed their agreement with 
MS. 

> What's under the license is the software in it.  And that license
> spirit requires the distributor to pass on the right to modify the
> software.

And since when did they have to enable people to use their hardware in 
violation of the licensing agreement they implicitly agree to when opening 
the package?

There is *NOTHING* stopping you from doing whatever you want with the code 
that runs on a TiVO (or any similar device). You (and everyone that thinks 
like you) are confusing a want to use the *HARDWARE* however you want with 
your GPL granted "right" to do what you want with the *SOFTWARE*.

> > I don't know if you've followed US politics very much over the last
> > six years, but there's been a lot of "protecting our freedoms" going
> > on. And it's been ugly. Maybe you could realize that sometimes
> > "protecting your freedom" is *anything*but*!
>
> Is this why you're overreacting?

No, he's making a point. RMS and the FSF, in drafting GPLv3 to include the 
language and clauses it does, is "protecting your freedom" the way a lot of 
the post 9/11 changes to the US Federal code does it. (ie: by saying "no, you 
can't do that anymore")

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > >> > find offensive, so I don't choose to use it. It's offensive because
> > >> > Tivo never did anything wrong, and the FSF even acknowledged that. The
> > >> > fact
> > >>
> > >> Not all of us agree with this for the benefit of future legal
> > >> interpretation.
> > >
> > > Well, even the FSF lawyers did,
> >
> > Or rather they didn't think an attempt to enforce that in the US would
> > prevail (or so I'm told).  That's not saying what TiVo did was right,
> > and that's not saying that what TiVo did was permitted by the license.
> > Only courts of law can do that.
> 
> Wrong! Anyone with half a brain can make the distinction. What TiVO did is 
> entirely legal - they fully complied with the GPLv2. Note that what they 
> *DON'T* allow people to do is run whatever version of whatever software they 
> want on their hardware. They have that right - its the "Free Software 
> Foundation" and the GPL - regardless of version - is a *SOFTWARE* license. 
>...

The GPLv2 says:

"For an executable work, complete source code means all the source code 
for all modules it contains, plus any associated interface definition 
files, plus the scripts used to control compilation and installation of 
the executable."

The question is whether this includes private keys.
Different people have different opinions regarding this issue.

If "the complete source code" includes private keys, the GPLv2 requires 
them to give any costumer the private keys.

Fact is that Harald Welte did in several cases successfully convince 
vendors that private keys are part of the source code if they are 
required for running the compiled binary on some hardware.

AFAIK there haven't been any court rulings on this issue, and it could 
even be that courts in different countries will decide differently.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 20:14:47 Neil Brown wrote:
> On Wednesday June 13, dhazelton@enter.net wrote:
> > 1: Sheeple (n): People that act like sheep - ie: they cannot think or
> > form opinions for themselves and always look to someone else for their
> > thoughts and parrot the opinions of some "trusted" figure.
>
> I recommend that you avoid definitions like this.  Using them simply
> makes you appear to have a very poor understanding of your fellow
> humans.
>
> "cannot" and "always" are absolutes that never apply (well, hardly
> ever).

In this case I gave a very hasty definition of the term, though it does fit 
very well. 

> In my experience, most people do think for themselves and do form
> opinions about areas where they have an interest/ability, and tend to
> follow trusted others in areas where they have less interest or
> ability.

Agreed. But those *aren't* "Sheeple".

> The problem that I think you see is particularly the "parrot the
> opinions" bit.  When a person tries to argue a case based largely on
> the opinion of someone else with little personal understanding, they
> are in very risky territory.   This is akin to 'fundamentalism' that
> you also mentioned in your post.
>
> Accusing people of arguing opinions that are not their own may well be
> appropriate. Accusing them of not be able to think is not.

I never accused any specific person of either. I was making an observation 
about the general nature of humanity as I've observed it. A better definition 
of "Sheeple" would be "Someone who continually parrots some "trusted" figures 
opinion even when presented with proof that that opinion is in error. They 
also, generally, do not "think deeply" on any topic."

And while I won't apologize to anyone that may have felt the "Sheeple" remark 
was directed at them, I will apologize for not taking more time to choose a 
better definition for the term originally.

DRH

> NeilBrown
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Thu, Jun 14, 2007 at 02:15:53AM +0200, Bongani Hlope wrote:
> On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:
> > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > The fact is, Tivo didn't take those rights away from you, yet the FSF
> > > says that what Tivo did was "against the spirit". That's *bullshit*.
> >
> > Oh, good, let's take this one.
> >
> >   if you distribute copies of such a program, [...]
> >   you must give the recipients all the rights that you have
> >
> > So, TiVo includes a copy of Linux in its DVR.
> 
> And they give you the same right that they had, which is obtain free software 
> that you can modify and redistribute. There's nothing in there that says they 
> should give you the tools they used after they received the software, which 
> is what you seem to be looking for.
>...

Wrong, the GPLv2 says:

"For an executable work, complete source code means all the source code 
for all modules it contains, plus any associated interface definition 
files, plus the scripts used to control compilation and installation of 
the executable."

The question is whether this includes private keys.
Different people have different opinions regarding this issue.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Bongani Hlope <bhlope@mweb.co.za> wrote:

> On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:

>> if you distribute copies of such a program, [...]
>> you must give the recipients all the rights that you have

>> So, TiVo includes a copy of Linux in its DVR.

> And they give you the same right that they had, which is obtain free software 
> that you can modify and redistribute. There's nothing in there that says they 
> should give you the tools they used after they received the software, which 
> is what you seem to be looking for.

Can they modify the software in their device?

Do they pass this right on?

>> TiVo retains the right to modify that copy of Linux as it sees fit.

>> It doesn't give the recipients the same right.

> It does, can't you modify their kernel source?

It's not the kernel source.  That's not where the TiVo anti-tampering
machinery blocks modifications.

It's about that copy of the kernel that ships in the device in object
code.  That's the one that TiVo customers ought to be entitled to
modify, if TiVo can modify it itself.

> Where does it say you should be able to run you modifications on the
> same hardware?

Where it says that you should pass on all the rights that you have.

While TiVo retains the ability to replace, upgrade, fix, break or make
any other change in the GPLed software in the device, it ought to pass
it on to its customers.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Jörn Engel <joern@logfs.org> wrote:

> On Wed, 13 June 2007 14:33:07 -0700, Linus Torvalds wrote:
>> 
>> The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and 
>> you can use it without having to drink the kool-aid.

> One could even add that "tit-for-tat" appears to be the best strategy
> in game theory for continuous runs of the prisoners dilemma.

It is, indeed.

Now the remaining piece of the proof is to show that the GPLv2 is
tit-for-tat.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Neil Brown <neilb@suse.de> wrote:

> Accusing people of arguing opinions that are not their own may well be
> appropriate. Accusing them of not be able to think is not.

I agree with the latter, but the former is seldom appropriate.  Any
accusation ought to be provable, and it's nearly impossible to prove
that an opinion held by someone is not his own.  People quite often
arrive at similar opinions independently.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 20:44:19 Adrian Bunk wrote:
> On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> > On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > > >> > find offensive, so I don't choose to use it. It's offensive
> > > >> > because Tivo never did anything wrong, and the FSF even
> > > >> > acknowledged that. The fact
> > > >>
> > > >> Not all of us agree with this for the benefit of future legal
> > > >> interpretation.
> > > >
> > > > Well, even the FSF lawyers did,
> > >
> > > Or rather they didn't think an attempt to enforce that in the US would
> > > prevail (or so I'm told).  That's not saying what TiVo did was right,
> > > and that's not saying that what TiVo did was permitted by the license.
> > > Only courts of law can do that.
> >
> > Wrong! Anyone with half a brain can make the distinction. What TiVO did
> > is entirely legal - they fully complied with the GPLv2. Note that what
> > they *DON'T* allow people to do is run whatever version of whatever
> > software they want on their hardware. They have that right - its the
> > "Free Software Foundation" and the GPL - regardless of version - is a
> > *SOFTWARE* license. ...
>
> The GPLv2 says:
>
> "For an executable work, complete source code means all the source code
> for all modules it contains, plus any associated interface definition
> files, plus the scripts used to control compilation and installation of
> the executable."
>
> The question is whether this includes private keys.
> Different people have different opinions regarding this issue.
>
> If "the complete source code" includes private keys, the GPLv2 requires
> them to give any costumer the private keys.
>
> Fact is that Harald Welte did in several cases successfully convince
> vendors that private keys are part of the source code if they are
> required for running the compiled binary on some hardware.

If the hardware was designed for the end-user to change the software running 
on it - including running software that it was never meant to run (ie: a 
complete webserver on cell phone) - then yes, the signing keys are a part of 
the source, as the software running on the device is designed to be updated 
by the user using the provided system.

If, on the other hand, the only "software updates" the user is expected to 
perform are the installation of newer versions of the existing code 
for "Security" or "Bug Fix" reasons then the signing keys aren't part of the 
source.

I haven't looked into what Harald Welte did, but I'd be surprised if someone 
tried following suit in America and had as much success.

>
> AFAIK there haven't been any court rulings on this issue, and it could
> even be that courts in different countries will decide differently.

Agreed.

DRH

>
> cu
> Adrian



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> Now stop parroting the FSF's worn and tired tripe.

Are you playing Linus' sheeple and parroting his lines just to make a
point, or are you like that all the time? ;-)

> PS: Looking at your .sig I guess maybe you can't do that without getting 
> kicked out of the FSF-LA

Don't worry, I'm not speaking even for FSFLA, and I'm entitled to my
own opinion.

I haven't consulted other FSFLA members about this.  This is all my
own personal opinion.

It just so happens that I'm very closely involved in the process, I've
spent a lot of time thinking about it, and I happen to share a similar
moral and ethical background with others involved in the process, so I
arrive at similar conclusions.

And then, I influence the process myself, so it's not like some of the
arguments I brought up here weren't taken into account while creating
the GPLv3, and adopted by its other proponents.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 20:55:52 Alexandre Oliva wrote:
> On Jun 13, 2007, Bongani Hlope <bhlope@mweb.co.za> wrote:
> > On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:
> >> if you distribute copies of such a program, [...]
> >> you must give the recipients all the rights that you have
> >>
> >> So, TiVo includes a copy of Linux in its DVR.
> >
> > And they give you the same right that they had, which is obtain free
> > software that you can modify and redistribute. There's nothing in there
> > that says they should give you the tools they used after they received
> > the software, which is what you seem to be looking for.
>
> Can they modify the software in their device?
>
> Do they pass this right on?

But this *ISN'T* a right that the GPLv2 *REQUIRES* be passed on. 

> >> TiVo retains the right to modify that copy of Linux as it sees fit.
> >>
> >> It doesn't give the recipients the same right.
> >
> > It does, can't you modify their kernel source?
>
> It's not the kernel source.  That's not where the TiVo anti-tampering
> machinery blocks modifications.
>
> It's about that copy of the kernel that ships in the device in object
> code.  That's the one that TiVo customers ought to be entitled to
> modify, if TiVo can modify it itself.

The GPLv2 makes no real provision for *DIRECTLY* modifying object code. What 
provisions the GPLv2 has apply to the source code.

And no, the end user *SHOULD* *NOT* be entitled to run whatever kernel they 
like on a TiVO. It was designed with the "install new kernel" functionality 
so that the TiVO corporation could update the kernel running on the hardware 
when security problems came up, when bugs were fixed or even when the new 
version gives better performance.

> > Where does it say you should be able to run you modifications on the
> > same hardware?
>
> Where it says that you should pass on all the rights that you have.
>
> While TiVo retains the ability to replace, upgrade, fix, break or make
> any other change in the GPLed software in the device, it ought to pass
> it on to its customers.

It *DOES* *NOT* say "All rights that you have". It says "All rights that are 
granted you by this license". If every piece of software released under the 
GPL had *ALL* rights passed on, then *ANYONE* could do the "I'm granting 
company X the right to use this software outside the GPL for $50,000USD." 
instead of just the *creator* of the software.

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> > Only courts of law can do that.
> 
> Wrong! Anyone with half a brain can make the distinction. What TiVO did is 

Maybe half a brain can, but anyone with a whole brain can assure you its
a bit more complex and you are wrong..

> version of it that we provide on our hardware". Why is that legal? Because 
> TiVO produces the hardware and sells it to you with a certain *LICENSE* - 

The keys required to make the code run with the hardware are part of the
software. The license requires the software and relevant scripts etc are
included. Thus there is a very good argument that the keys are part of
the software.

And since there is no court ruling to high enough level in the USA, UK or
any other jurisdiction on that it remains a matter of opinion.

Tivo may control the hardware but the authors control the software (via
the GPL), and subject to the limits of what may be specified by a
copyright license (as opposed to contract) can make such demands as they
see fit about their software and anything derivative of it.

> because it does contain hardware covered under any number of patents. That 
> license grants you the right to use the patents - in this case algorithms - 

You can't patent algorithms either

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 21:04:42 Alexandre Oliva wrote:
> On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > Now stop parroting the FSF's worn and tired tripe.
>
> Are you playing Linus' sheeple and parroting his lines just to make a
> point, or are you like that all the time? ;-)

Nope. I'm just tired of giving proof after proof that you're wrong and having 
you restate the same tripe.

> > PS: Looking at your .sig I guess maybe you can't do that without getting
> > kicked out of the FSF-LA
>
> Don't worry, I'm not speaking even for FSFLA, and I'm entitled to my
> own opinion.

Certainly. I never said otherwise. What I stated and then *implied* was that 
you are repeating the same false logic over and over again trying to make 
people believe that it isn't borked and that that false logic is exactly the 
same crap I've seen from the FSF numerous times.

> I haven't consulted other FSFLA members about this.  This is all my
> own personal opinion.

Where I am examining the facts and drawing a logical conclusion. That it 
happens to form an opinion is secondary to the truth.

> It just so happens that I'm very closely involved in the process, I've
> spent a lot of time thinking about it, and I happen to share a similar
> moral and ethical background with others involved in the process, so I
> arrive at similar conclusions.

Okay. Still doesn't explain why you have argued that the GPLv3 doesn't attempt 
to cover hardware and then provide proof that it does. 

> And then, I influence the process myself, so it's not like some of the
> arguments I brought up here weren't taken into account while creating
> the GPLv3, and adopted by its other proponents.

This is no surprise - I had a feeling this was the truth. Not that it changes 
my opinion at all. As I've said, I have never liked the GPL at all, but v2 is 
the best that exists - even though I've put together custom licenses myself, 
none of them have had the number of lawyers look at them that the GPLv2 has 
had.

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Wed, Jun 13, 2007 at 09:01:28PM -0400, Daniel Hazelton wrote:
> On Wednesday 13 June 2007 20:44:19 Adrian Bunk wrote:
> > On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> > > On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > > > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > > > >> > find offensive, so I don't choose to use it. It's offensive
> > > > >> > because Tivo never did anything wrong, and the FSF even
> > > > >> > acknowledged that. The fact
> > > > >>
> > > > >> Not all of us agree with this for the benefit of future legal
> > > > >> interpretation.
> > > > >
> > > > > Well, even the FSF lawyers did,
> > > >
> > > > Or rather they didn't think an attempt to enforce that in the US would
> > > > prevail (or so I'm told).  That's not saying what TiVo did was right,
> > > > and that's not saying that what TiVo did was permitted by the license.
> > > > Only courts of law can do that.
> > >
> > > Wrong! Anyone with half a brain can make the distinction. What TiVO did
> > > is entirely legal - they fully complied with the GPLv2. Note that what
> > > they *DON'T* allow people to do is run whatever version of whatever
> > > software they want on their hardware. They have that right - its the
> > > "Free Software Foundation" and the GPL - regardless of version - is a
> > > *SOFTWARE* license. ...
> >
> > The GPLv2 says:
> >
> > "For an executable work, complete source code means all the source code
> > for all modules it contains, plus any associated interface definition
> > files, plus the scripts used to control compilation and installation of
> > the executable."
> >
> > The question is whether this includes private keys.
> > Different people have different opinions regarding this issue.
> >
> > If "the complete source code" includes private keys, the GPLv2 requires
> > them to give any costumer the private keys.
> >
> > Fact is that Harald Welte did in several cases successfully convince
> > vendors that private keys are part of the source code if they are
> > required for running the compiled binary on some hardware.
> 
> If the hardware was designed for the end-user to change the software running 
> on it - including running software that it was never meant to run (ie: a 
> complete webserver on cell phone) - then yes, the signing keys are a part of 
> the source, as the software running on the device is designed to be updated 
> by the user using the provided system.
> 
> If, on the other hand, the only "software updates" the user is expected to 
> perform are the installation of newer versions of the existing code 
> for "Security" or "Bug Fix" reasons then the signing keys aren't part of the 
> source.

Are you an idiot, or do you just choose to ignore all proof that doesn't 
fit your preconceived beliefs?

The GPL doesn't give someone distributing the software the choice of how
much to limit the freedom of the user.

Either private keys required to run the kernel on the hardware are 
always considered part of "the complete source code" or they are never 
part of it.

> I haven't looked into what Harald Welte did, but I'd be surprised if someone 
> tried following suit in America and had as much success.
>...

Harald is in Germany, and he therefore takes legal action against people 
distributing products violating his copyright on the Linux kernel
in Germany at German courts based on German laws.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 21:16:19 Alan Cox wrote:
> > > Only courts of law can do that.
> >
> > Wrong! Anyone with half a brain can make the distinction. What TiVO did
> > is
>
> Maybe half a brain can, but anyone with a whole brain can assure you its
> a bit more complex and you are wrong..
>
> > version of it that we provide on our hardware". Why is that legal?
> > Because TiVO produces the hardware and sells it to you with a certain
> > *LICENSE* -
>
> The keys required to make the code run with the hardware are part of the
> software. The license requires the software and relevant scripts etc are
> included. Thus there is a very good argument that the keys are part of
> the software.

Good argument, but I'll stand by my interpretation of the law, the GPL and the 
situation until there is solid proof that a signing-key is part of the source 
code. Doubly so because the language of the GPLv2 makes it clear that "all 
relevant scripts, etc" are only needed to build and run the "covered work" - 
not for proper installation of it. (and, in the case of a TiVO, the signing 
keys are part of the installation, not the running or building. Besides 
needing the proper signing key, the kernel in a TiVO is run the same as any 
other Linux kernel) 

> And since there is no court ruling to high enough level in the USA, UK or
> any other jurisdiction on that it remains a matter of opinion.
>
> Tivo may control the hardware but the authors control the software (via
> the GPL), and subject to the limits of what may be specified by a
> copyright license (as opposed to contract) can make such demands as they
> see fit about their software and anything derivative of it.

Agreed. However, AFAICT, TiVO meets the provisions of the GPLv2 - they make 
the source of the GPL'd part of their system available. (And I'm not going to 
get into arguments over whether kernel modules are "derivative works" or not, 
since those invariably end up with "They aren't, even though we think they 
should be")

> > because it does contain hardware covered under any number of patents.
> > That license grants you the right to use the patents - in this case
> > algorithms -
>
> You can't patent algorithms either

Then explain the patents on the MP3 algorithm, the LZW algorithm, etc... Those 
patents are real and while the LZW one may have lapsed, still relevant.

DRH 

> Alan



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Chris Adams]

Once upon a time, Alexandre Oliva  <aoliva@redhat.com> said:
>  if you distribute copies of such a program, [...]
>  you must give the recipients all the rights that you have
>
>So, TiVo includes a copy of Linux in its DVR.  
>
>TiVo retains the right to modify that copy of Linux as it sees fit.
>
>It doesn't give the recipients the same right.

Sure it does; you received a program (the kernel) and you can modify it.
You also received hardware; they don't support modification of that.
Nowhere in the license does it say they have to, because the license
only covers the program.

Or are you claiming that putting software on hardware makes the result a
derivative work?  I think it falls under the "mere aggregation" clause.

What if TiVo had put the kernel in a burned-in ROM (not flash, or on a
flash ROM with no provision for reprogramming it)?  Would that also
violate the "spirit" of the GPL?  Must any device that wishes to include
GPL code include additional hardware to support replacing that code
(even if that hardware is otherwise superfluous)?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:
> On Wed, Jun 13, 2007 at 09:01:28PM -0400, Daniel Hazelton wrote:
> > On Wednesday 13 June 2007 20:44:19 Adrian Bunk wrote:
> > > On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> > > > On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > > > > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> 
wrote:
> > > > > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > > > > >> > find offensive, so I don't choose to use it. It's offensive
> > > > > >> > because Tivo never did anything wrong, and the FSF even
> > > > > >> > acknowledged that. The fact
> > > > > >>
> > > > > >> Not all of us agree with this for the benefit of future legal
> > > > > >> interpretation.
> > > > > >
> > > > > > Well, even the FSF lawyers did,
> > > > >
> > > > > Or rather they didn't think an attempt to enforce that in the US
> > > > > would prevail (or so I'm told).  That's not saying what TiVo did
> > > > > was right, and that's not saying that what TiVo did was permitted
> > > > > by the license. Only courts of law can do that.
> > > >
> > > > Wrong! Anyone with half a brain can make the distinction. What TiVO
> > > > did is entirely legal - they fully complied with the GPLv2. Note that
> > > > what they *DON'T* allow people to do is run whatever version of
> > > > whatever software they want on their hardware. They have that right -
> > > > its the "Free Software Foundation" and the GPL - regardless of
> > > > version - is a *SOFTWARE* license. ...
> > >
> > > The GPLv2 says:
> > >
> > > "For an executable work, complete source code means all the source code
> > > for all modules it contains, plus any associated interface definition
> > > files, plus the scripts used to control compilation and installation of
> > > the executable."
> > >
> > > The question is whether this includes private keys.
> > > Different people have different opinions regarding this issue.
> > >
> > > If "the complete source code" includes private keys, the GPLv2 requires
> > > them to give any costumer the private keys.
> > >
> > > Fact is that Harald Welte did in several cases successfully convince
> > > vendors that private keys are part of the source code if they are
> > > required for running the compiled binary on some hardware.
> >
> > If the hardware was designed for the end-user to change the software
> > running on it - including running software that it was never meant to run
> > (ie: a complete webserver on cell phone) - then yes, the signing keys are
> > a part of the source, as the software running on the device is designed
> > to be updated by the user using the provided system.
> >
> > If, on the other hand, the only "software updates" the user is expected
> > to perform are the installation of newer versions of the existing code
> > for "Security" or "Bug Fix" reasons then the signing keys aren't part of
> > the source.
>
> Are you an idiot, or do you just choose to ignore all proof that doesn't
> fit your preconceived beliefs?

Nope. Merely stating a distinction. Either a device is distributed, like the 
common PC, that is designed for the user to change and update the software 
on, or, like the PS2 it isn't designed for that. If I find a way to update my 
PS2 to run Linux and find that it doesn't want to start the "Linux Firmware" 
because I'm lacking a signing key...

In the case of a device that internally runs Linux (or any other GPL'd 
software) and wasn't designed for the end-user to change the software running 
on it then the signing keys aren't part of the source. OTOH, if I sell a PC 
running Linux that requires the kernel be signed then the signing keys *are* 
part of the source, since a PC is designed for the end-user to change the 
software running on it.

BTW, nice use of irony with that line. Makes me regret letting my fingers get 
ahead of my brain.

> The GPL doesn't give someone distributing the software the choice of how
> much to limit the freedom of the user.

Never claimed it did. I just wasn't as specific as I should have been when 
giving my examples.

> Either private keys required to run the kernel on the hardware are
> always considered part of "the complete source code" or they are never
> part of it.

No. It all depends on the use-case. If the hardware is designed for the user 
to install their own, custom versions of the code on then the signing keys 
are part of the source as defined by the GPLv2.

If, OTOH, the hardware was never meant for the end-user to install custom 
versions of the software on, then while the signing keys are still 
*technically* part of the source, in practice they are not. Why? Because in 
most of those cases the end-user isn't granted the right to install and run 
custom binaries on the hardware. If the manufacturer provided the signing 
keys they'd be facilitating the commission of a crime. (call it "Breach of 
Contract")

> > I haven't looked into what Harald Welte did, but I'd be surprised if
> > someone tried following suit in America and had as much success.
> >...
>
> Harald is in Germany, and he therefore takes legal action against people
> distributing products violating his copyright on the Linux kernel
> in Germany at German courts based on German laws.

I know this. As I said, I doubt that anyone who tried this in America would 
have the success he has had.

DRH

> cu
> Adrian



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> I haven't looked into what Harald Welte did, but I'd be surprised if someone 
> tried following suit in America and had as much success.

One of the distinct advantages of German law over the US system (and to a
large extent the UK system its based upon) is that German law favours
justice over money.

> > AFAIK there haven't been any court rulings on this issue, and it could
> > even be that courts in different countries will decide differently.
> 
> Agreed.

That in theory shouldn't happen as the conventions on copyright are
supposed to stop that mess occuring.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> What if TiVo had put the kernel in a burned-in ROM (not flash, or on a
> flash ROM with no provision for reprogramming it)?  Would that also
> violate the "spirit" of the GPL?  Must any device that wishes to include
> GPL code include additional hardware to support replacing that code
> (even if that hardware is otherwise superfluous)?

This is an area the GPLv3 tries to clarify and for good reason. 

Of course these days in the US someone would probably sue arguing that a
ROM is protection scheme ;)

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> What if TiVo had put the kernel in a burned-in ROM (not flash, or on a
> flash ROM with no provision for reprogramming it)?  Would that also
> violate the "spirit" of the GPL?  Must any device that wishes to include
> GPL code include additional hardware to support replacing that code
> (even if that hardware is otherwise superfluous)?

As a PS to the GPL3 comment here is the basic difference

ROM	-	I can't modify the code on the device
		The creator can't modify the code further on the device

Tivo	-	I can't modify the code on the device
		The owner can modify the code 

One is an implicit limitation of the hardware (just like I can't run
openoffice on a 4MB PC even though the license gives me the right to
try), the other is an artificial restriction.

One case is witholding freedom in the GPL sense by one party while
keeping it themselves, the other is a limitation of the system
inevitably imposed on everyone.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Wednesday 13 June 2007 20:55:52 Alexandre Oliva wrote:
>> On Jun 13, 2007, Bongani Hlope <bhlope@mweb.co.za> wrote:
>> > On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:
>> >> if you distribute copies of such a program, [...]
>> >> you must give the recipients all the rights that you have

>> >> So, TiVo includes a copy of Linux in its DVR.

>> Can they modify the software in their device?

>> Do they pass this right on?

> But this *ISN'T* a right that the GPLv2 *REQUIRES* be passed on. 

You may be right.  The spirit says it should, but the legalese may
have missed the mark.  So GPLv3 makes it clear that it is.

>> It's about that copy of the kernel that ships in the device in object
>> code.  That's the one that TiVo customers ought to be entitled to
>> modify, if TiVo can modify it itself.

> The GPLv2 makes no real provision for *DIRECTLY* modifying object
> code.

Sure.  And that's not what I'm talking about.

What I'm talking about is being able to replace, upgrade, fix, tweak,
hack, and otherwise modify the program on the machine in the same way
that the vendor still can.

> What provisions the GPLv2 has apply to the source code.

This is too narrow a view of the GPLv2 provisions.

> And no, the end user *SHOULD* *NOT* be entitled to run whatever kernel they 
> like on a TiVO. It was designed with the "install new kernel" functionality 
> so that the TiVO corporation could update the kernel running on the hardware 
> when security problems came up, when bugs were fixed or even when the new 
> version gives better performance.

I.e., it was designed such that TiVo could modify the installed
kernel, but the user couldn't.  That's an outright violation of the
spirit of the GPL.

>> > Where does it say you should be able to run you modifications on the
>> > same hardware?

>> Where it says that you should pass on all the rights that you have.

>> While TiVo retains the ability to replace, upgrade, fix, break or make
>> any other change in the GPLed software in the device, it ought to pass
>> it on to its customers.

> It *DOES* *NOT* say "All rights that you have". It says "All rights
> that are granted you by this license".

I suggest you to reboot into memtest ;-)  The preamble of GPLv2 says:

  For example, if you distribute copies of such a program, whether
  gratis or for a fee, you must give the recipients
  all the rights that you have.
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
> If every piece of software released under the GPL had *ALL* rights
> passed on, then *ANYONE* could do the "I'm granting company X the
> right to use this software outside the GPL for $50,000USD."

The requirement above applies to licensees, not to the licensor.  The
licensor doesn't have to pass on all the rights s/he has, s/he only
decides to respect the licensee's freedoms, conditioned to the respect
of others' freedoms by means of passing on all rights the licensee has
over the software.

Arguably, one could use this argument to state that any authors of
derived works ought to pass on the right to choose the license for the
derived work under the GPL, but since (a) the above is not in the
legal terms, and (b) the downstream recipients would be bound by the
terms of the GPL anyway, and that requires the use of the GPL itself,
this would make no difference whatsoever.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> Still doesn't explain why you have argued that the GPLv3 doesn't
> attempt to cover hardware and then provide proof that it does.

It doesn't cover hardware, in the same way that it doesn't cover
patents, and it doesn't cover pro-DRM laws.  It merely arranges, as
best as we've managed a copyright license to do, that they can't be
used as excuses (or tools) to disrespect the freedoms that the GPL
demands all licensees to respect for other users.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Wed, Jun 13, 2007 at 09:40:13PM -0400, Daniel Hazelton wrote:
> On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:
> > On Wed, Jun 13, 2007 at 09:01:28PM -0400, Daniel Hazelton wrote:
> > > On Wednesday 13 June 2007 20:44:19 Adrian Bunk wrote:
> > > > On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> > > > > On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > > > > > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> 
> wrote:
> > > > > > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > > > > > >> > find offensive, so I don't choose to use it. It's offensive
> > > > > > >> > because Tivo never did anything wrong, and the FSF even
> > > > > > >> > acknowledged that. The fact
> > > > > > >>
> > > > > > >> Not all of us agree with this for the benefit of future legal
> > > > > > >> interpretation.
> > > > > > >
> > > > > > > Well, even the FSF lawyers did,
> > > > > >
> > > > > > Or rather they didn't think an attempt to enforce that in the US
> > > > > > would prevail (or so I'm told).  That's not saying what TiVo did
> > > > > > was right, and that's not saying that what TiVo did was permitted
> > > > > > by the license. Only courts of law can do that.
> > > > >
> > > > > Wrong! Anyone with half a brain can make the distinction. What TiVO
> > > > > did is entirely legal - they fully complied with the GPLv2. Note that
> > > > > what they *DON'T* allow people to do is run whatever version of
> > > > > whatever software they want on their hardware. They have that right -
> > > > > its the "Free Software Foundation" and the GPL - regardless of
> > > > > version - is a *SOFTWARE* license. ...
> > > >
> > > > The GPLv2 says:
> > > >
> > > > "For an executable work, complete source code means all the source code
> > > > for all modules it contains, plus any associated interface definition
> > > > files, plus the scripts used to control compilation and installation of
> > > > the executable."
> > > >
> > > > The question is whether this includes private keys.
> > > > Different people have different opinions regarding this issue.
> > > >
> > > > If "the complete source code" includes private keys, the GPLv2 requires
> > > > them to give any costumer the private keys.
> > > >
> > > > Fact is that Harald Welte did in several cases successfully convince
> > > > vendors that private keys are part of the source code if they are
> > > > required for running the compiled binary on some hardware.
> > >
> > > If the hardware was designed for the end-user to change the software
> > > running on it - including running software that it was never meant to run
> > > (ie: a complete webserver on cell phone) - then yes, the signing keys are
> > > a part of the source, as the software running on the device is designed
> > > to be updated by the user using the provided system.
> > >
> > > If, on the other hand, the only "software updates" the user is expected
> > > to perform are the installation of newer versions of the existing code
> > > for "Security" or "Bug Fix" reasons then the signing keys aren't part of
> > > the source.
> >
> > Are you an idiot, or do you just choose to ignore all proof that doesn't
> > fit your preconceived beliefs?
> 
> Nope. Merely stating a distinction. Either a device is distributed, like the 
> common PC, that is designed for the user to change and update the software 
> on, or, like the PS2 it isn't designed for that. If I find a way to update my 
> PS2 to run Linux and find that it doesn't want to start the "Linux Firmware" 
> because I'm lacking a signing key...
> 
> In the case of a device that internally runs Linux (or any other GPL'd 
> software) and wasn't designed for the end-user to change the software running 
> on it then the signing keys aren't part of the source. OTOH, if I sell a PC 
> running Linux that requires the kernel be signed then the signing keys *are* 
> part of the source, since a PC is designed for the end-user to change the 
> software running on it.
> 
> BTW, nice use of irony with that line. Makes me regret letting my fingers get 
> ahead of my brain.
> 
> > The GPL doesn't give someone distributing the software the choice of how
> > much to limit the freedom of the user.
> 
> Never claimed it did. I just wasn't as specific as I should have been when 
> giving my examples.
> 
> > Either private keys required to run the kernel on the hardware are
> > always considered part of "the complete source code" or they are never
> > part of it.
> 
> No. It all depends on the use-case. If the hardware is designed for the user 
> to install their own, custom versions of the code on then the signing keys 
> are part of the source as defined by the GPLv2.
> 
> If, OTOH, the hardware was never meant for the end-user to install custom 
> versions of the software on, then while the signing keys are still 
> *technically* part of the source, in practice they are not. Why? Because in 
> most of those cases the end-user isn't granted the right to install and run 
> custom binaries on the hardware. If the manufacturer provided the signing 
> keys they'd be facilitating the commission of a crime. (call it "Breach of 
> Contract")
>...

Repetition doesn't let wrong things become true.

Where does the GPLv2 talk about the distinction you are trying to make 
based on distributor intentions?     

We are talking about the GPLv2 licence text, not about what you would
personally prefer.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Satyam Sharma]

[ Gmail did horrible things to the original post by giving it a base64
content transfer encoding, so majordomo@vger dropped it. It's just
an off-topic digression, but I cared enough to resend anyway, fwiw. ]

On 6/14/07, Jörn Engel <joern@logfs.org> wrote:
> On Wed, 13 June 2007 14:33:07 -0700, Linus Torvalds wrote:
> >
> > The beauty of the GPLv2 is exactly that it's a "tit-for-tat" license, and
> > you can use it without having to drink the kool-aid.
>
> One could even add that "tit-for-tat" appears to be the best strategy
> in game theory for continuous runs of the prisoners dilemma.

Tit-for-tat is the best *deterministic* strategy when playing
iterated prisoner's dilemma. But note that "deterministic" and
"rational" are not adjectives that go well with "humans", and
most real-world (social) situations are noisy environments --
miscommunication and misunderstandings are the usual noise.
A double-D noise perceived by any player would throw a
tit-for-tat-playing couple into a perennial spiral of D's, for example,
which is clearly not a Pareto-efficient solution for either.

> At times I
> wonder why game theory isn't taught in schools yet - it might shorten
> discussions like these.

Yes, and no.

Yes - for teaching game theory (and its social relevance) in schools;
and add behavioral economics to this list :-)

No - it doesn't shorten discussions, however. And it shouldn't either.
Human / social situations are complex, Jörn; tit-for-tat can win
computer contests, for example, but it's not a behaviour one person
would find as entirely agreeable in another.

Satyam

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Thu, Jun 14, 2007 at 02:45:40AM +0100, Alan Cox wrote:
>...
> > > AFAIK there haven't been any court rulings on this issue, and it could
> > > even be that courts in different countries will decide differently.
> > 
> > Agreed.
> 
> That in theory shouldn't happen as the conventions on copyright are
> supposed to stop that mess occuring.

Is there any way how this would be resolved?

I can easily imagine that two courts, no matter whether they are in the 
same or different countries, would decide differently in grey areas like 
non-GPL modules or the GPLv2 and private keys.

If the two courts are in the same country there's usually a higher court 
above both that can resolve this. But what if let's say the highest 
court in the USA and the highest court in Germany would disagree on such 
a matter?

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Forrest]

On Thu, Jun 14, 2007 at 02:52:48AM +0100, Alan Cox wrote:
> 
> As a PS to the GPL3 comment here is the basic difference
> 
> ROM	-	I can't modify the code on the device
> 		The creator can't modify the code further on the device
> 
> Tivo	-	I can't modify the code on the device
> 		The owner can modify the code
> 
> One is an implicit limitation of the hardware (just like I can't run
> openoffice on a 4MB PC even though the license gives me the right to
> try), the other is an artificial restriction.
> 
> One case is witholding freedom in the GPL sense by one party while
> keeping it themselves, the other is a limitation of the system
> inevitably imposed on everyone.

I've been following this discussion and I find this interesting.
Consider these two cases:

1.) I ship the device back to the manufacturer, they replace the ROM,
    and ship it back to me.

2.) I ship the device back to the manufacturer, they load new code
    into it, and ship it back to me.

How do these two differ?  Or is it now just a question of the ROM
being in a socket?  I can't see how the technicalities of how the
hardware is constructed can change the legality of the software.

-- 
Dan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alexandre Oliva wrote:
> 
> Let me see if I got this right.  There was a section entitled
> "3. Digital Restrictions Management" in GPLv3dd1.  Are you saying
> that, when people complained about the DRM clause, they actually meant
> the provisions in "1. Source Code"

Yes. I said that multiple times. It was obvious. But people didn't listen.

It's now in Section 7, or whatever.

The section 3 never mattered.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:

> Exactly. They don't. What TiVO prevents is using that modified version on 
> their hardware. And they have that right, because the Hardware *ISN'T* 
> covered by the GPL.

Indeed, TiVO has this legal right.  But then they must not use
software under the GPLv3 in it.  And, arguably, they must not use
software under the GPLv2 either.

> In the case of 99% of the hardware targeted by the clause of the GPLv3 you 
> elucidate on, the "ability to install modified versions of the software" was 
> *NOT* intended for that use, nor was it intended for *ANYONE* *EXCEPT* 
> trained service personell to have *ACCESS* to that functionality. Arguing 
> otherwise is just idiotic - I have never found a piece of "high tech" 
> hardware (like a TiVO) that was designed for the end-user to modify. (yes, 
> installing a new version of the linux kernel is "modifying" the system)

It's about time for a change for better, wouldn't you think?

In 95% of the desktop computers, you can't make changes to the OS that
runs on it.  Whom is this good for?

> And? They distribute the kernel source - as they recieved it - in
> compliance with the GPL.

This makes it seem like you think that passing on the source code is
enough to comply with the GPL.  Check your assumptions.  It's not.

>> to prohibit people from removing locks that stop them from doing
>> things they're legally entitled to do

> What "Legally Entitled" things?

Time shifting of any shows, creating copies of shows for personal use,
letting others do so.  Think fair use, and how TiVO software and DRM
in general gets in the way.

> And... You do realize that almost every difference between the GPLv2
> and the GPLv3 is going to cause a hell of a lot of problems?

For those who are not willing to abide by the spirit of the license,
yes.  Does it look like I'm concerned about them?  If they're willing
to look for and maybe even find holes in the license to disrespect
users' freedoms, why should I worry about the problems that plugging
these holes is going to cause them?  If they'd taken the spirit of the
GPL for what it is, instead of looking for loopholes, this improved
wording wouldn't be causing them any problems whatsoever.

> The fact that the GPLv3 is designed to prevent things that RMS
> *PERSONALLY* finds distasteful - DRM and the like - is a big
> turn-off for a *LOT* of people.

This is a pretty sad accusation.  2/3s of the Free Software packages
use the GPL with its existing spirit, and you still haven't shown that
any changes proposed in GPLv3 fail to abide by the same spirit.  That
some (many?) people misunderstood or disregarded the spirit is an
unfortunate fact, but trying to pose the patching that's going into
GPLv3 as if it was a matter of personal taste, rather than improved
compliance with the spirit, is unfair and uncalled for.

> (Personally I don't like *ANY* version of the GPL, because there are
> chunks I have problems with)

What are you doing lurking and spreading confusion in a list about a
project that chose to use it, then?

>> Do you expect Linux would have flourished if computers had locks that
>> stopped people from modifying Linux in them?

> But you aren't talking about a "computer" here. You're talking about
> a mass-market device that must comply with both US and International
> copyright law - and that's just a TiVO.

Oh, sorry.  I missed when the meaning of the word computer was
narrowed from "machine with a general-purpose microprocessor, memory
and other peripherals" to whatever you decide it is.

And then, the GPL doesn't talk about computers at all.  It's not about
the hardware, it's about the software, remember? ;-)

> if you upload a modified linux kernel to your wireless router that
> gives it a 2000 foot range, you've just broken the law

At which point, you get punished by the law system.

> *AND* violated the license on the hardware which states that you
> "won't modify it or the controlling software"

Err..  The hardware licensor who includes software under the GPL be
supposed to be a licensee of the software in order to have legal
permission to distribute it, at which point the following provision
kicks in:

  6. Each time you redistribute the Program (or any work based on the
  Program) [...] You may not impose any further restrictions on the
  recipients' exercise of the rights granted herein. [...]

And here's one of the rights granted herein that would be restricted
by this hardware license:

  2. You may modify your copy or copies of the Program or any portion
  of it, thus forming a work based on the Program,

So such a restriction in the hardware license seems to be failure to
comply with the GPL, which means the violator may lose the license.

> even things like the connectors used to upload the operating
> software at the factory that people now cannot have in a device that
> runs GPL(v3) covered software unless they ship the related
> "Installation Information".

This sounds like a reasonable point.  Please bring it up at
gplv3.fsf.org.  If it requires specialized hardware to modify the
software in the device, the hardware manufacturer can't modify the
software without cooperation from the user, and then perhaps it would
be fair for the user to need cooperation from the manufacturer.

> That, to me, reads like RMS got mad about TiVO and said "I don't
> like it, lets add a clause making it wrong to the next GPL". Hell,
> that *IS* what happened, and nothing the FSF or Eben Moglen says
> will convince me otherwise.

If you've already made your mind about this, in spite of not having
the facts, I guess it doesn't make sense for me to waste my time
trying to convince you, does it?

>> What's under the license is the software in it.  And that license
>> spirit requires the distributor to pass on the right to modify the
>> software.

> And since when did they have to enable people to use their hardware in 
> violation of the licensing agreement they implicitly agree to when opening 
> the package?

Since they got permission to distribute the software under the
condition of passing on the freedoms without imposing further
restrictions on their exercise.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> (and, in the case of a TiVO, the signing 
> keys are part of the installation, not the running or building.

Is installation not a precondition for running?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 22:08:27 Adrian Bunk wrote:
> On Wed, Jun 13, 2007 at 09:40:13PM -0400, Daniel Hazelton wrote:
> > On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:
> > > On Wed, Jun 13, 2007 at 09:01:28PM -0400, Daniel Hazelton wrote:
> > > > On Wednesday 13 June 2007 20:44:19 Adrian Bunk wrote:
> > > > > On Wed, Jun 13, 2007 at 07:46:15PM -0400, Daniel Hazelton wrote:
> > > > > > On Wednesday 13 June 2007 19:15:42 Alexandre Oliva wrote:
> > > > > > > On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org>
> >
> > wrote:
> > > > > > > > On Wed, 13 Jun 2007, Alan Cox wrote:
> > > > > > > >> > find offensive, so I don't choose to use it. It's
> > > > > > > >> > offensive because Tivo never did anything wrong, and the
> > > > > > > >> > FSF even acknowledged that. The fact
> > > > > > > >>
> > > > > > > >> Not all of us agree with this for the benefit of future
> > > > > > > >> legal interpretation.
> > > > > > > >
> > > > > > > > Well, even the FSF lawyers did,
> > > > > > >
> > > > > > > Or rather they didn't think an attempt to enforce that in the
> > > > > > > US would prevail (or so I'm told).  That's not saying what TiVo
> > > > > > > did was right, and that's not saying that what TiVo did was
> > > > > > > permitted by the license. Only courts of law can do that.
> > > > > >
> > > > > > Wrong! Anyone with half a brain can make the distinction. What
> > > > > > TiVO did is entirely legal - they fully complied with the GPLv2.
> > > > > > Note that what they *DON'T* allow people to do is run whatever
> > > > > > version of whatever software they want on their hardware. They
> > > > > > have that right - its the "Free Software Foundation" and the GPL
> > > > > > - regardless of version - is a *SOFTWARE* license. ...
> > > > >
> > > > > The GPLv2 says:
> > > > >
> > > > > "For an executable work, complete source code means all the source
> > > > > code for all modules it contains, plus any associated interface
> > > > > definition files, plus the scripts used to control compilation and
> > > > > installation of the executable."
> > > > >
> > > > > The question is whether this includes private keys.
> > > > > Different people have different opinions regarding this issue.
> > > > >
> > > > > If "the complete source code" includes private keys, the GPLv2
> > > > > requires them to give any costumer the private keys.
> > > > >
> > > > > Fact is that Harald Welte did in several cases successfully
> > > > > convince vendors that private keys are part of the source code if
> > > > > they are required for running the compiled binary on some hardware.
> > > >
> > > > If the hardware was designed for the end-user to change the software
> > > > running on it - including running software that it was never meant to
> > > > run (ie: a complete webserver on cell phone) - then yes, the signing
> > > > keys are a part of the source, as the software running on the device
> > > > is designed to be updated by the user using the provided system.
> > > >
> > > > If, on the other hand, the only "software updates" the user is
> > > > expected to perform are the installation of newer versions of the
> > > > existing code for "Security" or "Bug Fix" reasons then the signing
> > > > keys aren't part of the source.
> > >
> > > Are you an idiot, or do you just choose to ignore all proof that
> > > doesn't fit your preconceived beliefs?
> >
> > Nope. Merely stating a distinction. Either a device is distributed, like
> > the common PC, that is designed for the user to change and update the
> > software on, or, like the PS2 it isn't designed for that. If I find a way
> > to update my PS2 to run Linux and find that it doesn't want to start the
> > "Linux Firmware" because I'm lacking a signing key...
> >
> > In the case of a device that internally runs Linux (or any other GPL'd
> > software) and wasn't designed for the end-user to change the software
> > running on it then the signing keys aren't part of the source. OTOH, if I
> > sell a PC running Linux that requires the kernel be signed then the
> > signing keys *are* part of the source, since a PC is designed for the
> > end-user to change the software running on it.
> >
> > BTW, nice use of irony with that line. Makes me regret letting my fingers
> > get ahead of my brain.
> >
> > > The GPL doesn't give someone distributing the software the choice of
> > > how much to limit the freedom of the user.
> >
> > Never claimed it did. I just wasn't as specific as I should have been
> > when giving my examples.
> >
> > > Either private keys required to run the kernel on the hardware are
> > > always considered part of "the complete source code" or they are never
> > > part of it.
> >
> > No. It all depends on the use-case. If the hardware is designed for the
> > user to install their own, custom versions of the code on then the
> > signing keys are part of the source as defined by the GPLv2.
> >
> > If, OTOH, the hardware was never meant for the end-user to install custom
> > versions of the software on, then while the signing keys are still
> > *technically* part of the source, in practice they are not. Why? Because
> > in most of those cases the end-user isn't granted the right to install
> > and run custom binaries on the hardware. If the manufacturer provided the
> > signing keys they'd be facilitating the commission of a crime. (call it
> > "Breach of Contract")
> >...
>
> Repetition doesn't let wrong things become true.
>
> Where does the GPLv2 talk about the distinction you are trying to make
> based on distributor intentions?
>
> We are talking about the GPLv2 licence text, not about what you would
> personally prefer.

The GPLv2 doesn't have to cover this distinction to make it a reality. This 
distinction is *EXACTLY* the type of distinction a lawyer will make when 
arguing the point.

Yes, it's artificial. Yes, it does appear to violate the GPLv2 - *IF* you read 
the text of such in a specific manner. 

However, the GPL, until version 3, *NEVER* guaranteed the right to run a given 
piece of software on *ANY* hardware - not the hardware it *COMES* on. 

And please, I repeated myself only because your reply seemed to imply that you 
didn't understand the statement I had made. Since you have now informed me, 
in a backhanded way, informed me that my interpretation of your response was 
wrong, I will not repeat myself again.

Also note that I have re-examined the facts, in light of new information 
presented in this discussion, and have come to the conclusion that devices 
like the TiVO, in keeping the signing keys private (because of 
the "Facilitation of a Crime" thing I noted earlier), is violating the GPL, 
but not in the manner almost everyone is arguing. The violation is, rather, 
with the clause about the license being null and void in event of laws 
impacting the delivery of the source. (Because, as I also stated earlier, the 
signing keys are part of the source. Since, in some cases, the license on the 
hardware prevents running modified binaries (a reason for the digital 
signing) companies will keep said keys private - doing otherwise can (and I 
can assure you that some lawyer will do this) be construed as "Facilitating 
the Commission of a crime". In this case, it'd be "Breach of Contract" - 
IANAL, but IIRC, licenses fall under contract law))

DRH

>
> cu
> Adrian



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Chris Adams <cmadams@hiwaay.net> wrote:

> Once upon a time, Alexandre Oliva  <aoliva@redhat.com> said:
>> if you distribute copies of such a program, [...]
>> you must give the recipients all the rights that you have

>> So, TiVo includes a copy of Linux in its DVR.  

>> TiVo retains the right to modify that copy of Linux as it sees fit.

>> It doesn't give the recipients the same right.

> Sure it does; you received a program (the kernel) and you can modify it.

If I take the software I received, build it and install it on the same
hardware, it won't run.  Something is missing in the source code I
received, I guess..

If I make changes to the source code, build it, and install it on that
same computer, it won't run.  How is that being able to modify *that*
copy of the program?

If TiVo makes the same changes, builds tehm, and installs it on my
computer, it will run just fine.  How are they passing on the right
they had to me?

> You also received hardware; they don't support modification of that.

They don't have to support them.  They don't have to help me if it
breaks.  But if they can do it and I can't, they're failing to comply
with the spirit of the GPL.

> Nowhere in the license does it say they have to, because the license
> only covers the program.

They can't distribute the program while imposing restrictions on
modification not present in the software license itself.

> Or are you claiming that putting software on hardware makes the result a
> derivative work?  I think it falls under the "mere aggregation" clause.

I tend to agree, in this particular case, but IANAL.  I don't rule out
derivative works in future attempts to find loopholes in the GPL.

> What if TiVo had put the kernel in a burned-in ROM

Then they wouldn't have the ability to change it any more, so there
wouldn't be such a right to pass on to the users.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> Either a device is distributed, like the common PC, that is designed
> for the user to change and update the software on, or, like the PS2
> it isn't designed for that.

Have you ever installed GNU/Linux on a PC "Designed for Microsoft Windows"?

How dare you? ;-)

> If, OTOH, the hardware was never meant for the end-user to install custom 
> versions of the software on, then while the signing keys are still 
> *technically* part of the source, in practice they are not. Why? Because in 
> most of those cases the end-user isn't granted the right to install and run 
> custom binaries on the hardware.

And distributing the GPLed software under this restriction is quite
likely copyright infringement.

> I know this. As I said, I doubt that anyone who tried this in
> America would have the success he has had.

On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:

>> Are you an idiot, or do you just choose to ignore all proof that
>> doesn't fit your preconceived beliefs?

;-) :-P :-D

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Wed, 13 Jun 2007, Alexandre Oliva wrote:
> 
> So, TiVo includes a copy of Linux in its DVR.  

Stop right there.

You seem to make the mistake to think that software is something physical.

> TiVo retains the right to modify that copy of Linux as it sees fit.

No. If you were logical (which you are not), you would admit that 
 (a) physical property is very different from intellectual property (the 
     FSF seems to admit that when it suits their needs, not otherwise)
 (b) They never modified "a copy" of Linux - they simply replaced it with
     "another copy" of Linux. The only thing that actually got *modified* 
     was their hardware!

The first copy didn't "morph" into a second copy. There was no "physical" 
software that was molded.  They do need to follow the GPLv2, since clearly 
they _do_ distribute Linux, but you have all the same rights as they do 
with regard to the *software*. 

The fact that they maintained some control of the *hardware* (and some 
software they wrote too) they designed is _their_ choice.

What Tivo did and do, is to distribute hardware that can *contain* a copy 
of Linux (or just about anything else, for that matter - again, there's 
a difference between physical and intellectual property).

And their hardware (and firmware) will run some integrity checks on 
*whatever* copies of software they have.  This is all totally outside 
Linux itself.

Btw, according to your _insane_ notion of "a copy" of software, you can 
never distribute GPL'd software on a CD-ROM, since you've taken away the 
right of people to modify that CD-ROM by burning and fixating it. So 
according to your (obvously incorrect) reading of the GPLv2, every time 
Red Hat sends anybody a CD-ROM, they have restricted peoples right to 
modify the software on that CD-ROM bymaking it write-only.

See? Your reading of the license doesn't _work_. Mine does. What I say is 
that when you distribute software, you don't distribute "a copy" of 
software, you distribute the _information_ about the software, so that 
others can take it and modify it. And notice? My reading of the license 
must be the correct one, since my reading actually makes sense, unlike 
yours.

And yes, when Tivo distributes Linux, they give everybody else all the 
same rights they have - with respect to Linux! No, not with respect to 
their hardware, but that's a totally different thing, and if you cannot 
wrap your mind around the difference between "the software that is on a 
CD" and the "piece of plastic that is the CD", and see that when you 
replace "CD" with any other medium, the equation doesn't change, I don't 
know what to say.

> It doesn't give the recipients the same right.
> 
> Oops.
> 
> Sounds like a violation of the spirit to me.

Only if you extend the license to the *hardware*. Oops. Which it never did 
before. 

In other words, you basically try to change the rules. The GPLv2 clearly 
states that it's about software, not hardware. All the language you quoted 
talks about software.

In other words, the only way to argue that I'm wrong is to try to twist 
the meanings of the words, and say that words only mean one specific thing 
that _you_ claim are their meaning.

And I'm saying you act like Humpty Dumpty when you do. You can argue that 
way all you like, but your argument is nonsensical. It's akin to the 
argument that "God is perfect. Perfect implies existence. Therefore God 
exists".

That kind of argument only works if you *define* the words to suit your 
argument. But it's a logical fallacy.

And I'm saying that the GPLv2 can mroe straightforwardly be read the way I 
read it - to talk about software, and to realize that software is not "a 
copy", it's a more abstract thing. You get Linux when you buy a Tivo (or 
preferably - don't buy it, since you don't like it), and that means that 
they have to give you access to and control over the SOFTWARE. But nowhere 
in the GPL (in the preamble or anywhere else) does it talk about giving 
you control over the HARDWARE, and the only way you can twist the GPLv2 to 
say that is by trying to re-define what the words mean.

And then you call *me* confused? After you yourself admitted that the FSF 
actually agrees with me, and that what Tivo did was not a license 
violation?

Trust me, I'm not the confused person here.

I'm perfectly fine with other people wanting to extend the license to 
cover the hardware, but I am *not* perfectly fine with people then trying 
to claim I'm confused just because I don't agree with them.

Face it: the GPLv3 is a _new_ license. Making funamentally _different_ and 
_new_ restrictions that do not exist in the GPLv2, and do not exist in the 
preamble. Any language attempts to make it appear otherwise are just 
sophistry.

And btw, just to make it clear: as far as I'm concerned, you can read the 
preamble and the word "freedom" and "rigths" _your_ way. I'm not objecting 
to that at all. If you read it so that you think it's wrong to distribute 
GPL'd software on a CD-ROM, that's really not my problem. You do whatever 
you want to, and think the license means whatever you want to.

What I'm objecting to is how you claim that anybody that doesn't follow 
your interpretation is "confused". When clearly even the FSF lawyers agree 
that my interpretation was _correct_, and I don't think your 
interpretation even makes sense!

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Wed, Jun 13, 2007 at 10:43:14PM -0400, Daniel Hazelton wrote:
> On Wednesday 13 June 2007 22:08:27 Adrian Bunk wrote:
> > On Wed, Jun 13, 2007 at 09:40:13PM -0400, Daniel Hazelton wrote:
> > > On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:
>...
> > > > Either private keys required to run the kernel on the hardware are
> > > > always considered part of "the complete source code" or they are never
> > > > part of it.
> > >
> > > No. It all depends on the use-case. If the hardware is designed for the
> > > user to install their own, custom versions of the code on then the
> > > signing keys are part of the source as defined by the GPLv2.
> > >
> > > If, OTOH, the hardware was never meant for the end-user to install custom
> > > versions of the software on, then while the signing keys are still
> > > *technically* part of the source, in practice they are not. Why? Because
> > > in most of those cases the end-user isn't granted the right to install
> > > and run custom binaries on the hardware. If the manufacturer provided the
> > > signing keys they'd be facilitating the commission of a crime. (call it
> > > "Breach of Contract")
> > >...
> >
> > Repetition doesn't let wrong things become true.
> >
> > Where does the GPLv2 talk about the distinction you are trying to make
> > based on distributor intentions?
> >
> > We are talking about the GPLv2 licence text, not about what you would
> > personally prefer.
> 
> The GPLv2 doesn't have to cover this distinction to make it a reality. This 
> distinction is *EXACTLY* the type of distinction a lawyer will make when 
> arguing the point.
>...

Reality check:

Harald convinced companies that they have to provide the private keys 
required to run the Linux kernel they ship on their hardware.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[david]

since the latest draft of the GPLv3 now discriminates against some uses 
(industrial vs commercial I think are the terms used) does it even qualify 
as a Open Source lincense anymore by the OSI terms?

David Lang

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 22:04:04 Alexandre Oliva wrote:
> On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > Still doesn't explain why you have argued that the GPLv3 doesn't
> > attempt to cover hardware and then provide proof that it does.
>
> It doesn't cover hardware, in the same way that it doesn't cover
> patents, and it doesn't cover pro-DRM laws.  It merely arranges, as
> best as we've managed a copyright license to do, that they can't be
> used as excuses (or tools) to disrespect the freedoms that the GPL
> demands all licensees to respect for other users.

Consider this scenario:
Small company A is manufacturing a new WiFi router.
They decide to have it run HURD as the OS.
In complying with the GPLv3 they supply the signing keys and everything else 
needed to install a new kernel on the hardware.
User B buys the router and modifies the kernel so it drives the WiFi to an 
output power twice that which it is licensed to carry.
FCC finds out and prosecutes User B for violating the regulations.
FCC then pulls the small companies license until they change their hardware so 
the driver can't push it to transmit at a higher power level and levies a 
fine.
Small company A loses the money paid on the fine, has to recall all the 
devices that can be modified (through software) to break the law at a massive 
cost *AND* has to redesign their hardware. The total cost drives the company 
into bankruptcy.

Small companies C,D and E, in order to avoid the fate of small company A, 
purchases a license for proprietary OS "F" to drive their new hardware.

Net loss: A lot of the users and publicity that "Free Software" used to get, 
because GPLv3 contains language that opens the companies to lawsuits that 
they wouldn't otherwise face.

Which is better: Growing the base of installed GPL covered software, 
or "ethics and morals" that demand the language that has been added to the 
GPLv3 ? Personally I'd like to see proprietary software driven into a very 
small "niche" market or entirely out of existence. However much I want this 
to happen, I cannot be anything *BUT* scared of the GPLv3 simply because I 
see it creating massive problems - and all because of a *small* portion of 
the new language it contains. It has taken almost 15 years for "Free 
Software" to make a dent in the market, and, IMHO, a lot of that is both 
Linux and the "holes" in GPLv2.

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Adrian Bunk wrote:
> 
> "For an executable work, complete source code means all the source code 
> for all modules it contains, plus any associated interface definition 
> files, plus the scripts used to control compilation and installation of 
> the executable."
> 
> The question is whether this includes private keys.

No. That's the question as the FSF would like to frame it.

But the real fact is that it *not* the right question.

You can install Linux on a Tivo all you like. Take out the harddisk, 
install your own version of Linux on it, and put it back in. That's pretty 
much how Tivo installs Linux on the things too, afaik, although they don't 
need to take the disk out (since they just assemble it).

No magic needed. In fact, no keys needed.

Now, maybe the hardware/firmware knows to expect a certain SHA1 on that 
disk, that's a different issue. Tivo could even tell you exactly what the 
SHA1 they are checking is. Maybe they have a list of SHA1's, and maybe 
they have a way to upgrade THEIR OWN FIRMWARE with new SHA1's, and they 
could still tell you all of them, and be very open.

And you could actually replace their copy of Linux with another one. It 
would have to have the same SHA1 to actually start _running_, but that's 
the hardware's choice. 

See? No private keys needed. No magic install scripts. It really _is_ that 
easy.

Of course, using private keys, and signing the image with them is possibly 
a technically more flexible/easier/more obvious way to do it, but in the 
end, do you really want to argue technical details?

But I think the whole thing is totally misguided, because the fact is, the 
GPLv2 doesn't talk about "in place" or "on the same hardware". 

So take another example: I obviously distribute code that is copyrighted 
by others under the GPLv2. Do I follow the GPLv2? I sure as hell do! But 
do I give you the same rights as I have to modify the copy on 
master.kernel.org as I have? I sure as hell DO NOT!

So by the idiotic logic of "modifying in place", I'm violating the GPLv2 
every time I'm makign a release - because I make Linux available, but I 
don't actually give people the "same rights" to that particular copy that 
I have! Oh horrors of horrors! You need to make a _copy_ of the thing I 
distribute, and then you have the same rights I have to that _copy_, but 
you never had the same rights to the thing I actually distributed!

And here's a big clue for people: anybody who thinks that I'm violating 
the GPLv2 by not giving out my private SSH key to master.kernel.org is a 
f*cking moron! You have the right to modify *copies* of the kernel I 
distribute, but you cannot actually modify the _actual_ entity that I made 
available!

See any parallels here? Any parallel to a CD-ROM distribution, or a Tivo 
distribution? The rights that the GPLv2 gives to "the software", is to 
something much bigger than "the particular copy of the software". 

Can people really not see the difference between "the software" and "a 
particular encoded copy of the software"? 

I'm sorry, but people who cannot see that difference are just stupid.

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 22:38:05 Alexandre Oliva wrote:
> On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:
> >
> > Exactly. They don't. What TiVO prevents is using that modified version on
> > their hardware. And they have that right, because the Hardware *ISN'T*
> > covered by the GPL.
>
> Indeed, TiVO has this legal right.  But then they must not use
> software under the GPLv3 in it.  And, arguably, they must not use
> software under the GPLv2 either.

Yes. It can be argued. But I cannot find *ANYTHING* in the GPLv2 that stops 
anyone from doing that, unless you add extra meaning to one specific clause. 
(And I am willing to admit that *MOST* people do give it that extra meaning)

> > In the case of 99% of the hardware targeted by the clause of the GPLv3
> > you elucidate on, the "ability to install modified versions of the
> > software" was *NOT* intended for that use, nor was it intended for
> > *ANYONE* *EXCEPT* trained service personell to have *ACCESS* to that
> > functionality. Arguing otherwise is just idiotic - I have never found a
> > piece of "high tech" hardware (like a TiVO) that was designed for the
> > end-user to modify. (yes, installing a new version of the linux kernel is
> > "modifying" the system)
>
> It's about time for a change for better, wouldn't you think?

I've never had a reason to want to change the way any device like a TiVO 
works. So I can't comment on this.

> In 95% of the desktop computers, you can't make changes to the OS that
> runs on it.  Whom is this good for?

Faulty logic. I have yet to find a computer that I couldn't change the OS on. 
I have run Linux on 3 different Mac's, every x86 machine I've ever owned and 
even had it running on my Palm. Whats more is that I have *never* heard of a 
person that knows what they are doing not being able to change the OS on a 
desktop computer.

> > And? They distribute the kernel source - as they recieved it - in
> > compliance with the GPL.
>
> This makes it seem like you think that passing on the source code is
> enough to comply with the GPL.  Check your assumptions.  It's not.

Hrm. Strange, but thats what most companies think. Hell, it even says that you 
have to do just that in the GPL. If you're talking about the fact that it can 
be argued that they are "distributing" Linux by selling their boxes and its a 
modified version then I'll agree with you. 

> >> to prohibit people from removing locks that stop them from doing
> >> things they're legally entitled to do
> >
> > What "Legally Entitled" things?
>
> Time shifting of any shows, creating copies of shows for personal use,
> letting others do so.  Think fair use, and how TiVO software and DRM
> in general gets in the way.

I thought that time shifting and creating personal copies was what the TiVO 
did already. Or do you mean "transferring the recorded copies off the TiVO 
and on to a different medium"? If that is what you mean by "Creating Copies" 
then, IIRC, you're wrong. DRM, I do agree, gets in the way of "Fair Use".

> > And... You do realize that almost every difference between the GPLv2
> > and the GPLv3 is going to cause a hell of a lot of problems?
>
> For those who are not willing to abide by the spirit of the license,
> yes.  Does it look like I'm concerned about them?  If they're willing
> to look for and maybe even find holes in the license to disrespect
> users' freedoms, why should I worry about the problems that plugging
> these holes is going to cause them?  If they'd taken the spirit of the
> GPL for what it is, instead of looking for loopholes, this improved
> wording wouldn't be causing them any problems whatsoever.

Okay. So you're not concerned that you're potentially pushing companies that 
would otherwise be major consumers of GPL'd software away? That doesn't make 
sense to me.

> > The fact that the GPLv3 is designed to prevent things that RMS
> > *PERSONALLY* finds distasteful - DRM and the like - is a big
> > turn-off for a *LOT* of people.
>
> This is a pretty sad accusation.  2/3s of the Free Software packages
> use the GPL with its existing spirit, and you still haven't shown that
> any changes proposed in GPLv3 fail to abide by the same spirit.  That
> some (many?) people misunderstood or disregarded the spirit is an
> unfortunate fact, but trying to pose the patching that's going into
> GPLv3 as if it was a matter of personal taste, rather than improved
> compliance with the spirit, is unfair and uncalled for.

Why should I repeat Linus' explanation of the ways that GPLv3 violates the 
spirit of GPLv2?

And why shouldn't I pose it as a matter of "Personal Taste"? The biggest and 
most powerful voice in the FSF says "I don't like Tivoization" and "I don't 
like DRM" and when the GPLv3 appears it has language that makes those 
violations of the license. Just like people have started using "GNU/Linux" 
or "GNU+Linux" to refer to Linux - a big voice spoke and said "It should be 
GNU/Linux" and it happens. (Not that I really have anything against that - 
though I feel that "GNU/Linux" gives the wrong impression)

> > (Personally I don't like *ANY* version of the GPL, because there are
> > chunks I have problems with)
>
> What are you doing lurking and spreading confusion in a list about a
> project that chose to use it, then?

Just because I don't like the license doesn't make disqualify me from liking 
something that uses the license. And I doubt I've "confused" anyone - where I 
have someone has caught it and called me on it.

> >> Do you expect Linux would have flourished if computers had locks that
> >> stopped people from modifying Linux in them?
> >
> > But you aren't talking about a "computer" here. You're talking about
> > a mass-market device that must comply with both US and International
> > copyright law - and that's just a TiVO.
>
> Oh, sorry.  I missed when the meaning of the word computer was
> narrowed from "machine with a general-purpose microprocessor, memory
> and other peripherals" to whatever you decide it is.

The word "Computer", in the manner I used it there, means "General Purpose 
Computational Device". A "TiVO" is not, and has never been, a "General 
Purpose Computational Device". 

(Perhaps I should have made the above clear)

> And then, the GPL doesn't talk about computers at all.  It's not about
> the hardware, it's about the software, remember? ;-)

Exactly. And I don't see anything about a TiVO (or any device that, like a 
TiVO, requires binaries that run on it to be digitally signed) that stops you 
from exercising the "freedoms" guaranteed by the GPL. As I said before, what 
it does is stop you from violating the license on the hardware.

> > if you upload a modified linux kernel to your wireless router that
> > gives it a 2000 foot range, you've just broken the law
>
> At which point, you get punished by the law system.

But the GPLv2 gives companies a chance to protect themselves from legal 
actions by people that are sure to follow. With the GPLv3 they have none, 
because it becomes a case of "I was just doing what I have the right to do 
under the license for the software. If I wasn't supposed to do it, they 
shouldn't have made it possible."

> > *AND* violated the license on the hardware which states that you
> > "won't modify it or the controlling software"
>
> Err..  The hardware licensor who includes software under the GPL be
> supposed to be a licensee of the software in order to have legal
> permission to distribute it, at which point the following provision
> kicks in:
>
>   6. Each time you redistribute the Program (or any work based on the
>   Program) [...] You may not impose any further restrictions on the
>   recipients' exercise of the rights granted herein. [...]

Hrm... I'd forgotten that clause entirely. That must be why I can't find a 
WRT-54GL anywhere in the US anymore...

> And here's one of the rights granted herein that would be restricted
> by this hardware license:
>
>   2. You may modify your copy or copies of the Program or any portion
>   of it, thus forming a work based on the Program,
>
> So such a restriction in the hardware license seems to be failure to
> comply with the GPL, which means the violator may lose the license.

My apologies. As I said above I totally forgot about section 6. 

> > even things like the connectors used to upload the operating
> > software at the factory that people now cannot have in a device that
> > runs GPL(v3) covered software unless they ship the related
> > "Installation Information".
>
> This sounds like a reasonable point.  Please bring it up at
> gplv3.fsf.org.  If it requires specialized hardware to modify the
> software in the device, the hardware manufacturer can't modify the
> software without cooperation from the user, and then perhaps it would
> be fair for the user to need cooperation from the manufacturer.

Will do. I'm surprised that this situation never came up before, actually. 

> > That, to me, reads like RMS got mad about TiVO and said "I don't
> > like it, lets add a clause making it wrong to the next GPL". Hell,
> > that *IS* what happened, and nothing the FSF or Eben Moglen says
> > will convince me otherwise.
>
> If you've already made your mind about this, in spite of not having
> the facts, I guess it doesn't make sense for me to waste my time
> trying to convince you, does it?

You've made some well reasoned and supported arguments here. I respect RMS and 
Eben Moglen because of what they've done, but that doesn't mean I have to 
like them. However, you are free to (try) to change my mind about them.

> >> What's under the license is the software in it.  And that license
> >> spirit requires the distributor to pass on the right to modify the
> >> software.
> >
> > And since when did they have to enable people to use their hardware in
> > violation of the licensing agreement they implicitly agree to when
> > opening the package?
>
> Since they got permission to distribute the software under the
> condition of passing on the freedoms without imposing further
> restrictions on their exercise.

Okay. Given the interpretation of GPLv2 you seem to have I will grant you this 
point. Personally, I do not feel that the "spirit" of the GPLv2 has anything 
to do with running modified binaries on a "closed" hardware platform (like a 
TiVO). But, as I said, I'll grant you the point, because our views on 
the "spirit" of the GPL and interpretations of it seem to be completely 
incompatible.

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 533 bytes --]

On Wed, 13 Jun 2007 23:40:47 -0300, Alexandre Oliva said:
> On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> 
> > (and, in the case of a TiVO, the signing 
> > keys are part of the installation, not the running or building.
> 
> Is installation not a precondition for running?

If a company sells you hardware that includes a ROM that contains GPL'ed
software, are they in violation of the GPL if they don't include a ROM burner
in the hardware?  Or are ROM burners like compilers, where you have to supply
your own?


[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Wednesday 13 June 2007 22:56:40 Adrian Bunk wrote:
> On Wed, Jun 13, 2007 at 10:43:14PM -0400, Daniel Hazelton wrote:
> > On Wednesday 13 June 2007 22:08:27 Adrian Bunk wrote:
> > > On Wed, Jun 13, 2007 at 09:40:13PM -0400, Daniel Hazelton wrote:
> > > > On Wednesday 13 June 2007 21:24:01 Adrian Bunk wrote:
> >
> >...
> >
> > > > > Either private keys required to run the kernel on the hardware are
> > > > > always considered part of "the complete source code" or they are
> > > > > never part of it.
> > > >
> > > > No. It all depends on the use-case. If the hardware is designed for
> > > > the user to install their own, custom versions of the code on then
> > > > the signing keys are part of the source as defined by the GPLv2.
> > > >
> > > > If, OTOH, the hardware was never meant for the end-user to install
> > > > custom versions of the software on, then while the signing keys are
> > > > still *technically* part of the source, in practice they are not.
> > > > Why? Because in most of those cases the end-user isn't granted the
> > > > right to install and run custom binaries on the hardware. If the
> > > > manufacturer provided the signing keys they'd be facilitating the
> > > > commission of a crime. (call it "Breach of Contract")
> > > >...
> > >
> > > Repetition doesn't let wrong things become true.
> > >
> > > Where does the GPLv2 talk about the distinction you are trying to make
> > > based on distributor intentions?
> > >
> > > We are talking about the GPLv2 licence text, not about what you would
> > > personally prefer.
> >
> > The GPLv2 doesn't have to cover this distinction to make it a reality.
> > This distinction is *EXACTLY* the type of distinction a lawyer will make
> > when arguing the point.
> >...
>
> Reality check:
>
> Harald convinced companies that they have to provide the private keys
> required to run the Linux kernel they ship on their hardware.

In Germany, not America. I should have qualified my statement to make it clear 
I mean "In America". Sorry about the confusion.

DRH

>
> cu
> Adrian



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 1125 bytes --]

On Thu, 14 Jun 2007 04:56:40 +0200, Adrian Bunk said:

> Reality check:
> 
> Harald convinced companies that they have to provide the private keys 
> required to run the Linux kernel they ship on their hardware.

No, the *real* reality check:

The operative words here are "convinced companies" - as opposed to "convinced
a judge to rule that private keys are required to be disclosed". (I just
checked around on gpl-violations.org, and I don't see any news items that say
they actually generated citable case law on the topic of keys...)

Harald convinced companies that it was easier/cheaper/faster to provide the
private keys than to continue in a long legal battle with an uncertain outcome.
If the company estimates the total loss due to keys being released is US$100K,
but the costs of taking it to court are estimated at US$200K, it's obviously
a win (lesser loss, actually) for the company to just fold.

Incidentally, this same logic is what drives the average successful patent
troll lawsuit - the sued company will buy a license for $25K, just because
they know that fighting the lawsuit will cost $100K and up.



[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bron Gondwana]

On Thu, Jun 14, 2007 at 02:52:48AM +0100, Alan Cox wrote:
> > What if TiVo had put the kernel in a burned-in ROM (not flash, or on a
> > flash ROM with no provision for reprogramming it)?  Would that also
> > violate the "spirit" of the GPL?  Must any device that wishes to include
> > GPL code include additional hardware to support replacing that code
> > (even if that hardware is otherwise superfluous)?
> 
> As a PS to the GPL3 comment here is the basic difference
> 
> ROM	-	I can't modify the code on the device
> 		The creator can't modify the code further on the device
> 
> Tivo	-	I can't modify the code on the device
> 		The owner can modify the code 

Tivo gets sick of the endless flamewars on lkml, signs a copy
of QNX, pushes it out to the hardware.  No more Linux on Tivo.

You also can't replace that but Tivo can.  As I see it the two
are completely orthagonal:

a) Can anyone but the manufacturer upload new software into a
   a device without taking extreme measures (soldering a new
   public-key-containing-chip onto the board)

b) Is the software currently installed on a device licenced under
   a rule which requires the distributor to also distribute source
   code upon request.

Now I think it would reasonable to ask that the source code be able
to be built by [same compiler, same flags, same ...] to produce an
identical binary to the one running on the device so you can confirm
that it's exactly the same code.  That's separate from being able to
upload a changed binary.

Bron.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Willy Tarreau]

On Wed, Jun 13, 2007 at 03:06:51PM -0700, Linus Torvalds wrote:
> And I actually am of the very firm opinion that a world with gray areas 
> (and purple, and pink, and green) is a hell of a lot better than one where 
> everything is black-and-white.

agreed, because you cannot imagine at the beginning all fair uses of your
project. It's a good thing that people can use it, thinking "hey, it's not
explicitly allowed but I think I can defend my case".

> Only lawyers want a black-and-white world.

not really. They would lose their job. They need a gray world to get
customers, but they want to decide what half is black and what half
is white in front of the judge depending on their customer's needs.

> Indeed. And it's _fine_ to even be in it "just to make a quick buck". We 
> do want all kinds of input. I think the community is much healthier having 
> lots of different reasons for people wanting to be involved, rather than 
> concentrating on just some specific reason.
> 
> For some it's the technology. For some it's the license. For some it's 
> just a thing to pass boredom. Others like to learn. Whatever. It's all 
> good!

And I think that for many people (including myself), it's all of these in
this order :
  - something to learn (when you're at school)
  - something to pass boredom (when you're at school too)
  - the technology (when you're working on designing new products)
  - the license (when you finally try to put your products on the market)

Regards,
Willy

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 2704 bytes --]

> > > The fact is, Tivo didn't take those rights away from you, yet the FSF
> > > says that what Tivo did was "against the spirit". That's *bullshit*.
> >
> > Oh, good, let's take this one.
> >
> >   if you distribute copies of such a program, [...]
> >   you must give the recipients all the rights that you have
> >
> > So, TiVo includes a copy of Linux in its DVR.
> >
> 
> And they give you the same right that they had, which is obtain free software 
> that you can modify and redistribute. There's nothing in there that says they 
> should give you the tools they used after they received the software, which 
> is what you seem to be looking for.

IANAL so I won't comment on the legal aspects of TiVo's doing.

However it definitely is against _MY_ understanding of the spirit
of the GPL. At least to me that's quite obvious. I'm sure you all
know the story of the printer driver RMS couldn't fix that reportedly
made him start the whole FSF business.

Looking at what TiVo did I realize glaring similarities.

<disclaimer>
I'm in no way related with the FSF. I hereby state I'm not parroting
anyone's else position but have come to this conclusion solely on
my own.
</disclaimer>

> > TiVo retains the right to modify that copy of Linux as it sees fit.
> >
> > It doesn't give the recipients the same right.
> 
> It does, can't you modify their kernel source? Where does it say you should be 
> able to run you modifications on the same hardware?

Come on! The whole idea of software is to have it run on some HW.
Why would I want to change it in the first place if I can't run it ?

If what they did is actually allowed by the wording of the legal phrases
of the GPLv2 then that IMO is a loophole w/r to the spirit (as I understand)
it and IMO should be plugged.

> The only fear that I have with the whole Tivo saga, is that companies like 
> Dell can use the same thing to say: "Our hardware will only run Company's X 
> distribution of Linux". 

Would not such a restriction voilate the spirit of the GPL ?

Anyway, my simplistic view is:
Once it is under the GPL I could change it and actually make the
changes work as I see fit.

That's what I think my freedom as of the GPL is about.

Now all that needs to be done is make sure the legal phrases are such
that they convince the judges they actually mean this in court too.

Best wishes,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Bron Gondwana <brong@fastmail.fm> wrote:

> Tivo gets sick of the endless flamewars on lkml, signs a copy
> of QNX, pushes it out to the hardware.  No more Linux on Tivo.

What do we lose?

Do we actually get any benefit whatsoever from TiVO's choice of Linux
as the kernel for its device?

Do TiVO customers lose anything from the change from one non-Free
software to another?  (the Linux binary, as shipped in the TiVO, has
become non-Free)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Daniel Forrest <forrest@lmcg.wisc.edu> wrote:

> 1.) I ship the device back to the manufacturer, they replace the ROM,
>     and ship it back to me.

> 2.) I ship the device back to the manufacturer, they load new code
>     into it, and ship it back to me.

> How do these two differ?  Or is it now just a question of the ROM
> being in a socket?  I can't see how the technicalities of how the
> hardware is constructed can change the legality of the software.

I don't see that they differ.  If the software can be replaced, the
manufacturer ought to tell you how to do it.  It doesn't have to do it
for you, it doesn't have to give you the hardware tools needed to do
it, but if you're not able to start from the source code and the
information provided by the manufacturer and get to a modified version
of the software on the device, while the manufacturer could do it,
then the manufacturer is locking you in, and therefore you're not
free.  This is a clear violation of the spirit of the license, even if
the legalese might make room for some such misbehavior.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> User B buys the router and modifies the kernel so it drives the WiFi to an 
> output power twice that which it is licensed to carry.
> FCC finds out and prosecutes User B for violating the regulations.

Ok so far.

> FCC then pulls the small companies license until they change their
> hardware so the driver can't push it to transmit at a higher power
> level and levies a fine.

I'd say this is unfair, but if it can happen, then maybe the small
company could have been more careful about the regulations.  There are
various ways to prevent these changes that don't involve imposing
restrictions of modification on any software in the device, all the
way from hardware-constrained output power to hardware-verified
authorized configuration parameters.

> Growing the base of installed GPL covered software,

When this doesn't bring freedom to people, when people can't actually
enjoy the freedoms that the software is supposed to provide, I don't
see why this would be a good thing.  What's the merit in being able to
claim "vendor X chose my Free Software and locked it down such that
users don't get the freedoms I meant for them, and I'm happy about it?"

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 06:44:25AM +0200, Michael Gerdau wrote:
> > able to run you modifications on the same hardware?
                                         ^^^^
> Come on! The whole idea of software is to have it run on some HW.
                                                           ^^^^
> Why would I want to change it in the first place if I can't run it ?


See the difference?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 1765 bytes --]

> > As a PS to the GPL3 comment here is the basic difference
> > 
> > ROM	-	I can't modify the code on the device
> > 		The creator can't modify the code further on the device
> > 
> > Tivo	-	I can't modify the code on the device
> > 		The owner can modify the code
> > 
> > One is an implicit limitation of the hardware (just like I can't run
> > openoffice on a 4MB PC even though the license gives me the right to
> > try), the other is an artificial restriction.
> > 
> > One case is witholding freedom in the GPL sense by one party while
> > keeping it themselves, the other is a limitation of the system
> > inevitably imposed on everyone.
> 
> I've been following this discussion and I find this interesting.
> Consider these two cases:
> 
> 1.) I ship the device back to the manufacturer, they replace the ROM,
>     and ship it back to me.
> 
> 2.) I ship the device back to the manufacturer, they load new code
>     into it, and ship it back to me.
> 
> How do these two differ?  Or is it now just a question of the ROM
> being in a socket?  I can't see how the technicalities of how the
> hardware is constructed can change the legality of the software.

At first glance I think a construct where the manufacturer is obliged
to load _MY_ modified software in a timely fashion and at a reasonable
price into the device would fit my understanding of the GPL's spirit
though this leaves room for the definition of timely...

Best,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On Wednesday 13 June 2007 21:59, Alexandre Oliva wrote:
> > It *DOES* *NOT* say "All rights that you have". It says "All rights
> > that are granted you by this license".
> 
> I suggest you to reboot into memtest ;-)  The preamble of GPLv2 says:
> 
>   For example, if you distribute copies of such a program, whether
>   gratis or for a fee, you must give the recipients
>   all the rights that you have.
>   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^

So if I am a sole author of a program and I chose to distribute it under
GPL then all recepients will get _all_ my rights, including right to
re-license the program under BSD or a proprietory license? Yeah, riiight...

Thankfully it is just preamble and not the actual license text.
 
-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 497 bytes --]

> In Germany, not America. I should have qualified my statement to make it clear 
> I mean "In America". Sorry about the confusion.

You shouldn't say "America" when you mean the "US".

Best wishes,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Theodore Tso]

On Thu, Jun 14, 2007 at 12:00:17AM -0400, Valdis.Kletnieks@vt.edu wrote:
> Incidentally, this same logic is what drives the average successful patent
> troll lawsuit - the sued company will buy a license for $25K, just because
> they know that fighting the lawsuit will cost $100K and up.

You're off by a factor of 10-50.  The usual estimates I've heard from
people who ought to know is the minimum ante for fighting a patent
lawsuit is $1 million to $5 million.  Lawyer time and expert witness
time to give the judge a granduate education in the technologies
involved is *expensive* (since the judge may be really smart, but most
judges have no engineering background to speak of, so you have to
explain the technologies involved in terms that make sense to someone
with an honors education with a Bachelor of Arts degree).

Basically, in the US, you get the best justice money can buy.  :-)

							- Ted

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> I've never had a reason to want to change the way any device like a TiVO 
> works. So I can't comment on this.

Have you never wanted to improve any aspect of the software in your
cell phone?  In your TV, VCR, DVD player, anything?  In the microwave
oven, maybe?

> On Wednesday 13 June 2007 22:38:05 Alexandre Oliva wrote:
>> In 95% of the desktop computers, you can't make changes to the OS that
>> runs on it.  Whom is this good for?

> Faulty logic. I have yet to find a computer that I couldn't change the OS on. 

I was not talking about installing another OS, I was talking about
making changes to the OS.  As in, improving one particular driver,
avoiding a blue screen, stuff like that.

> Or do you mean "transferring the recorded copies off the TiVO 
> and on to a different medium"?

Sure.  Such that I can watch shows while wasting time in public
transportation, in an airplane, whatever.

> DRM, I do agree, gets in the way of "Fair Use".

And the fact that TiVO can be, and has been modified remotely to add
restrictions on what users could do, means nothing you do with it is
safe.  You, and everything you've recorded with the TiVO, are at the
mercy of this one company.

> So you're not concerned that you're potentially pushing companies
> that would otherwise be major consumers of GPL'd software away? That
> doesn't make sense to me.

What would their consuming GPL software buy us, if they won't respect
users' freedoms, which is the very reason behind the GPL?

Heck, if they don't want to play by the rules, that's up to them.  But
then they shouldn't use the software at all.

Yeah, I wish they'd rather play by the rules, but if they don't want
to, too bad, for us and for them.

> Why should I repeat Linus' explanation of the ways that GPLv3 violates the 
> spirit of GPLv2?

Don't worry about parrotting here, he hasn't provided that explanation
yet ;-)  Please give it a try.

BTW, what license is Linux licensed under?  It's GPLv2 plus userland
exception, right?  (There's some additional module exception, right?)

> And why shouldn't I pose it as a matter of "Personal Taste"? The
> biggest and most powerful voice in the FSF says "I don't like
> Tivoization" and "I don't like DRM" and when the GPLv3 appears it
> has language that makes those violations of the license.

Have you ever wondered *why* he doesn't like them?

Could it possibly be because they harm the goal of his life, which is
to enable people to live their digital lives in freedom?

> Just like people have started using "GNU/Linux" or "GNU+Linux" to
> refer to Linux

No, no, you got it wrong.  Linux is the kernel.  GNU was the
nearly-complete operating system it fit in.  GNU+Linux is a complete
operating system.

And you don't have to believe me, believe Linus, the initial author of
Linux:

http://www.kernel.org/pub/linux/kernel/Historic/old-versions/RELNOTES-0.01

  Although linux is a complete kernel

  Sadly, a kernel by itself gets you nowhere. To get a working system
  you need a shell, compilers, a library etc. These are separate parts
  and may be under a stricter (or even looser) copyright. Most of the
  tools used with linux are GNU software and are under the GNU
  copyleft.

> A "TiVO" is not, and has never been, a "General Purpose
> Computational Device".

Err...  Last I looked it was a bunch of general-purpose components,
packaged in a way that made it not look like a general-purpose
computer.  Who gets to decide?  And with what motivations?

> Exactly. And I don't see anything about a TiVO (or any device that, like a 
> TiVO, requires binaries that run on it to be digitally signed) that stops you 
> from exercising the "freedoms" guaranteed by the GPL.

  2. You may modify your copy or copies of the Program or any portion
  of it

>> > if you upload a modified linux kernel to your wireless router that
>> > gives it a 2000 foot range, you've just broken the law

>> At which point, you get punished by the law system.

> But the GPLv2 gives companies a chance to protect themselves from legal 
> actions by people that are sure to follow.

So does the GPLv3.  It might be a bit narrower, to cut on other kinds
of abuses, but all constraints I'm aware of that are mandated by law
can still be achieved.  The point is to forbid disrespecting users'
freedoms to modify the software.  Configuration parameters for the
hardware, needed to comply with regulations, can be easily taken care
of without disrespecting users' freedoms.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, david@lang.hm wrote:

> since the latest draft of the GPLv3 now discriminates against some
> uses (industrial vs commercial I think are the terms used) 

  A "User Product" is either (1) a "consumer product," which means any
  tangible personal property which is normally used for personal,
  family, or household purposes, or (2) anything designed or sold for
  incorporation into a dwelling.

> does it even qualify as a Open Source lincense anymore by the OSI
> terms?

The definition is about the hardware, not the software, so it may
still qualify.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Valdis.Kletnieks@vt.edu wrote:

> If a company sells you hardware that includes a ROM that contains GPL'ed
> software, are they in violation of the GPL if they don't include a ROM burner
> in the hardware?  Or are ROM burners like compilers, where you have to supply
> your own?

  this requirement does not apply if neither you nor any third party
  retains the ability to install modified object code on the User
  Product (for example, the work has been installed in ROM).

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Adrian Bunk <bunk@stusta.de> wrote:

> If the two courts are in the same country there's usually a higher court 
> above both that can resolve this. But what if let's say the highest 
> court in the USA and the highest court in Germany would disagree on such 
> a matter?

Upgrade the license so as to provide guidance as to the intent of the
authors, such that the disagreement doesn't happen again.

If there's room in each country's laws to fix the problem, that is.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bongani Hlope]

On Thursday 14 June 2007 02:55:52 Alexandre Oliva wrote:
> On Jun 13, 2007, Bongani Hlope <bhlope@mweb.co.za> wrote:
> > On Thursday 14 June 2007 01:49:23 Alexandre Oliva wrote:
> >> if you distribute copies of such a program, [...]
> >> you must give the recipients all the rights that you have
> >>
> >> So, TiVo includes a copy of Linux in its DVR.
> >
> > And they give you the same right that they had, which is obtain free
> > software that you can modify and redistribute. There's nothing in there
> > that says they should give you the tools they used after they received
> > the software, which is what you seem to be looking for.
>
> Can they modify the software in their device?
>
> Do they pass this right on?
>
> >> TiVo retains the right to modify that copy of Linux as it sees fit.
> >>
> >> It doesn't give the recipients the same right.
> >
> > It does, can't you modify their kernel source?
>
> It's not the kernel source.  That's not where the TiVo anti-tampering
> machinery blocks modifications.
>
> It's about that copy of the kernel that ships in the device in object
> code.  That's the one that TiVo customers ought to be entitled to
> modify, if TiVo can modify it itself.
>
> > Where does it say you should be able to run you modifications on the
> > same hardware?
>
> Where it says that you should pass on all the rights that you have.
>
> While TiVo retains the ability to replace, upgrade, fix, break or make
> any other change in the GPLed software in the device, it ought to pass
> it on to its customers.

So according to your logic, I can go to Sharp's website and download the GPL 
source code for their Zaurus. But I don't own a Sharp Zaurus; to keep with 
your interpretation of the spirit of GPL, they have to give me a Zaurus so 
that I can run my modifications on the same hardware?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Wed, 13 Jun 2007, Alexandre Oliva wrote:
>> 
>> So, TiVo includes a copy of Linux in its DVR.  

> Stop right there.

> You seem to make the mistake to think that software is something physical.

Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
require some physical support.  That's the hard disk in the TiVO DVR,
in this case.  I don't see how this matters, though.

>> TiVo retains the right to modify that copy of Linux as it sees fit.

>  (b) They never modified "a copy" of Linux - they simply replaced it with
>      "another copy" of Linux. The only thing that actually got *modified* 
>      was their hardware!

Per this reasoning, nobody never modifies software.  When you open a
source file in your editor, you make changes to it, then save it,
you're not modifying it, you're replacing it with another copy, and
the only thing that actually got modified was the hardware.

Maybe look what "modify" means in legal context?

Then refer to the GPL:

  2. You may modify your copy or copies of the Program or any portion
  of it, thus forming a work based on the Program,

> And their hardware (and firmware) will run some integrity checks on 
> *whatever* copies of software they have.  This is all totally outside 
> Linux itself.

Agreed.  But as it turns out they use these checks to stop people from
modifying the copy of Linux they ship in the device, and this
restriction is a GPL violation because they don't provide information
you need to build a functioning modified version.

> Btw, according to your _insane_ notion of "a copy" of software, you can 
> never distribute GPL'd software on a CD-ROM, since you've taken away the 
> right of people to modify that CD-ROM by burning and fixating it.

You don't retain that right yourself.  When you pass that copy on, you
pass it on with all the rights that you have.  No problem here.  This
is no different from the software on ROM.

> And I'm saying that the GPLv2 can mroe straightforwardly be read the way I 
> read it - to talk about software, and to realize that software is not "a 
> copy", it's a more abstract thing.

If you choose to disregard the legal meaning of the legal terms used
in the GPLv2, you may have a point.

> that means that they have to give you access to and control over the
> SOFTWARE.

Yes.  That's all I'm saying.  You just can't use the hardware to take
that control away.  That would be a violation of the license.

> Face it: the GPLv3 is a _new_ license. Making funamentally _different_ and 
> _new_ restrictions that do not exist in the GPLv2,

This is true.

> and do not exist in the preamble.

This is not true.

The spirit remains the same: let people modify and share the software.

If the binary you got can't be created out of the corresponding
sources, something is missing.  If it won't run without this missing
bit, you're missing functional portions of the source code.  This all
means the hardware is being used to impose a restriction on
modification of the software, which is against the spirit of the GPL,
and quite likely against its letter as well.

If you don't want it to be so, you can always add an additional
permission that clarifies this bit, such that TiVO and you will be
happy.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Adrian Bunk wrote:
>> 
>> "For an executable work, complete source code means all the source code 
>> for all modules it contains, plus any associated interface definition 
>> files, plus the scripts used to control compilation and installation of 
>> the executable."
>> 
>> The question is whether this includes private keys.

> No. That's the question as the FSF would like to frame it.

No.  The FSF actually does *not* want to take this position.  That's
why it chose the formulation of Installation Instructions.  It doesn't
share my view that the keys needed to sign a binary in order for it to
work are part of the source code.

> And you could actually replace their copy of Linux with another one. It 
> would have to have the same SHA1 to actually start _running_, but that's 
> the hardware's choice. 

That's the hardware imposing a restriction on modification of the
software.  It doesn't matter how elaborate the excuse is to justify
denying users' freedoms: it's against the spirit of the GPL, and the
GPL will be amended as needed to plug such holes.

> So take another example: I obviously distribute code that is copyrighted 
> by others under the GPLv2. Do I follow the GPLv2? I sure as hell do! But 
> do I give you the same rights as I have to modify the copy on 
> master.kernel.org as I have? I sure as hell DO NOT!

That's an interesting argument.

People don't get your copy, so they're not entitled to anything about
it.

When they download the software, they get another copy, and they have
a right to modify that copy.

> And here's a big clue for people: anybody who thinks that I'm violating 
> the GPLv2 by not giving out my private SSH key to master.kernel.org is a 
> f*cking moron!

Agreed, except I'd probably use a lighter term.

> See any parallels here? Any parallel to a CD-ROM distribution, or a Tivo 
> distribution?

Yes.  You see how TiVO is different?  It is modifyable, and I actually
receive the copy that TiVO can still modify, but I can't.

> The rights that the GPLv2 gives to "the software", is to something
> much bigger than "the particular copy of the software".

Indeed, it's something bigger.  But this doesn't exclude the smaller
things, does it?

> Can people really not see the difference between "the software" and "a 
> particular encoded copy of the software"? 

There is a difference.  But the GPL doesn't limit itself to the
former.  It explicitly talks about "copies".

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 01:51:13 Alexandre Oliva wrote:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > I've never had a reason to want to change the way any device like a TiVO
> > works. So I can't comment on this.
>
> Have you never wanted to improve any aspect of the software in your
> cell phone?  In your TV, VCR, DVD player, anything?  In the microwave
> oven, maybe?

Nope. I've been tempted several times, but decided that the extra bits I'd 
thought about wouldn't add anything to the device.

> > On Wednesday 13 June 2007 22:38:05 Alexandre Oliva wrote:
> >> In 95% of the desktop computers, you can't make changes to the OS that
> >> runs on it.  Whom is this good for?
> >
> > Faulty logic. I have yet to find a computer that I couldn't change the OS
> > on.
>
> I was not talking about installing another OS, I was talking about
> making changes to the OS.  As in, improving one particular driver,
> avoiding a blue screen, stuff like that.

Ah, well... In the case of "Windos" and other proprietary OS's I try to 
educate people and get them to switch. I don't, personally, have any 
computers that run Windows (and I switched my Palm back to PalmOS because it 
wasn't getting the same performance under Linux - which rather surprised me. 
And rather than fight with it I just switched it back.

> > Or do you mean "transferring the recorded copies off the TiVO
> > and on to a different medium"?
>
> Sure.  Such that I can watch shows while wasting time in public
> transportation, in an airplane, whatever.

Under the US Copyright law I'm not sure that making a "second copy" like that 
is legal. IIRC, "Fair Use" only allows for one copy.

> > DRM, I do agree, gets in the way of "Fair Use".
>
> And the fact that TiVO can be, and has been modified remotely to add
> restrictions on what users could do, means nothing you do with it is
> safe.  You, and everything you've recorded with the TiVO, are at the
> mercy of this one company.

As has been noted in their TOS and the licenses for the hardware from the 
start. The FSF itself explicitly reserves the right to change the GPL at any 
time - which is no different. (when you remove all the bits explaining the 
purpose of the license)

> > So you're not concerned that you're potentially pushing companies
> > that would otherwise be major consumers of GPL'd software away? That
> > doesn't make sense to me.
>
> What would their consuming GPL software buy us, if they won't respect
> users' freedoms, which is the very reason behind the GPL?

I'm not referring to companies that are embedding GPL'd software in their 
products. The companies I'm referring to are the ones that would like to use 
GPL'd software internally. A lot of them would probably have private 
modifications that would never be distributed - and under the GPLv2 it is 
clear that you can keep modifications private as long as you don't distribute 
them. "Pushing them away" means that they'd not do that because they would be 
concerned that the license will change under them in such a way that even 
those private modifications need to be released to the public.

(and don't try to argue that even though those modifications are truly private 
(to the company) they should be released anyway to comply with the "spirit" 
of the license. It is made clear that it isn't by the text of the license 
itself)

> Heck, if they don't want to play by the rules, that's up to them.  But
> then they shouldn't use the software at all.
>
> Yeah, I wish they'd rather play by the rules, but if they don't want
> to, too bad, for us and for them.
>
> > Why should I repeat Linus' explanation of the ways that GPLv3 violates
> > the spirit of GPLv2?
>
> Don't worry about parrotting here, he hasn't provided that explanation
> yet ;-)  Please give it a try.

But he has. Whether you have accepted that his explanations are valid or not 
doesn't change the fact.

> BTW, what license is Linux licensed under?  It's GPLv2 plus userland
> exception, right?  (There's some additional module exception, right?)

The kernel itself is GPLv2 (only). Individual components - even individual 
files - have other licenses or retain the "any later version" clause. 
(Someone pointed out, earlier in this thread, that there is GPLv1.1 code in 
the kernel)

> > And why shouldn't I pose it as a matter of "Personal Taste"? The
> > biggest and most powerful voice in the FSF says "I don't like
> > Tivoization" and "I don't like DRM" and when the GPLv3 appears it
> > has language that makes those violations of the license.
>
> Have you ever wondered *why* he doesn't like them?

Not really. I've always figured he had reasons similar to mine for not liking 
DRM. As to his dislike of "Tivoization", well, that I've always attributed to 
the fact that someone at that company managed to outsmart him. (and no, I 
wasn't being serious with that last line)

> Could it possibly be because they harm the goal of his life, which is
> to enable people to live their digital lives in freedom?
>
> > Just like people have started using "GNU/Linux" or "GNU+Linux" to
> > refer to Linux
>
> No, no, you got it wrong.  Linux is the kernel.  GNU was the
> nearly-complete operating system it fit in.  GNU+Linux is a complete
> operating system.

Yet I still find people that insist that the *ENTIRE* system - kernel and 
all - is a GNU project. Not just the common idiot you'd find on almost any 
street in the US, but also educated people in technical fields. The reason is 
the name. 

*AND* you cut out the bit where I said "I have no problems with it"

> And you don't have to believe me, believe Linus, the initial author of
> Linux:
>
> http://www.kernel.org/pub/linux/kernel/Historic/old-versions/RELNOTES-0.01
>
>   Although linux is a complete kernel
>
>   Sadly, a kernel by itself gets you nowhere. To get a working system
>   you need a shell, compilers, a library etc. These are separate parts
>   and may be under a stricter (or even looser) copyright. Most of the
>   tools used with linux are GNU software and are under the GNU
>   copyleft.

Never claimed otherwise. The problem is that using a composite name like that 
*does* confuse a hell of a lot of people. See my statements above about that.

> > A "TiVO" is not, and has never been, a "General Purpose
> > Computational Device".
>
> Err...  Last I looked it was a bunch of general-purpose components,
> packaged in a way that made it not look like a general-purpose
> computer.  Who gets to decide?  And with what motivations?

And so is every game console. But until the original XBox was released nobody 
tried using one as a "General Purpose" machine. The TiVO wasn't designed as a 
general purpose machine - it was designed for a specific purpose. That the 
*easiest* design to produce uses a bunch of general purpose components is an 
economic choice, nothing else.

I will not, however, argue about this anymore. As with other bits, we've 
reached a point where we disagree and no amount of explanation will change 
the others viewpoint. (I hope you understand mine as well as (I think) I 
understand yours)

> > Exactly. And I don't see anything about a TiVO (or any device that, like
> > a TiVO, requires binaries that run on it to be digitally signed) that
> > stops you from exercising the "freedoms" guaranteed by the GPL.
>
>   2. You may modify your copy or copies of the Program or any portion
>   of it
>
> >> > if you upload a modified linux kernel to your wireless router that
> >> > gives it a 2000 foot range, you've just broken the law
> >>
> >> At which point, you get punished by the law system.
> >
> > But the GPLv2 gives companies a chance to protect themselves from legal
> > actions by people that are sure to follow.
>
> So does the GPLv3.  It might be a bit narrower, to cut on other kinds
> of abuses, but all constraints I'm aware of that are mandated by law
> can still be achieved.  The point is to forbid disrespecting users'
> freedoms to modify the software.  Configuration parameters for the
> hardware, needed to comply with regulations, can be easily taken care
> of without disrespecting users' freedoms.

Let me quote Linus here:

But I think the whole thing is totally misguided, because the fact is, the 
GPLv2 doesn't talk about "in place" or "on the same hardware".

In other words, GPLv3 is breaking with its predecessor - it's adding a 
requirement that doesn't exist in previous versions. *AND* it's dictating 
terms for *HARDWARE* when it isn't a hardware license. If I release software 
under the GPL and somebody modifies it to run on a different hardware 
platform I'll be happy, even if they don't send me a patchset for the new 
version. If I create a piece of hardware and run Linux on it, but have it 
locked to a specific version or versions from a specific source (ie: me) and 
release it to the public, I *WILL* release the version of Linux I'm running 
on it. What I won't do is release whatever tools and such that are needed to 
make the hardware run a different version of the kernel. Why? Because: the 
hardware was designed so that a specific version of the kernel runs without 
problems, there is hardware that is very picky and running a customized 
kernel could cause that hardware to fail, etc... There are more reasons than 
the legal protection I previously mentioned. Not a single one of them 
is "Because I want to restrict peoples freedoms re: the GPL'd software". 

Admittedly the two examples I chose aren't very realistic, but those were the 
first two examples that came to mind.

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 01:39:13 Michael Gerdau wrote:
> > In Germany, not America. I should have qualified my statement to make it
> > clear I mean "In America". Sorry about the confusion.
>
> You shouldn't say "America" when you mean the "US".

Sorry, I slipped. I'm still trying to rid myself of the uniquely "US" belief 
that "America" == "USA". Thanks for the reminder.

DRH

>
> Best wishes,
> Michael



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Dmitry Torokhov <dtor@insightbb.com> wrote:

> On Wednesday 13 June 2007 21:59, Alexandre Oliva wrote:
>>   For example, if you distribute copies of such a program, whether
>>   gratis or for a fee, you must give the recipients
>>   all the rights that you have.
>>   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^

> So if I am a sole author of a program and I chose to distribute it under
> GPL

then you're not a licensee, you're a licensor, and these terms don't
apply to you.  Already covered upthread BTW.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Bongani Hlope <bhlope@mweb.co.za> wrote:

> On Thursday 14 June 2007 02:55:52 Alexandre Oliva wrote:

>> While TiVo retains the ability to replace, upgrade, fix, break or make
>> any other change in the GPLed software in the device, it ought to pass
>> it on to its customers.

> So according to your logic, I can go to Sharp's website and download the GPL 
> source code for their Zaurus. But I don't own a Sharp Zaurus; to keep with 
> your interpretation of the spirit of GPL, they have to give me a Zaurus so 
> that I can run my modifications on the same hardware?

Sharp can modify the copy of the code in your Zaurus as much as you
do, when you don't have a Zaurus.  I don't see how you can get to the
conclusion that they have to give you a Zaurus, when all you're
getting is software.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bron Gondwana]

On Thu, Jun 14, 2007 at 01:58:26AM -0300, Alexandre Oliva wrote:
> On Jun 14, 2007, Bron Gondwana <brong@fastmail.fm> wrote:
> 
> > Tivo gets sick of the endless flamewars on lkml, signs a copy
> > of QNX, pushes it out to the hardware.  No more Linux on Tivo.
> 
> What do we lose?
> 
> Do we actually get any benefit whatsoever from TiVO's choice of Linux
> as the kernel for its device?

Sure, if they make any changes or fixes to Linux.  Other than that,
only the same benefit that Microsoft get from Windows piracy - TiVo
employees become familiar with Linux and are more likely to use it
and maybe contribute more in another job later.

What we don't get is TiVo having a better kernel than everyone else
because they've put some work into extending it without giving that
work back.

I see stuff in arch/powerpc/kernel/ which is Copyright "TiVo, Inc"
and more recent stuff in usb/net/asix.c and usb/net/mcs7830.c which
is more than I've ever contributed to the kernel, despite making
extensive use and even selling services where I ran servers with
Linux on them but didn't allow my customers to change the kernel
on the servers if there was some feature they wanted to play with.

> Do TiVO customers lose anything from the change from one non-Free
> software to another?  (the Linux binary, as shipped in the TiVO, has
> become non-Free)

Not particularly, no.  Other than maybe some nice features that TiVo
gains from being able to use Linux.

Bron.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 02:36:12 Alexandre Oliva wrote:
> On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > On Thu, 14 Jun 2007, Adrian Bunk wrote:
> >> "For an executable work, complete source code means all the source code
> >> for all modules it contains, plus any associated interface definition
> >> files, plus the scripts used to control compilation and installation of
> >> the executable."
> >>
> >> The question is whether this includes private keys.
> >
> > No. That's the question as the FSF would like to frame it.
>
> No.  The FSF actually does *not* want to take this position.  That's
> why it chose the formulation of Installation Instructions.  It doesn't
> share my view that the keys needed to sign a binary in order for it to
> work are part of the source code.
>
> > And you could actually replace their copy of Linux with another one. It
> > would have to have the same SHA1 to actually start _running_, but that's
> > the hardware's choice.
>
> That's the hardware imposing a restriction on modification of the
> software.  It doesn't matter how elaborate the excuse is to justify
> denying users' freedoms: it's against the spirit of the GPL, and the
> GPL will be amended as needed to plug such holes.

And? There is *absolutely* *nothing* in any version of the GPL *prior* to 3 
that says that hardware cannot impose restrictions. What the GPL *does* say 
is that you can't "add additional restrictions to the license" - (IMHO) a 
piece of hardware having a restriction isn't an "additional restriction added 
to the license". As well, as Linus stated, there is nothing *anywhere* - 
AFAICT, not even in GPLv3 - that says that you have to be able to run the 
software "in place" or "on the same hardware".

If a hardware manufacturer - like TiVO - uses GPL'd code in their product - 
and complies with the terms of the license - they aren't required to allow 
you to run modified code on that hardware. Without it mentioned anywhere in 
the GPL *OR* the assorted writings of RMS (who founded the FSF and wrote the 
original GPL) that "modified software must be able to run on the same 
hardware" then it cannot be in the "spirit" of the license to allow this.

> > So take another example: I obviously distribute code that is copyrighted
> > by others under the GPLv2. Do I follow the GPLv2? I sure as hell do! But
> > do I give you the same rights as I have to modify the copy on
> > master.kernel.org as I have? I sure as hell DO NOT!
>
> That's an interesting argument.
>
> People don't get your copy, so they're not entitled to anything about
> it.
>
> When they download the software, they get another copy, and they have
> a right to modify that copy.

But you get the TiVO corporations copy of the software? I smell a logical 
fallacy here, but can't remember the name for it.

> > And here's a big clue for people: anybody who thinks that I'm violating
> > the GPLv2 by not giving out my private SSH key to master.kernel.org is a
> > f*cking moron!
>
> Agreed, except I'd probably use a lighter term.
>
> > See any parallels here? Any parallel to a CD-ROM distribution, or a Tivo
> > distribution?
>
> Yes.  You see how TiVO is different?  It is modifyable, and I actually
> receive the copy that TiVO can still modify, but I can't.

I don't. You don't get the TiVO corporations copy of the software. You get 
your own copy, with all the rights that TiVO had when receiving the software. 
The right to install and run the kernel in the TiVO device is independent of 
the rights to copy, modify, distribute and run the software. (because the GPL 
never guarantees you the right to run the software on a particular piece of 
hardware.)

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Thursday 14 June 2007 01:51:13 Alexandre Oliva wrote:
>> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
>> > On Wednesday 13 June 2007 22:38:05 Alexandre Oliva wrote:
>> >> In 95% of the desktop computers, you can't make changes to the OS that
>> >> runs on it.  Whom is this good for?

>> > Faulty logic. I have yet to find a computer that I couldn't change the OS
>> > on.

>> I was not talking about installing another OS, I was talking about
>> making changes to the OS.  As in, improving one particular driver,
>> avoiding a blue screen, stuff like that.

> Ah, well... In the case of "Windos" and other proprietary OS's I try to 
> educate people and get them to switch.

Good.  So I presume you'd tell them to switch away from a
turned-proprietary GNU/Linux operating system as well, right?

So, again, what do we gain if companies abuse the GPL and disrespect
users' rights that we meant them to respect?

>> > Or do you mean "transferring the recorded copies off the TiVO
>> > and on to a different medium"?

>> Sure.  Such that I can watch shows while wasting time in public
>> transportation, in an airplane, whatever.

> Under the US Copyright law I'm not sure that making a "second copy"
> like that is legal. IIRC, "Fair Use" only allows for one copy.

Even if you delete the "first copy"?

Actually, I thought fair use in US entitled you to make a backup copy.
So the copy in your TiVO would be your original, and the external copy
would be your fair-use backup.

> As has been noted in their TOS and the licenses for the hardware from the 
> start.

If it is used to disrespect the inalienable freedoms associated with
the GPL software in the device, it seems like a license violation to
me.

> The FSF itself explicitly reserves the right to change the GPL at any 
> time - which is no different.

Actually, it's completely different.

If the FSF revises the GPL, the old version remains available for
anyone to use for any new software, and all software released under
the old version remains available under that old version.

In contrast, your TiVO may get a software upgrade without your
permission that will take your rights away from that point on, and
there's very little you can do about it, other than unplugging it from
the network to avoid the upgrade if it's not too late already.

> A lot of them would probably have private modifications that would
> never be distributed - and under the GPLv2 it is clear that you can
> keep modifications private as long as you don't distribute them.

Likewise with GPLv3.

> "Pushing them away" means that they'd not do that because they would
> be concerned that the license will change under them in such a way
> that even those private modifications need to be released to the
> public.

This would not only change the spirit of the license, but turn it into
a non-Free Software license.

And then, again, the license can't possibly be changed from under
them.  A new revision of the GPL would only affect software licensed
under that new revision.  If you already got it under an earlier
revision, you know what you got, and nobody can take that away from
you.

> (and don't try to argue that even though those modifications are
> truly private (to the company) they should be released anyway to
> comply with the "spirit" of the license. It is made clear that it
> isn't by the text of the license itself)

How could you possibly come to the conclusion that forcing anyone to
release private modifications would be in compliance with the spirit
of the license?  can != must

>> > Why should I repeat Linus' explanation of the ways that GPLv3 violates
>> > the spirit of GPLv2?

>> Don't worry about parrotting here, he hasn't provided that explanation
>> yet ;-)  Please give it a try.

> But he has. Whether you have accepted that his explanations are
> valid or not doesn't change the fact.

His explanation is based on a reading of the license that doesn't
match what its authors meant.  I guess the authors know better what
they meant the spirit of the license to be than someone else who
studied it a lot but that until very recently couldn't even tell the
spirit from the legal terms.

>> > Just like people have started using "GNU/Linux" or "GNU+Linux" to
>> > refer to Linux

>> No, no, you got it wrong.  Linux is the kernel.  GNU was the
>> nearly-complete operating system it fit in.  GNU+Linux is a complete
>> operating system.

> *AND* you cut out the bit where I said "I have no problems with it"

Referring to Linux as GNU/Linux would be wrong, because Linux is the
kernel, and that's unrelated with the GNU operating system.  It's the
combination of them that forms GNU+Linux.  And it's referring to this
combination as Linux that is wrong.

I'm sorry that I got the impression that you meant the combination
when you wrote "refer to Linux" above.  It looked like you meant the
combination, since I've never seen anyone call the kernel GNU/Linux or
GNU+Linux.

> Never claimed otherwise. The problem is that using a composite name like that 
> *does* confuse a hell of a lot of people.

Pronouncing the '+' or '/' helps a lot.  GNU plus Linux makes a lot of
sense, and so does GNU on Linux.

>> > A "TiVO" is not, and has never been, a "General Purpose
>> > Computational Device".

>> Err...  Last I looked it was a bunch of general-purpose components,
>> packaged in a way that made it not look like a general-purpose
>> computer.  Who gets to decide?  And with what motivations?

> And so is every game console. But until the original XBox was released nobody 
> tried using one as a "General Purpose" machine. The TiVO wasn't designed as a 
> general purpose machine - it was designed for a specific purpose. That the 
> *easiest* design to produce uses a bunch of general purpose components is an 
> economic choice, nothing else.

So, if I put together a general-purpose computer, a general-purpose
operating system, adding a label "not a general-purpose computer" is
enough to make it so, just so that I can escape the obligations to
respect users' freedoms?

> I will not, however, argue about this anymore.

Fair enough.

> Let me quote Linus here:

> But I think the whole thing is totally misguided, because the fact is, the 
> GPLv2 doesn't talk about "in place" or "on the same hardware".

> In other words, GPLv3 is breaking with its predecessor - it's adding a 
> requirement that doesn't exist in previous versions.

No dispute about this.  The requirements are being added to the legal
terms, precisely such that they better reflect the spirit, under the
light of the new threats that appeared since GPLv2 was published.

But the new requirements do abide by the same spirit, and that was a
promise the FSF made WRT revisions of the GPL.

> *AND* it's dictating terms for *HARDWARE* when it isn't a hardware
> license.

Only in as much as you try to use the hardware as means to disrespect
the spirit of the license and escape from the obligation to respect
users' freedoms.

> If I release software under the GPL and somebody modifies it to run
> on a different hardware platform I'll be happy, even if they don't
> send me a patchset for the new version.

Yup.  See the bit about GPL not being tit-for-tat.

> If I create a piece of hardware and run Linux on it, but have it
> locked to a specific version or versions from a specific source (ie:
> me) and release it to the public, I *WILL* release the version of
> Linux I'm running on it.

Good.

> What I won't do is release whatever tools and such that are needed to 
> make the hardware run a different version of the kernel. Why? Because: the 
> hardware was designed so that a specific version of the kernel runs without 
> problems, there is hardware that is very picky and running a customized 
> kernel could cause that hardware to fail, etc...

Why do you care?  It's no longer your hardware, it's theirs.

Why would you refrain from providing information to others such that
they *could* make the software do what *they* want in their hardware
that they got from you?

If you let them change it and they break it, they get to keep all the
pieces.  Your job is done.  Why get out of your way to stop them from
making the best out of *their* hardware?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Paul Mundt]

On Thu, Jun 14, 2007 at 02:51:13AM -0300, Alexandre Oliva wrote:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> 
> > I've never had a reason to want to change the way any device like a TiVO 
> > works. So I can't comment on this.
> 
> Have you never wanted to improve any aspect of the software in your
> cell phone?  In your TV, VCR, DVD player, anything?  In the microwave
> oven, maybe?
> 
This is perhaps the part that's the most interesting. For the very small
number of people that _do_ want to change these things (usually at the
expense of a voided warranty, in the consumer device case), there's
always a way to make these changes, even if you must resort to hardware
hacking. Trying to mandate this sort of functionality in the license
might make it easier for a few people to get their code loaded, but the
vast majority of users have zero interest in anything like this.

I don't see how you can claim that the vendor is infringing on your
freedom, _you_ made the decision to go out and buy the product knowing
that the vendor wasn't going to go out of their way to help you hack
the device. In many cases the vendor doesn't even have the option
(802.11b channels and certification come to mind, GSM, etc.) of opening
things up to the end user, and making changes to the license isn't going
to magically change any of this.

If you don't like what the vendor has done with the product, you have the
freedom to not support the vendor, and to try and encourage people to
follow suit. As an example, I simply opted not to buy a tivo since I
wasn't able to do what I wanted with it out of the box, rather than
opting to rant about it (or coin an idiotic buzzword) much to the dismay
of every other person on a mailing list. This was neither something I
lost a great deal of sleep over, nor did I at any time feel like my
freedom was being eroded. True story.

If the vendor's bottom line is measurably impacted, they may even
reevaluate their position on supporting device hacking, but it's
certainly not going to be through draconian licensing that vendors
suddenly decide to play nice.

There were certainly enough vendors that followed the letter of the
GPLv2 without following the spirit of the license, with varied benefit
(especially with consumer device vendors). Imposing additional
constraints under the guise of the FSF's current version of "freedom"
isn't going to get these sorts of vendors working any better with the
community. You could of course argue that these vendors have nothing to
offer the community, and so they shouldn't be tolerated at all, but that
assumes that being able to load arbitrary code on their hardware and
the usefulness of whatever contributions they have to make to the software
are directly coupled some how. This has never been the case, and it only
seems to be a mindset that has started circulating when the GPLv3 came
about.

One idly wonders who exactly the FSF feels they're speaking for at this
point..

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Thu, Jun 14, 2007 at 03:07:17AM -0300, Alexandre Oliva wrote:
> On Jun 13, 2007, Adrian Bunk <bunk@stusta.de> wrote:
> 
> > If the two courts are in the same country there's usually a higher court 
> > above both that can resolve this. But what if let's say the highest 
> > court in the USA and the highest court in Germany would disagree on such 
> > a matter?
> 
> Upgrade the license so as to provide guidance as to the intent of the
> authors, such that the disagreement doesn't happen again.
> 
> If there's room in each country's laws to fix the problem, that is.

I don't think that's an option.

Consider the question of whether non-GPL kernel modules are legal at 
all and the number and different opinions of Linux kernel authors.

Plus the general question whether any "upgrade the license" would be 
valid in all jurisdictions ("GPL version 2 or any later version"  
as licence might make it possible, but in all other cases I'd have 
serious doubts).

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Matt Keenan]

Alexandre Oliva wrote:
> On Jun 13, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
>   
>> On Wed, 13 Jun 2007, Alexandre Oliva wrote:
>>     
>>> So, TiVo includes a copy of Linux in its DVR.  
>>>       
>
>   
>> Stop right there.
>>     
>
>   
>> You seem to make the mistake to think that software is something physical.
>>     
>
> Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
> require some physical support.  That's the hard disk in the TiVO DVR,
> in this case.  I don't see how this matters, though.
>
>   
I'm now intrigued, where are these (Brazilian and US) definitions
stipulated, and under what authority?

Matt

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 2685 bytes --]

On Wednesday 13 June 2007 22:14, Krzysztof Halasa wrote:
> It seems so.
>
> > But it has this upgrade option, and one possible interpretation of
>
>       ^^
>
> > Linus' comment is "no, it doesn't have this update option".
>
> It? What "it"?
> I don't get it. If you say the licence is v2 only, then how can it have
> options?

By section 9. The license is v2, and basically allows to update the 
license - and it makes this a choice of the user (who also has rights to 
change stuff and redistribute it).

> > If I use GPL as license, I'm under "GPL regime", i.e. the terms of the
> > GPL apply.
>
> First, the local and international laws apply. It's not like selling your
> soul to the devil.

Contract law means that first and foremost the contract itself defines the 
rules, and only if it is not or contradicts the law, the law jumps in. The 
GPL is not really a contract, it's a license, but the law is not much 
different here, especially once you accept the GPL. If you put your code 
under GPL, the text in the GPL is the deal. The law is only the framework 
under which the deal works.

If you accept the M$ EULA, international law still applies, yet you are 
selling your soul to the devil (because the EULA sais so).

> > Now, I may rewrite those few "GPLv2 only" files, and
> > then I have a GPLv2-or later compatible linux-some.version-bp kernel.
>
> Sure, you can rewrite all non "GPLv2 or later" code and have v3 Linux.
> The problem is you think only "few" files are v2.

Because only few files say so, and they must say what they mean, because GPL 
is rather clear that if you put a file which doesn't say which version 
applies under GPL, it's "any GPL". Why is it so difficult to grok section 9 
of the current GPLv2, which people claim is well understood?

A number of kernel hacker deliberately want their work under GPLv2 only 
(like Al Viro), and they are fully entitled to do that - but they must 
announce it in a propper place (not lkml or lwn.org), and a comment in 
COPYING signed by Linus Torvalds doesn't seem to be propper to me, 
especially when the GPLv2 gives a procedure how to do it (look for the 
appendix: "How to Apply These Terms to Your New Programs").

There are good reasons to follow the advice there, and those who did follow 
the advice in the Linux kernel in the vast majority said "GPLv2 or later". 
Verbatim copy without understanding? Or is it rather that the other people 
who didn't follow the advice didn't read the GPL, and therefore understand 
it even less ;-)?

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 03:11:45 Alexandre Oliva wrote:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Thursday 14 June 2007 01:51:13 Alexandre Oliva wrote:
> >> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> >> > On Wednesday 13 June 2007 22:38:05 Alexandre Oliva wrote:
> >> >> In 95% of the desktop computers, you can't make changes to the OS
> >> >> that runs on it.  Whom is this good for?
> >> >
> >> > Faulty logic. I have yet to find a computer that I couldn't change the
> >> > OS on.
> >>
> >> I was not talking about installing another OS, I was talking about
> >> making changes to the OS.  As in, improving one particular driver,
> >> avoiding a blue screen, stuff like that.
> >
> > Ah, well... In the case of "Windos" and other proprietary OS's I try to
> > educate people and get them to switch.
>
> Good.  So I presume you'd tell them to switch away from a
> turned-proprietary GNU/Linux operating system as well, right?

If that happened I'd be lost. I've tried the various BSD's and found they had 
problems with hardware support and getting a new version of the BSD kernel to 
compile and boot is something of a black art.

The point is moot, though. It can never happen.

> So, again, what do we gain if companies abuse the GPL and disrespect
> users' rights that we meant them to respect?
>
> >> > Or do you mean "transferring the recorded copies off the TiVO
> >> > and on to a different medium"?
> >>
> >> Sure.  Such that I can watch shows while wasting time in public
> >> transportation, in an airplane, whatever.
> >
> > Under the US Copyright law I'm not sure that making a "second copy"
> > like that is legal. IIRC, "Fair Use" only allows for one copy.
>
> Even if you delete the "first copy"?
>
> Actually, I thought fair use in US entitled you to make a backup copy.
> So the copy in your TiVO would be your original, and the external copy
> would be your fair-use backup.

Hrm... Perhaps. 

> > As has been noted in their TOS and the licenses for the hardware from the
> > start.
>
> If it is used to disrespect the inalienable freedoms associated with
> the GPL software in the device, it seems like a license violation to
> me.

As much as the US "Declaration of Independence" and other sources want people 
to believe otherwise there is no such thing as "inalienable rights" 
or "inalienable freedoms". In this case I have been unable to find 
this "inalienable freedom" to run custom versions of software "on the same 
machine" that you received the original copy on anywhere before the GPLv3 - 
and even then it isn't explicitly clear. There is no restriction on your 
right to modify, copy, distribute or run the software as provided by versions 
of the GPL prior to version 3. If this "run modified copies on the same 
hardware you received the original on" *IS* the "spirit" of the license, then 
why isn't it stated anywhere before GPLv3? (After all, the FSF has have 20+ 
years to mention it)

> > The FSF itself explicitly reserves the right to change the GPL at any
> > time - which is no different.
>
> Actually, it's completely different.
>
> If the FSF revises the GPL, the old version remains available for
> anyone to use for any new software, and all software released under
> the old version remains available under that old version.

I'll grant you that. But, at this point, where can I find a copy of the GPLv1 
without having to dig around the net ?

> In contrast, your TiVO may get a software upgrade without your
> permission that will take your rights away from that point on, and
> there's very little you can do about it, other than unplugging it from
> the network to avoid the upgrade if it's not too late already.

And because its a device that connects to their network - and TiVO isn't a 
telecommunications company - they have the right to upgrade and configure the 
software inside however they want. (In the US at least)

> > A lot of them would probably have private modifications that would
> > never be distributed - and under the GPLv2 it is clear that you can
> > keep modifications private as long as you don't distribute them.
>
> Likewise with GPLv3.

I can see this, but will a company see this?

> > "Pushing them away" means that they'd not do that because they would
> > be concerned that the license will change under them in such a way
> > that even those private modifications need to be released to the
> > public.
>
> This would not only change the spirit of the license, but turn it into
> a non-Free Software license.

Point. But once again - would a company pay attention to that fact?

> And then, again, the license can't possibly be changed from under
> them.  A new revision of the GPL would only affect software licensed
> under that new revision.  If you already got it under an earlier
> revision, you know what you got, and nobody can take that away from
> you.

True. But that doesn't save them from lawsuits trying to force them to obey 
the terms of the new revision even though they received the software under an 
earlier version.

> > (and don't try to argue that even though those modifications are
> > truly private (to the company) they should be released anyway to
> > comply with the "spirit" of the license. It is made clear that it
> > isn't by the text of the license itself)
>
> How could you possibly come to the conclusion that forcing anyone to
> release private modifications would be in compliance with the spirit
> of the license?  can != must

I was trying to be sarcastic and inject a little humor here. Guess I should 
have used the old <sarcasm> tag :)

> >> > Why should I repeat Linus' explanation of the ways that GPLv3 violates
> >> > the spirit of GPLv2?
> >>
> >> Don't worry about parrotting here, he hasn't provided that explanation
> >> yet ;-)  Please give it a try.
> >
> > But he has. Whether you have accepted that his explanations are
> > valid or not doesn't change the fact.
>
> His explanation is based on a reading of the license that doesn't
> match what its authors meant.  I guess the authors know better what
> they meant the spirit of the license to be than someone else who
> studied it a lot but that until very recently couldn't even tell the
> spirit from the legal terms.

And his interpretation is no less valid than that of anyone else. In fact, 
after a recent conversation with a couple of lawyers that I know, I can state 
that his interpretation isn't that far off from theirs.

> >> > Just like people have started using "GNU/Linux" or "GNU+Linux" to
> >> > refer to Linux
> >>
> >> No, no, you got it wrong.  Linux is the kernel.  GNU was the
> >> nearly-complete operating system it fit in.  GNU+Linux is a complete
> >> operating system.
> >
> > *AND* you cut out the bit where I said "I have no problems with it"
>
> Referring to Linux as GNU/Linux would be wrong, because Linux is the
> kernel, and that's unrelated with the GNU operating system.  It's the
> combination of them that forms GNU+Linux.  And it's referring to this
> combination as Linux that is wrong.
>
> I'm sorry that I got the impression that you meant the combination
> when you wrote "refer to Linux" above.  It looked like you meant the
> combination, since I've never seen anyone call the kernel GNU/Linux or
> GNU+Linux.

Then you're lucky. I've had a lot of people say something similar to the 
following: "Oh, I've heard about that. So which version of the GNU-Linux 
kernel are you running?" 

> > Never claimed otherwise. The problem is that using a composite name like
> > that *does* confuse a hell of a lot of people.
>
> Pronouncing the '+' or '/' helps a lot.  GNU plus Linux makes a lot of
> sense, and so does GNU on Linux.

Yes, it does. While pronouncing the '/' or '+' sounds a bit odd it does get 
the point across that it's the GNU userspace running on top of the Linux 
kernel. (as does "GNU on Linux")

> >> > A "TiVO" is not, and has never been, a "General Purpose
> >> > Computational Device".
> >>
> >> Err...  Last I looked it was a bunch of general-purpose components,
> >> packaged in a way that made it not look like a general-purpose
> >> computer.  Who gets to decide?  And with what motivations?
> >
> > And so is every game console. But until the original XBox was released
> > nobody tried using one as a "General Purpose" machine. The TiVO wasn't
> > designed as a general purpose machine - it was designed for a specific
> > purpose. That the *easiest* design to produce uses a bunch of general
> > purpose components is an economic choice, nothing else.
>
> So, if I put together a general-purpose computer, a general-purpose
> operating system, adding a label "not a general-purpose computer" is
> enough to make it so, just so that I can escape the obligations to
> respect users' freedoms?
>
> > I will not, however, argue about this anymore.
>
> Fair enough.
>
> > Let me quote Linus here:
> >
> > But I think the whole thing is totally misguided, because the fact is,
> > the GPLv2 doesn't talk about "in place" or "on the same hardware".
> >
> > In other words, GPLv3 is breaking with its predecessor - it's adding a
> > requirement that doesn't exist in previous versions.
>
> No dispute about this.  The requirements are being added to the legal
> terms, precisely such that they better reflect the spirit, under the
> light of the new threats that appeared since GPLv2 was published.
>
> But the new requirements do abide by the same spirit, and that was a
> promise the FSF made WRT revisions of the GPL.

As I've stated before - I can find nothing in the history of the GPL or the 
FSF that makes the "on the same hardware" requirement clear and part of 
the "spirit" of "Free Software". The closest anything comes is the "printer 
driver" that was the (in)famous "last straw" for RMS and caused him to create 
the FSF and the GNU Project.

> > *AND* it's dictating terms for *HARDWARE* when it isn't a hardware
> > license.
>
> Only in as much as you try to use the hardware as means to disrespect
> the spirit of the license and escape from the obligation to respect
> users' freedoms.

Shouldn't matter. As I've repeated quite a bit I cannot find a single mention 
that "on the same hardware" has been a goal of the FSF or part of 
the "spirit" of the GPL at any point before the drafting of GPLv3.

> > If I release software under the GPL and somebody modifies it to run
> > on a different hardware platform I'll be happy, even if they don't
> > send me a patchset for the new version.
>
> Yup.  See the bit about GPL not being tit-for-tat.
>
> > If I create a piece of hardware and run Linux on it, but have it
> > locked to a specific version or versions from a specific source (ie:
> > me) and release it to the public, I *WILL* release the version of
> > Linux I'm running on it.
>
> Good.
>
> > What I won't do is release whatever tools and such that are needed to
> > make the hardware run a different version of the kernel. Why? Because:
> > the hardware was designed so that a specific version of the kernel runs
> > without problems, there is hardware that is very picky and running a
> > customized kernel could cause that hardware to fail, etc...
>
> Why do you care?  It's no longer your hardware, it's theirs.

Legal requirements in some countries that require manufacturers to provide 
support for their product for a period of time after it has been purchased.

> Why would you refrain from providing information to others such that
> they *could* make the software do what *they* want in their hardware
> that they got from you?

See above. 

> If you let them change it and they break it, they get to keep all the
> pieces.  Your job is done.  Why get out of your way to stop them from
> making the best out of *their* hardware?

*thinks* Okay - looks like I had a thinko there. Arguments on this subject 
(ie: the theoretical hardware) are withdrawn and the point is conceded.

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 2096 bytes --]

On Thursday 14 June 2007 01:49, Alexandre Oliva wrote:
> Oh, good, let's take this one.
>
>   if you distribute copies of such a program, [...]
>   you must give the recipients all the rights that you have
>
> So, TiVo includes a copy of Linux in its DVR.
>
> TiVo retains the right to modify that copy of Linux as it sees fit.
>
> It doesn't give the recipients the same right.
>
> Oops.
>
> Sounds like a violation of the spirit to me.
>
> Sounds like plugging this hole would retain the same spirit.

Note that Harald Welte has already managed to force Siemens to unlock 
a "tivoized" Linux router with the GPLv2 in Germany. German contract law 
values intention when the contract has no specific clause that deals with 
the issue, and in German law, an accepted license is a contract.

So the fact that tivoizing Linux is against the spirit of the GPLv2 is a 
court-proof fact, not just some speculation.

What about if your GPL program ends up in a piece of hardware (e.g. a ROM, 
or an embedded ROM, or if it's some GPL code from OpenCores, as gate 
netlist in silicon)? My interpretation is that you need a permission from 
the author for doing that, unless there's an easy way to replace it with a 
modified copy (e.g. if you put the OpenCores stuff into an FPGA, replacing 
the configuration PROM would do it).

Some people have difficulties with intentions of contracts rather than 
direct rules. That may be due to different rules in different countries. In 
continental Europe, contract law usually bases on Code Napoleon, and 
there, "good faith" is an important principle (and "good faith" means that 
the intention is more important than the actual coded practices). In the 
roman law that was used before and has survived in countries who didn't let 
Napoleon in (like the UK and the USA), it's slightly different. But a 
contract or a license still is not a program where anything that isn't said 
explicitely isn't said at all.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Petrovitsch]

On Wed, 2007-06-13 at 23:38 -0300, Alexandre Oliva wrote:
> On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:
> 
> > Exactly. They don't. What TiVO prevents is using that modified version on 
> > their hardware. And they have that right, because the Hardware *ISN'T* 
       ^^^^^^^^^^^^^^
BTW as soon as I bought that thing, it is *my* hardware and no longer
*theirs* (whoever "theirs" was).

> > covered by the GPL.
> 
> Indeed, TiVO has this legal right.  But then they must not use

Do they? At least in .at, it is usually impossible to (legally) limit
the rights of the *owner* a (tangible) thing (and if I bought it, I *am*
the owner and no one else) - even if you put it in the sales contract
since this is discussion about/within sales law.

One usual example is "you buy a car and neither the car producer nor the
(re)seller can restrict the brands of the tires you may use or the brand
of the fuel etc.".

And the same holds for pretty much everything. No one can forbid you to
open a TV set and fix it (or let it fix by whoever I choose to).

Yes, there are exceptions in several laws for specific things (e.g. for
really dangerous ones like airbags in cars) but in general, you are
allowed to do almost anything (including the simple destruction of it).

And yes, if you *rent* the thing, you are not the owner and this is a
totally different thing.

> software under the GPLv3 in it.  And, arguably, they must not use
> software under the GPLv2 either.

	Bernd
-- 
Firmix Software GmbH                   http://www.firmix.at/
mobil: +43 664 4416156                 fax: +43 1 7890849-55
          Embedded Linux Development and Services

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[jimmy bahuleyan]

Theodore Tso wrote:
> Basically, in the US, you get the best justice money can buy.  :-)

that has to be one of the best one-liners ever! :)

> 
> 							- Ted

-jb
-- 
Tact is the art of making a point without making an enemy.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Gerdau]

[-- Attachment #1: Type: text/plain, Size: 2170 bytes --]

> > > able to run you modifications on the same hardware?
>                                          ^^^^
> > Come on! The whole idea of software is to have it run on some HW.
>                                                            ^^^^
> > Why would I want to change it in the first place if I can't run it ?
> 
> See the difference?

Forgive my poor mastery of the english language and me letting slip
this inconsistency.

The first sentence you cited was a general remark IMO valid outside
of this context and possibly ill placed as it was.

The second sentence pertains the key msg I was trying to deliver and
apparently I did a poor job in phrasing it so let me redo it:
Why would I want to change the SW targetted for some HW if I can't run
the changed version on said HW ?

[note that for the TiVo case I possibly would not own or be able to own
similar HW being able to run my modified SW; so even some HW would not
be triggered either]                                 ^^^^

Remember I'm discussing my understanding of the spirit of the GPL,
not whether the legal part actually does give me that right enforceable
in court.

Here is another stmt which is valid outside of this context AFAIAC:
If the GPLv2 does not legally give me the right that I think its spirit
gives me then the legal phrases should be changed to achieve that.

Whether or not others share my view of what the spirit of the GPL
implies is completely theirs to decide and if they differ they likely
won't agree on my previous stmt either. Fine with me.

And this leads to another observation:
IMO this thread is partly fueled by a fundamental mixing of PoVs.
Some argue based on their perceived view of the spirit of the GPL
and some based on the actual legal phrases in GPLv2 and GPLv3 and
whether or how they reflect the perceived spirit.

Best wishes,
Michael
-- 
 Technosis GmbH, Geschäftsführer: Michael Gerdau, Tobias Dittmar
 Sitz Hamburg; HRB 89145 Amtsgericht Hamburg
 Vote against SPAM - see http://www.politik-digital.de/spam/
 Michael Gerdau       email: mgd@technosis.de
 GPG-keys available on request or at public keyserver

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 04:37:55 Bernd Petrovitsch wrote:
> On Wed, 2007-06-13 at 23:38 -0300, Alexandre Oliva wrote:
> > On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > > On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:
> > >
> > > Exactly. They don't. What TiVO prevents is using that modified version
> > > on their hardware. And they have that right, because the Hardware
> > > *ISN'T*
>
>        ^^^^^^^^^^^^^^
> BTW as soon as I bought that thing, it is *my* hardware and no longer
> *theirs* (whoever "theirs" was).

eh. Perhaps I should have said that differently. And TiVO could handle it 
differently. I'm not going to argue about it anymore. It's pointless.

> > > covered by the GPL.
> >
> > Indeed, TiVO has this legal right.  But then they must not use
>
> Do they? At least in .at, it is usually impossible to (legally) limit
> the rights of the *owner* a (tangible) thing (and if I bought it, I *am*
> the owner and no one else) - even if you put it in the sales contract
> since this is discussion about/within sales law.
>
> One usual example is "you buy a car and neither the car producer nor the
> (re)seller can restrict the brands of the tires you may use or the brand
> of the fuel etc.".

No argument there. However, that is not to say that "you bought it, now you're 
free to do with it whatever you please" is always what the law says (at least 
in the US)

In the TiVO case there may be restrictions placed on the manufacturer for 
legal reasons or contractual reasons. Seeing as I'm not privy to the 
contracts between TiVO and the various production and broadcasting companies 
I can't comment on what contracts they have. As to the legal side there are 
restrictions in copyright law.

> And the same holds for pretty much everything. No one can forbid you to
> open a TV set and fix it (or let it fix by whoever I choose to).

I know of at least one company that will sell you the parts to repair your TV 
if its out of warranty.

DRH

> Yes, there are exceptions in several laws for specific things (e.g. for
> really dangerous ones like airbags in cars) but in general, you are
> allowed to do almost anything (including the simple destruction of it).
>
> And yes, if you *rent* the thing, you are not the owner and this is a
> totally different thing.
>
> > software under the GPLv3 in it.  And, arguably, they must not use
> > software under the GPLv2 either.
>
> 	Bernd



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 2609 bytes --]

On Thursday 14 June 2007 09:32, Paul Mundt wrote:
> This is perhaps the part that's the most interesting. For the very small
> number of people that _do_ want to change these things (usually at the
> expense of a voided warranty, in the consumer device case), there's
> always a way to make these changes, even if you must resort to hardware
> hacking. Trying to mandate this sort of functionality in the license
> might make it easier for a few people to get their code loaded, but the
> vast majority of users have zero interest in anything like this.

I don't feel this is a very conclusive argument.

How many computer users do want to change their OS? I mean not only want to 
change the OS in the sense of "apply patches released by Microsoft", but on 
their own? Many typical computer users ask for help to "fix their computer" 
when turning it off and on again already "fixes it". They would never ever 
change the source code of their OS even if they technically could do it - 
they are not programmers.

However, if there is the technical possibility to change the firmware of an 
appliance, somebody does it, and often mere users upload these changes to 
their own device (like the OpenWRT stuff).

Let me give one example: My parents own a DVB-T DVR. It was a cheap one, and 
it was cheap because the software is lackluster. Unfortunately it isn't 
free. Many users of this device complain to the manufacturer about the 
stability and quality of the software, but with no avail - there haven't 
been any updates in the last two years. I suppose I would be able to fix 
the problem, most other users probably wouldn't (and my parents neither). 
But if I did fix the problem, and provided them with an updated firmware, 
they would install it on their device.

That's the "help your neighbour" right in the GNU manifesto. It's as 
important as the "help yourself" right, maybe even more. It was the 
original motivation of RMS to make free software - the frustration of not 
being able to help his neighbours. He had an NDA to help himself.

What people want is software that works. If the firmware of your microwave 
or DVR works, you don't care so much if it is free or not. You only care if 
it doesn't work, and you feel the urge to fix it (and turning it off and on 
again doesn't fix it). That's why people complain loud about ATI drivers 
not being open, and don't care that much about the Nvidia driver, which is 
just as closed, but works.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Krzysztof Halasa]

Daniel Hazelton <dhazelton@enter.net> writes:

> Exactly. And I don't see anything about a TiVO (or any device that, like a 
> TiVO, requires binaries that run on it to be digitally signed) that stops
> you 
> from exercising the "freedoms" guaranteed by the GPL. As I said before, what 
> it does is stop you from violating the license on the hardware.

BTW: don't they sell their hardware (as well)? I think it should be
easy to replace the ROMs (EPROMs? flash ROMs?) using some diagnostic
clip and/or JTAG. Unless the CPU itself verifies ROM signatures,
they shouldn't matter.
-- 
Krzysztof Halasa

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 762 bytes --]

On Thursday 14 June 2007 03:24, Adrian Bunk wrote:
> Harald is in Germany, and he therefore takes legal action against people
> distributing products violating his copyright on the Linux kernel
> in Germany at German courts based on German laws.

And if Tivo did sell their crap in Germany, I bet, Harald had brought them 
down years ago (as he did in the "tivoized" Siemens router case). But Tivo 
doesn't (they started in the UK, and stopped doing so right after Harald 
unlocked that Siemens router ;-), and in the US, courts may think 
different. Or they rely that there simply is no Harald Welte in the US, who 
goes after the violators.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Krzysztof Halasa]

Alexandre Oliva <aoliva@redhat.com> writes:

> BTW, what license is Linux licensed under?  It's GPLv2 plus userland
> exception, right?  (There's some additional module exception, right?)

Pure GPLv2.

Userland exception? Never heard of.
Module exception? Perhaps you mean "interpretation"?
-- 
Krzysztof Halasa

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Schmidt]

Alexandre Oliva wrote:
> On Jun 14, 2007, Bron Gondwana <brong@fastmail.fm> wrote:
> 
>> Tivo gets sick of the endless flamewars on lkml, signs a copy
>> of QNX, pushes it out to the hardware.  No more Linux on Tivo.
> 
> What do we lose?
> 
> Do we actually get any benefit whatsoever from TiVO's choice of Linux
> as the kernel for its device?

Do they contribute back any code that makes Linux better?
If Tivo doesn't, what about other vendors who may be in a similar situation?


Bernd

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Petrovitsch]

On Thu, 2007-06-14 at 05:05 -0400, Daniel Hazelton wrote:
> On Thursday 14 June 2007 04:37:55 Bernd Petrovitsch wrote:
[...]
> > > > covered by the GPL.
> > >
> > > Indeed, TiVO has this legal right.  But then they must not use
> >
> > Do they? At least in .at, it is usually impossible to (legally) limit
> > the rights of the *owner* a (tangible) thing (and if I bought it, I *am*
> > the owner and no one else) - even if you put it in the sales contract
> > since this is discussion about/within sales law.
> >
> > One usual example is "you buy a car and neither the car producer nor the
> > (re)seller can restrict the brands of the tires you may use or the brand
> > of the fuel etc.".
> 
> No argument there. However, that is not to say that "you bought it, now you're 
> free to do with it whatever you please" is always what the law says (at least 
> in the US)

Of course not (and I neither stated nor implied it) - there are lots of
laws which forbid killing other people etc.
But the seller of the car is not in the position to forbid anything
(which is not forbidden by the law), e.g. ha cannot forbid to replace
the motor or similar thing. I may loose guarantee or have to cope with
other consequences (if it is done badly),  but that is my problem and
decision.

> In the TiVO case there may be restrictions placed on the manufacturer for 
> legal reasons or contractual reasons. Seeing as I'm not privy to the 

Frankly, I really don't care that much about legal and contractual
reasons of the *manufacturer* (starting from waste disposal regulations
up to tax regulations, etc.) and they are irrelevant to me anyways.
At most I can have
*) legal restrictions (obviously coming from the law) on the *usage* of
     the device or
*) from a contract (obviously with the seller of the device since there
    is no other involved - and this contract may contain inapplicable
    clauses - e.g. sth. like "you are not allowed hear German music with
    this device").
And I don't have a contract with the manufacturer so there can't be any
limitation by the manufacturer.

> contracts between TiVO and the various production and broadcasting companies 
> I can't comment on what contracts they have. As to the legal side there are 

And they are pretty irrelevant anyways to everyone else.

> restrictions in copyright law.

ACK. But copyright law (at least the equivalent in .at and very
probably .de - and IMHO it is probably everywhere else similar simply
because copyright/authors rights laws was actually designed and written
to deal with music, literature, etc. which are intangible by nature)
simply doesn't apply to hardware as such (pun intended;-).

	Bernd
-- 
Firmix Software GmbH                   http://www.firmix.at/
mobil: +43 664 4416156                 fax: +43 1 7890849-55
          Embedded Linux Development and Services

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> I've been following this discussion and I find this interesting.
> Consider these two cases:
> 
> 1.) I ship the device back to the manufacturer, they replace the ROM,
>     and ship it back to me.
> 
> 2.) I ship the device back to the manufacturer, they load new code
>     into it, and ship it back to me.
> 
> How do these two differ?  Or is it now just a question of the ROM

Thats one thing I don't like about the GPL3 special casing.

> being in a socket?  I can't see how the technicalities of how the
> hardware is constructed can change the legality of the software.

In the replace/reflash the ROM case its about access to the righ tools -
I could do it myself, send it to another company to load my code etc.

In the Tivo case its about one company having the ability to make such
mods and blocking others from doing so.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> the new language it contains. It has taken almost 15 years for "Free 
> Software" to make a dent in the market, and, IMHO, a lot of that is both 
> Linux and the "holes" in GPLv2.


You appear terminally confused. The purpose of the GPL as defined by its
authors is not commercial success, world domination or making zillions of
dollars - it is keeping the software protected by that license "free" in
terms of liberty as measured against the set of freedoms to
run/modify/etc they discuss in the licence document.

The fact this is a good license for making zillions of dollars, producing
good software and the like is either incidental or a logical result of the
protection of freedoms depending upon which views you believe.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:

> > the new language it contains. It has taken almost 15 years for "Free 
> > Software" to make a dent in the market, and, IMHO, a lot of that is both 
> > Linux and the "holes" in GPLv2.
> 
> You appear terminally confused. The purpose of the GPL as defined by 
> its authors is not commercial success, world domination or making 
> zillions of dollars - it is keeping the software protected by that 
> license "free" in terms of liberty as measured against the set of 
> freedoms to run/modify/etc they discuss in the licence document.
> 
> The fact this is a good license for making zillions of dollars, 
> producing good software and the like is either incidental or a logical 
> result of the protection of freedoms depending upon which views you 
> believe.

that's fine, but the fundamental question is: where is the moral 
boundary of the power that the copyright license gives? The FSF seems to 
believe "nowhere, anything that copyright law allows us to achieve our 
goals is a fair game" - and the GPLv3 shows that belief. I dont 
subscribe to that view. I think the proper limit is the boundary where 
the limit of the software is - because that's the only sane and globally 
workable way to stop the power-hungry. I.e. the information we produce 
is covered by the rules of the GPL. It might be used in ways 
inconvenient to us, it might be put on hardware we dont like (be that a 
Tivo, a landmine or an abortion instrument) but that does not change the 
fundamental fact: it's outside the _moral scope_ of our power. Whether 
some jurisdictions allow the control of _other_ information via our 
information is immaterial. If a jurisdiction allows the control of 
hardware that is associated with our software, so what? If a 
jurisdiction allows the controlling of various aspects of movie theaters 
that happen to play copyrighted movies, does it make it morally right?

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> that's fine, but the fundamental question is: where is the moral 
> boundary of the power that the copyright license gives? The FSF seems to 

Assuming a democratic state then the laws of the land ought to reflect the
'general will' (if you believe Rousseau anyway). They should thus define
the boundary ['derivative work' generally ] according to the general good
and with the consent of the people.

> hardware that is associated with our software, so what? If a 
> jurisdiction allows the controlling of various aspects of movie theaters 
> that happen to play copyrighted movies, does it make it morally right?

Does that question not suppose some positivist absolute morality ? I
suspect many would argue that it is moral to do so if the end goal of the
controls is moral. You might also want to apply the tests in Fuller's
Internal Morality of Law ?

I'm not sure Philosophy is on topic for l/k however 8)

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Poole]

Matt Keenan writes:

> Alexandre Oliva wrote:
>>
>> Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
>> require some physical support.  That's the hard disk in the TiVO DVR,
>> in this case.  I don't see how this matters, though.
>>
>>   
> I'm now intrigued, where are these (Brazilian and US) definitions
> stipulated, and under what authority?

In the US, 17 USC 101 (the "Definitions" section of the title dealing
with Copyright) makes this definition:

    A "computer program" is a set of statements or instructions to be
    used directly or indirectly in a computer in order to bring about
    a certain result.

As its purpose is to outline the scope of copyright law, this
definition is made under the authority granted to Congress by Article
I, Section 8 of the United States Constitution.

Michael Poole

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 2474 bytes --]

On Thursday 14 June 2007 12:38, Ingo Molnar wrote:
> that's fine, but the fundamental question is: where is the moral
> boundary of the power that the copyright license gives? The FSF seems to
> believe "nowhere, anything that copyright law allows us to achieve our
> goals is a fair game" - and the GPLv3 shows that belief. I dont
> subscribe to that view. I think the proper limit is the boundary where
> the limit of the software is - because that's the only sane and globally
> workable way to stop the power-hungry. I.e. the information we produce
> is covered by the rules of the GPL. It might be used in ways
> inconvenient to us, it might be put on hardware we dont like (be that a
> Tivo, a landmine or an abortion instrument) but that does not change the
> fundamental fact: it's outside the _moral scope_ of our power.

Where is the boundary between hard- and software? I'm employed as hardware 
designer, and for this purpose, I write programs in a hardware description 
language, which can be converted into hardware through a synthesis 
software. I write firmware, which is assembled into binary and gets placed 
on on-chip memory (ROM or NVM). I've even studied computer science, and 
electric engineering was just a side-course. I know how transistors work, 
and how gates are implemented in terms of transistors, but in essense, it's 
not that relevant unless you want to do analog circuits. Usually, during 
the development phase, we put the Verilog into an FPGA, where the 
configuration file still is obviously "software" in any sense it can be. 
I've even released descriptions of some parts of the work I do under GPL 
for people to put it into their own FPGAs.

There is no boundary between hard- and software in the sense of that 
hardware is something fundamentally different. Hardware is just another way 
to implement programs, and it uses other languages (but SystemC even looks 
quite close to C). If there is a boundary, it's way below the distinction 
between a Tivo and a PC, because these two basically consist of a 
processor, some RAM, some flash, a harddisk, and a video driver.

What's true: We don't have the moral power to define *where* the software 
goes, but we have the moral power to define *how* users can change the 
software when they own the hardware (the physical representation).

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Cox]

> If the two courts are in the same country there's usually a higher court 
> above both that can resolve this. But what if let's say the highest 
> court in the USA and the highest court in Germany would disagree on such 
> a matter?

It would be very unusual for both cases to proceed in parallel. The one
court will then strongly consider the decision of the other.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:

> > that's fine, but the fundamental question is: where is the moral 
> > boundary of the power that the copyright license gives? The FSF 
> > seems to
> 
> Assuming a democratic state then the laws of the land ought to reflect 
> the 'general will' (if you believe Rousseau anyway). They should thus 
> define the boundary ['derivative work' generally ] according to the 
> general good and with the consent of the people.

uhm, so if the MPAA and the RIAA pays for another nice piece of 
legislation that extends the power of copyright owners, do you find it 
morally justified to use those powers, as long as it's argued to be in 
favor of some long-term goal that you judge to be moral, even if it 
results in some "temporary injustice"?

i think that could be the main difference in thinking. I argue that the 
only way to be moral is to be moral _now_, not "later, once this very 
important fight for the common good is over". I think the moral approach 
to this is to say _no_ to attempts to extend the license to beyond the 
"moral scope" of the software we wrote - regardless of what new powers 
are legislated into the hands of copyright owners. It's naturally hard 
to do, because giving up power is always hard to do.

In other words: we need to apply our concepts of freedom and fairness 
not only to the end result, but to the means and methods of achieving 
those end results as well. The end goals are often forgotten, it's the 
process that matters to the end result.

Or in yet another set of words: this concept of morality also happens to 
be expressed fairly accurately in the thousands of years of 'quid pro 
quo' concept. (shared amongst many, many cultures on this planet, shared 
amongst far more cultures than the western 'freedom' concept.) (which 
concept of quid-pro-quo fairness is likely coded into our brains and 
into our thinking genetically - because it's a simple and very efficient 
group survival method.)

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Bernd Paysan <bernd.paysan@gmx.de> wrote:

> On Thursday 14 June 2007 12:38, Ingo Molnar wrote:
> > that's fine, but the fundamental question is: where is the moral
> > boundary of the power that the copyright license gives? The FSF seems to
> > believe "nowhere, anything that copyright law allows us to achieve our
> > goals is a fair game" - and the GPLv3 shows that belief. I dont
> > subscribe to that view. I think the proper limit is the boundary where
> > the limit of the software is - because that's the only sane and globally
> > workable way to stop the power-hungry. I.e. the information we produce
> > is covered by the rules of the GPL. It might be used in ways
> > inconvenient to us, it might be put on hardware we dont like (be that a
> > Tivo, a landmine or an abortion instrument) but that does not change the
> > fundamental fact: it's outside the _moral scope_ of our power.
> 
> Where is the boundary between hard- and software? [...]

this is largely irrelevant to my argument: the FSF is clearly trying to 
extend the scope of the GPL to restrict the distribution of certain 
hardware+software combinations. The FSF is not really arguing that the 
boundary between software and hardware is diffuse. (which btw. it 
clearly is) The FSF simply wants to be able to say via the GPLv3: "to be 
able to distribute GPL-ed software, the hardware is required to do this 
and this".

please note an important thing here: "required to do this and this" is 
the _precise antithesis_ of "freedom". The only significant restriction 
on freedom the GPLv2 allows is that the covered work (the software) is 
not to be restricted. And that is a fair deal. Even if any additional 
restrictions would otherwise be for the "common good" and would further 
"freedom" in creative ways.

If the FSF's argument and approach was correct then it would be fine to 
add these restrictions to GPLv4:

 - do not distribute non-GPL-compatible software with GPL software on 
   the same hardware.

 - send at least 10 free samples of the hardware to the FSF 
   headquarters. (after all true freedom is only achieved if developers 
   are not only allowed to modify the hardware, but are allowed to test 
   it as well, for the freedom of the community.)

 - donate $10 to the FSF.

 - spray "Linus sucks because he stole RMS's GNU thunder in the 90s and
   never gave it back!" graffiti on 3 separate walls in your
   neighborhood.

Each of these items is an additional restriction on either the 
hardware+software combination that is being distributed or on the person 
who does the distribution, and each of these items - some abstractly, 
some more directly - advance a notion of the "four GNU freedoms" in some 
way. And each of these items has a basis in copyright law and might be 
legally put into a license and might be enforceable. (ok, probably not 
the last item ;)

think about it, the list of things that one can do via license to 
"achieve more freedom" just doesnt stop! My point is: it has to stop at 
the only boundary that makes sense, and which boundary is clearly 
spelled out in the spirit and in the letter of the GPLv2: "our work is 
our work, your work is your work".

Any additional restriction to "help achieve more freedom" just puts us 
into divisive political and moralistic games that will only fracture us, 
that will eventually erode the value of our software and hence makes any 
'power' we have over that software meaningless in the real world. In the 
end no-one but the Microsofts of this world will win.

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alan Milnes]

On 14/06/07, Ingo Molnar <mingo@elte.hu> wrote:

>    My point is: it has to stop at the only boundary that makes sense, and which boundary
> is clearly spelled out in the spirit and in the letter of the GPLv2: "our work is our work, your
> work is your work".

Agreed - if you want to take my work you are welcome as long as you
contribute back your changes.  That's the deal that GPL2 enforces and
why it has been so successful. GPL3 is a very different beast with a
much wider agenda, which makes it far more difficult to achieve
consensus on what it should contain.

Personally I would have liked to seen a GPL2.x which fixes some of the
issues but stays true to the more limited objective.

Alan

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Robin Getz]

On Thu 14 Jun 2007 01:07, Alexandre Oliva pondered:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> 
> > User B buys the router and modifies the kernel so it drives the WiFi
> to an 
> > output power twice that which it is licensed to carry.
> > FCC finds out and prosecutes User B for violating the regulations.
> 
> Ok so far.
> 
> > FCC then pulls the small companies license until they change their
> > hardware so the driver can't push it to transmit at a higher power
> > level and levies a fine.
> 
> I'd say this is unfair, but if it can happen, 

it can (at least hypothetically in the US - I do not know of any actual 
cases - anyone?).

> then maybe the small 
> company could have been more careful about the regulations.  There are
> various ways to prevent these changes that don't involve imposing
> restrictions of modification on any software in the device, all the
> way from hardware-constrained output power to hardware-verified
> authorized configuration parameters.

As a person pretty familiar with the hardware in these types of devices - this 
just isn't practical. 

The power output goes to the air, has some to do with any external power amp, 
the antenna design/connectors, the layout of the printed circuit board (PCB), 
the materials of the PCB, and numerous other factors that are outside the 
chip where the register sits that controls the "output power". This is why 
the register is there - so many things exist in the real world that require 
it to be changed to actually get desired power output in the air.

What you are asking is possible - that people who make the standard CMOS 
radios add EEPROM or FLASH to their processes - or they could add an 
interface to a serial EEPROM where these hardware limits could sit - and 
during manufacturing could program these to add a hardware limit.

But this seems - just silly - adding extra cost/complexity to the hardware - 
because you limit things in the zero-cost space?

Or run the driver in user space, and forget about it. (Like everyone is 
starting to do on the desktop).

> When this doesn't bring freedom to people, when people can't actually
> enjoy the freedoms that the software is supposed to provide, I don't
> see why this would be a good thing.  What's the merit in being able to
> claim "vendor X chose my Free Software and locked it down such that
> users don't get the freedoms I meant for them, and I'm happy about it?"

Sometimes joy comes from knowing that your contribution was valuable.

I would ask about two other examples:

 - medical devices - which must have their software certified by the FDA (at 
least in the US) - and the manufacture can not allow non-certified software 
to be loaded on it - and would be classified "designed or sold for 
incorporation into a dwelling" (per the GPL3 dd1v3). Think devices like :

http://www.powerheart.com/products/ [1]

Are Linux developers "happy" knowing their contributions will never help save 
a life? And this joy goes to Microsoft or VxWorks? Today it can go to Linux 
developers, since the GPL2 license doesn't restrict this like the GPL3 does.


 - gambling devices - which must have their software certified by various 
government agencies - to make sure that the odds are known, and there are no 
backdoors, and consumers don't get screwed - the manufacture can not allow 
non-certified software to be loaded on it. If these are in a hotel - where 
various people live - is that considered "incorporation into a dwelling"?

Not wanting to start a debate about the morality of being involved in the 
gambling industry - (if the statically challenged are giving the government 
money to keep my taxes down, I am mostly OK with it) - but I'm not "happy" 
thinking that someone can ledgistate restrictions on embedded OS choice, just 
because it must be verified by a third party.


Hypothetically - tragedy strikes - use of a modified Nao robot [2] contributes 
to someone's death (remember - modifications of the OS removed the three 
laws). Someone (maybe Microsoft or VxWorks, or GreenHills) steps up, and 
lobbies the government - that embedded OSes need tighter control to ensure 
safety - and need to be validated by the government. (like the FCC and FDA is 
already doing in the US).

Since the GPL3 states" - 
"If you cannot convey the Program, or other covered work, so as to satisfy 
simultaneously your obligations under this License and any other pertinent 
obligations, then as a consequence you may not convey it at all."

Forget DRM - forget Tivo - anyone who develops a product which must be 
certified can not use anything that is under the GPL3.

On Sun, 10 Jun 2007 10:29:04 Linus Torvalds pondered:
> I still think GPLv2 is simply the better license.
>
> I consider dual-licensing unlikely (and technically quite hard), but at 
> least _possible_ in theory. I have yet to see any actual *reasons* for 
> licensing under the GPLv3, though.

I think there are actual reasons _not_ to.

What I have told people who have asked me (who are thinking about putting 
Linux in a medical device) is kernel version 2.6.xx (current version at the 
time), is released under GPL2.  No one can go back and change the license on 
a previously released package. 

It might be easier to say that the kernel licence will only be reviewed during 
a major bump in the kernel version number (3.x.x) and that all 2.x kernels 
will be under the same license as what exists today.

-Robin

[1] - I don't know if this device has Linux in it or not - but this just 
represents classes of devices that may never be able to use Linux.

[2]  http://linuxdevices.com/news/NS6263763539.html

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Poole]

Ingo Molnar writes:

> this is largely irrelevant to my argument: the FSF is clearly trying to 
> extend the scope of the GPL to restrict the distribution of certain 
> hardware+software combinations. The FSF is not really arguing that the 
> boundary between software and hardware is diffuse. (which btw. it 
> clearly is) The FSF simply wants to be able to say via the GPLv3: "to be 
> able to distribute GPL-ed software, the hardware is required to do this 
> and this".

Most people arguing for the expansive interpretation do not really
care what hardware is combined with what software.  They care about
the ability for the user (in the GPLv2's terms, someone who receives
GPL'ed software) to have comparable ability to modify and
(re-)distribute the software as the software distributor does.  The
issue of GPLed software on DRMed hardware applies equally to digital
video recorders, where the hardware and software distributor are
usually the same, and video game consoles, where they are not.

There is no good reason to treat a "GPL-incompatible" hardware
platform (for example, incompatible due to restrictions on the keys to
generate digital signatures) differently than a "GPL-incompatible"
patent area.  If a software distributor cannot simultaneously comply
with the GPL and his other obligations, he should either not
distribute the software or be prepared to face the liability from
breaching his obligations.

Michael Poole

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Matt Keenan]

Michael Poole wrote:
> Matt Keenan writes:
>
>   
>> Alexandre Oliva wrote:
>>     
>>> Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
>>> require some physical support.  That's the hard disk in the TiVO DVR,
>>> in this case.  I don't see how this matters, though.
>>>
>>>   
>>>       
>> I'm now intrigued, where are these (Brazilian and US) definitions
>> stipulated, and under what authority?
>>     
>
> In the US, 17 USC 101 (the "Definitions" section of the title dealing
> with Copyright) makes this definition:
>
>     A "computer program" is a set of statements or instructions to be
>     used directly or indirectly in a computer in order to bring about
>     a certain result.
>
> As its purpose is to outline the scope of copyright law, this
> definition is made under the authority granted to Congress by Article
> I, Section 8 of the United States Constitution.
>
>   
But where is the part that says it "requires some physical support"? It
says what it is; "a set of statements or instructions", how it should be
used; "to be used directly or indirectly in a computer", and what
purpose it serves; "in order to bring about a certain result", but it
doesn't seem to indicate that it "requires physical support" aka needing
some physical representation. I suspect this argument boils down to the
philosophical debate of whether ideas (in this case software) can be
truely devoid of the physical.

Matt

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Bernd Petrovitsch wrote:

> On Wed, 2007-06-13 at 23:38 -0300, Alexandre Oliva wrote:
> > On Jun 13, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > > On Wednesday 13 June 2007 19:49:23 Alexandre Oliva wrote:
> > 
> > > Exactly. They don't. What TiVO prevents is using that modified version on 
> > > their hardware. And they have that right, because the Hardware *ISN'T* 
>        ^^^^^^^^^^^^^^
> BTW as soon as I bought that thing, it is *my* hardware and no longer
> *theirs* (whoever "theirs" was).

You bought *their* design. It was your choice. 

And yes, you own the hardware, and you can hack it any which way you like 
(modulo laws and any other contracts you signed when you bought it). But 
they had the right to design it certain ways, and part of that design may 
be making it _harder_ for you to hack.

For example, they may have used glue to put the thing together rather than 
standard phillips screws. Or poured resin over some of the chips. All of 
which has been done (not necessarily with Linux, but this really is an 
issue that has nothing to do with Linux per se). Making the firmware or 
hardware harder to access or modify is their choice.

Your choice is whether you buy it, despite the fact that you know it's not 
necessarily all that easy to hack.

> > Indeed, TiVO has this legal right.  But then they must not use
> 
> Do they? At least in .at, it is usually impossible to (legally) limit
> the rights of the *owner* a (tangible) thing (and if I bought it, I *am*
> the owner and no one else) - even if you put it in the sales contract
> since this is discussion about/within sales law.

The "when I buy it, I own it" argument is a favourite of the GPLv3 shills, 
but it's irrelevant. The *design* was done long before you bought it, and 
yes, Tivo had the right to design and build it, any which way they wanted 
to. 

> One usual example is "you buy a car and neither the car producer nor the
> (re)seller can restrict the brands of the tires you may use or the brand
> of the fuel etc.".
> 
> And the same holds for pretty much everything. No one can forbid you to
> open a TV set and fix it (or let it fix by whoever I choose to).

You are missing the picture. Sure, you can do whatever you want to (within 
any applicable laws) _after_ you bought it. But that doesn't take away the 
right from the manufacturer to design it his way.

And you're also *wrong*. Tivo doesn't limit the brands of electricity it 
uses or anything idiotic like that. You can put after-market rubber bumps 
on the thing to make it look sleeker, and I seriously doubt that Tivo will 
do aythign at all. It's about going into the innards, and different car 
manufacturers make that harder too, for various reasons.

If the car manufacturer makes things harder to hack, it's your choice. For 
example, car hackers *do* actually prefer certain brands. Apparently the 
Subaru's are popular, and German cars are a pain to try to change. I'm 
told that even somethign as simple as upgrading the sound system is just 
_harder_ in a German car, apparently because they make things fit together 
so tightly, that doing after-market cabling is just much more of a 
problem.

Same goes for things like electronic engine controls. Look it up. Try 
chipping a car lately? For some, it's literally buying a chip online, and 
some fairly simple work. For others, it's almost impossible, and you have 
to take your car in to somebody who really knows what he's doing. And you 
know what? Exactly like with a Tivo, the car manufacturer won't have 
anything to do with the car afterwards. If you broke it by chipping it, 
you voided your warranty.

See? If you are actually looking for a car to hack on, you'd buy a car 
with that in mind. Do the exact same thing with your Tivo. Don't buy it if 
you want to hack it: buy a Neuros OSD device instead!  I'm serious: the 
Neuros people do *not* limit you, and in fact they encourage hacking. 
Instead of whining about Tivo, do something *positive*, and support Neuros 
for their better policies!

				Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Thu, Jun 14, 2007 at 12:00:17AM -0400, Valdis.Kletnieks@vt.edu wrote:
> On Thu, 14 Jun 2007 04:56:40 +0200, Adrian Bunk said:
> 
> > Reality check:
> > 
> > Harald convinced companies that they have to provide the private keys 
> > required to run the Linux kernel they ship on their hardware.
> 
> No, the *real* reality check:
> 
> The operative words here are "convinced companies" - as opposed to "convinced
> a judge to rule that private keys are required to be disclosed". (I just
> checked around on gpl-violations.org, and I don't see any news items that say
> they actually generated citable case law on the topic of keys...)
> 
> Harald convinced companies that it was easier/cheaper/faster to provide the
> private keys than to continue in a long legal battle with an uncertain outcome.
> If the company estimates the total loss due to keys being released is US$100K,
> but the costs of taking it to court are estimated at US$200K, it's obviously
> a win (lesser loss, actually) for the company to just fold.
>...

Here in Germany, the rules at court are roughly "the loser pays 
everything including the costs of the winner", so if a big company is 
sure they will win at court there's no reason not to go there.

And if they did the effort of using private keys to only allow running 
an official firmware, they must have seen an advantage from doing so.

I'm not saying it legally clear the other way round, my statement was 
an answer to Daniel's emails claiming it was clear what such companies 
do was legal.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> People don't get your copy, so they're not entitled to anything about
> it.
> 
> When they download the software, they get another copy, and they have
> a right to modify that copy.

Umm. I notice how you must have known how *idiotic* your response was, 
because you snipped away the part where I talked about Red Haty 
distributing CD-ROM's.

In other words, Red Hat distributes copies (and yes, you *get* that copy), 
and you cannot modify that copy that you got.

So the "right to make changes" _must_ be separate from the actual copy of 
the image.

And don't get fooled by the "all the rights that you have". That 
_obviously_ and clearly talks about "the program", which in turn 
equally obviously and clearly has to be about something bigger than the 
"one copy", since the GPLv2 requires you have the right to change it.

So you edited out the part where I talked about CD's. That's the proof 
that your reading is untenable, because obviously you cannot change the 
program on the CD: you got a copy, but the right to make modifications 
wasn't ON THAT HARDWARE.

> > And here's a big clue for people: anybody who thinks that I'm violating 
> > the GPLv2 by not giving out my private SSH key to master.kernel.org is a 
> > f*cking moron!
> 
> Agreed, except I'd probably use a lighter term.

Hey, I'm not exactly known for being polite. I tell it how I see it, and I 
tend to be pretty damn blunt about  it. 

> > See any parallels here? Any parallel to a CD-ROM distribution, or a Tivo 
> > distribution?
> 
> Yes.  You see how TiVO is different?  It is modifyable, and I actually
> receive the copy that TiVO can still modify, but I can't.

You keep on harping on that "modifyable", but no-where in the GPLv2 is 
that an issue. I claim that it *cannot* be an issue, since CD's are 
obviously ok.

So the "modifyable" part is a totally new thing to the GPLv3.

You cannot use that as an argument that the GPLv3 didn't change things, 
that's a circular agument: "the GPLv3 says so, so thus the GPLv3 is in the 
same spirit as the GPLv2". Doesn't make sense.

The fact is, the GPLv3 does fundamentally new things. Things I didn't sign 
up for (and things that nobody _else_ signed up for either) when I chose 
the GPLv2 for the kernel.

The fact that some people would like to change the kernel license to GPLv3 
is no different from the fact that some other people would like to cgange 
the kernel license to the BSD license.

Those people who have argued for using the BSD license, btw, argued so in 
the name of "freedom". No different from you. Do you think they were 
right? If so, why the hell do you think _you_ are right?

So here's what it fundamentally boils down to:

 - do you admit that the GPLv3 is a new license that does new and 
   different things?

 - do you admit that I chose the GPLv2, and have argued that I chose it 
   because I understood what it said?

 - do you admit that authors have the right to choose their own licenses?

And here's the fundamental answer:

 - if you answered "no" to any of the above questions, you're either 
   stupid (the first two questions) or a douche-bag (the third one)

 - if you answered "yes" to all the above questions, HOW THE HELL can you 
   call me confused, and argue against me when I say that the GPLv2 is a 
   better license? It wasn't your choice. 

It really is that easy. 

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 10:23:20AM +0200, Bernd Paysan wrote:
> A number of kernel hacker deliberately want their work under GPLv2 only 
> (like Al Viro), and they are fully entitled to do that - but they must 
> announce it in a propper place (not lkml or lwn.org)

A court deposition if somebody tries to do relicensing.  At that point
I believe that I made myself sufficiently clear, so I really doubt that
"all files without explicit license get the license defendant would like
and not the one located in the tree" would fly.  But you are welcome to
test that, of course - will make for nice punitive damages.  Just make
sure to test it yourself - giving somebody else an advice that will land
them in trouble is not nice...

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 1549 bytes --]

On Thursday 14 June 2007 16:08, Alan Milnes wrote:
> Agreed - if you want to take my work you are welcome as long as you
> contribute back your changes.  That's the deal that GPL2 enforces and
> why it has been so successful.

That may be a side effect of the GPL, but it's actually not how the GPLv2 
works (nor is it the intention). "Contribute back" means upstream. There's 
no such provision in the GPLv2, you contribute only downstream. And there 
are cases where you don't need to contribute at all.

E.g. the kernel hacking I'm doing at the moment: I have bought a uClinux 
blackfin board, for testing my digital audio amplifier. For that, I took 
one of the blackfin alsa audio drivers, and changed it so that it could 
talk to my digital audio amplifier. I'm not distributing this software, 
it's a complete in-house project, so I'm not obliged to contribute back. At 
the moment, I'm the only person in the world who has both access to the 
digital audio amplifier and the blackfin board, so releasing this driver in 
that early stage is a rather pointless excercise.

I think this above explains fairly well the "misunderstandings" that are 
appearing here. The GPL is not reflective (tit-for-tat), it's transient. If 
there's a loop in the transient propagation, it becomes reflective through 
the loop, but not by itself. This was the case in GPLv1, is the case in 
GPLv2, and will be the case in GPLv3.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Michael Poole]

Matt Keenan writes:

> Michael Poole wrote:
>> Matt Keenan writes:
>>
>>   
>>> Alexandre Oliva wrote:
>>>     
>>>> Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
>>>> require some physical support.  That's the hard disk in the TiVO DVR,
>>>> in this case.  I don't see how this matters, though.
>>>>
>>>>   
>>>>       
>>> I'm now intrigued, where are these (Brazilian and US) definitions
>>> stipulated, and under what authority?
>>>     
>>
>> In the US, 17 USC 101 (the "Definitions" section of the title dealing
>> with Copyright) makes this definition:
>>
>>     A "computer program" is a set of statements or instructions to be
>>     used directly or indirectly in a computer in order to bring about
>>     a certain result.
>>
>> As its purpose is to outline the scope of copyright law, this
>> definition is made under the authority granted to Congress by Article
>> I, Section 8 of the United States Constitution.
>>
>>   
> But where is the part that says it "requires some physical support"? It
> says what it is; "a set of statements or instructions", how it should be
> used; "to be used directly or indirectly in a computer", and what
> purpose it serves; "in order to bring about a certain result", but it
> doesn't seem to indicate that it "requires physical support" aka needing
> some physical representation. I suspect this argument boils down to the
> philosophical debate of whether ideas (in this case software) can be
> truely devoid of the physical.

Sets of statements or instructions that cannot "be used directly or
indirectly in a computer in order to bring about a certain result"
are, for the purposes of copyright law, not software.  "A computer" is
a physical device.  It always has been a physical device, except when
"computer" referred to a person who performed computations -- and that
meaning fell out of common use 40 years ago.  Any suggestion that the
requirement to be usable on a physical device is significantly
different from "require[s] some physical support" is laughably stupid.

17 USC 102 requires that copyright protection only subsists in works
that are "fixed in any tangible medium of expression" -- which
obviously includes paper and hard drives, and has been ruled to
include volatile program memory (the 9th Circuit's holding to this
effect in MAI Systems Corp. v. Peak Computer, Inc. is what inspired
the addition of 17 USC 117(c)).  If the set of instructions exist only
in transmission or in someone's head, they are not protected by
copyright law.

Michael Poole

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 14, 2007, Dmitry Torokhov <dtor@insightbb.com> wrote:
>
> > On Wednesday 13 June 2007 21:59, Alexandre Oliva wrote:
> >> For example, if you distribute copies of such a program, whether
> >> gratis or for a fee, you must give the recipients
> >> all the rights that you have.
> >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
> > So if I am a sole author of a program and I chose to distribute it under
> > GPL
>
> then you're not a licensee, you're a licensor, and these terms don't
> apply to you.

Heh. When you change a GPLed program and pass your changes you are the
licensor for the new code. You still have a right and license pieces
of the code you wrote under different license but you do not pass that
right to recepient of modified work.

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Adrian Bunk wrote:
> 
> Here in Germany, the rules at court are roughly "the loser pays 
> everything including the costs of the winner", so if a big company is 
> sure they will win at court there's no reason not to go there.

Well, the thing is (and I've said this before), a lawsuit is (and _should_ 
be) very much a last resort.

I think that the Open Source community (and the FSF too) is much better 
off *not* concentrating so much on "legal rules" of what can and cannot be 
done, and instead spend much more effort on showing people why the whole 
"Open Source" thing actually works.

And in fact, I think that's _exactly_ what Linux has been doing for the 
last decade!

A lot of companies are actually doing the Right Thing (tm).

Not because of anybody "forcing" them, but because they have literally 
bought into the whole "Open Source can do things better" mentality. 

In fact, the whole "coercive" approach is counter-productive. It makes 
people dislike you. It makes companies _resist_ open source, rather than 
see it as a potential ally. 

And no, I'm not speaking out of my *ss. Anybody who goes back fifteen 
years and looks at how the FSF was acting wrt the GPL (v2, back then), and 
how many friends - and enemies - they were making, should see that as a 
big clue. Linux really *did* change the landscape - for the better (*). By 
being much less contrary.

So look at Intel in the open source space. They're doing well. Look at 
Sun. They aren't _forced_ to open-source, they see others open-sourcing, 
and they see that it works damn well.

In the "Tivo space", look at Neuros. 

In other words, we're just *much* better off with a friendly license and 
not trying to force people to choose sides, than with the rabid idealism 
that was - and still is - the FSF. The FSF always makes for this horrible 
"you're with us, or you're against us" black-and-white mentality, where 
there are "evil" companies (Tivo) and "good" companies (although I dunno 
if the FSF really sees anybody as truly "good").

I'd much rather just see "individuals" and "companies". They're not evil 
or good, they are all in it for their own reasons (and their reasons are 
*NOT* the same reasons they are for me, you, or anybody else), and we 
should show them that the whole "Open Source" approach really does work 
for them.

It's totally pointless to try to "force" people to be good. That's like 
"curing" gay people. Not going to happen. 

			Linus

(*) Not just Linux, of course, but I do claim that this is actually an 
area where Linux was a big influence. Not the only one, but a major 
player.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Kevin Fox]

On Wed, 2007-06-13 at 20:42 -0400, Daniel Hazelton wrote:
<SNIP>
> >
> > Do you deny that TiVo prevents you (or at least a random customer)
> > from modifying the copy of Linux that they ship in their DVR?
> 
> Exactly. They don't. What TiVO prevents is using that modified version on 
> their hardware. And they have that right, because the Hardware *ISN'T* 
> covered by the GPL.

The hardware isn't directly covered by the GPL, correct. But, if they
want to use the software on the hardware, they have to comply with the
GPL. The software license can then influence hardware IF they want to
use it badly enough.

For example, the hardware is perfectly capable of being used to break
the terms of the GPL by being used to distribute a modified binary
without releasing the source. But the hardware's behavior is restricted
by the software for the betterment of all.

This whole argument is about the spirit of the GPL. Linus and others
think the spirit is one thing, the FSF guys think its something else.
Since the license is clearly owned by the FSF, I think they get the
final vote on what they "intended" it to be when they wrote it, no? If
they say they intended it to not allow Tivoization then believe them,
because they are the only ones that know what they were thinking when
they wrote it! The GPLv2 seems to allow it though. If Linus and friends
want to allow it, then they can stay with the GPLv2. For those who want
to disallow Tivoization, choose v3. No worries guys.

> Do you understand that, or do I need to break out the finger-puppets next ?

Guys, we are all friends here. No reason to be so insulting. Its just a
difference of opinion. People seem to be talking past each other instead
of to one another. This usually happens when people are basing their
underlying assumptions on different things and not listening to the
other. Please take a step back and think about it.

<SNIP>

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Kevin Fox wrote:
> 
> The hardware isn't directly covered by the GPL, correct. But, if they
> want to use the software on the hardware, they have to comply with the
> GPL.

Only with the GPLv3.

Again, don't confuse the *new* requirements in the GPLv3 with any "GPL 
requirements". They didn't exist before. The kernel never signed up to 
them. They are irrelevant for the discussion.

So hardware details have *nothing* to do with compying with the GPLv2.

Could you write *another* license that puts limitations on the hardware or 
environment that you have to comply with? Sure can. And the GPLv3 does 
that. But the GPLv2 does not, and that's a fundmanetal *improvement* over 
the GPLv3 in my opinion.

Do you like licenses that force the licensee to give money back?

So why do you like licenses that force the licensee to give access to 
hardware back? It's a form of "extra compensation" that the GPLv2 never 
had. The GPLv2 talks about giving access to the *source* code. The GPLv3 
talks about giving access to the *hardware*. 

Can people really not see the difference, and why I might think it's a 
fundamental difference, and why I might choose to say that the GPLv3 is a 
worse license?

And *why* would I ever downgrade to a worse license? There had better be 
some really pressing reason to choose the worse version of the GPL. And I 
just don't find that reason in the GPLv3 itself - although, as mentioned, 
the reason could become *external* (ie I might accept a worse license it 
it comes with external code attached to it that I think makes up for the 
license deficiency).

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bernd Paysan]

[-- Attachment #1: Type: text/plain, Size: 3963 bytes --]

On Thursday 14 June 2007 17:39, Al Viro wrote:
> On Thu, Jun 14, 2007 at 10:23:20AM +0200, Bernd Paysan wrote:
> > A number of kernel hacker deliberately want their work under GPLv2 only
> > (like Al Viro), and they are fully entitled to do that - but they must
> > announce it in a propper place (not lkml or lwn.org)
>
> A court deposition if somebody tries to do relicensing.  At that point
> I believe that I made myself sufficiently clear, so I really doubt that
> "all files without explicit license get the license defendant would like
> and not the one located in the tree" would fly.  But you are welcome to
> test that, of course - will make for nice punitive damages.

If I test it, it would be in Germany, and I really doubt that relicensing 
one copyleft to another can ever cause puntative damages here. You are only 
entitled to collect damages here when you have actual losses (that's why 
Harald Welte never gets a dime except for his defense expenses and 
voluntary donations to the FSF), but you can demand compliance. That would 
basically mean that the hypothetical linux-something.subversion-bp with 
GPLv3 parts in it can't be shipped further, because I can't fulfill all my 
obligations.

It's probably completely hypothetical, but if I really liked to be nasty, I 
could release the blackfin sound driver I've written for our digital 
amplifyer under GPLv3 or later. The code I've modified is explicitely under 
GPLv2 or later.

> Just make 
> sure to test it yourself - giving somebody else an advice that will land
> them in trouble is not nice...

As a non-lawyer, I can't give anybody legal advice in Germany, and I'd like 
to extend that to the rest of the world. This is my opinion, my 
interpretation of the GPLv2 and what's my logical reasoning what these 
three lines on top of /usr/src/linux/COPYING really mean. And there are 
only two possibilities:

* Either it means what it says, then it's quite likely a copyright 
infingement done by Linus to all those authors of linux-2.4.0-test8 and 
before, and you all may need to stop distributing Linux*, since you can't 
meet your obligations (and restart from linux-2.4.0-test8, which is the 
last legal version), or

* it does not exactly mean what it says, then you still can distribute 
Linux, but you can't really stop anyone who's updating it to GPLv3 - except 
for those few files that have explicit version numbers assigned.

BTW: If I grep through Linux, I find two files where you have noted your 
copyright and the release conditions (GPL v2), and I think last time I did 
the same thing, I found two GPLv2-files, as well - all other files with "Al 
Viro" in it apparently have multiple authors. These two files may be the 
same ones, or maybe there are two other files, making it four in total (or 
some further I missed, the text of v2 only is not as normed as the text 
for "v2 or later", but in general it's rare). These files clearly have to 
be rewritten or premission has to be asked when updating COPYING to GPLv3. 
But that's not a show-stopper.

Example, to test the legal issue:

/usr/src/linux/drivers/net/bmac.c

is Copyright (C) 1998 Randy Gobbel, and was modified by you in 1999, and has 
no specific GPL license. That's both before 2.4.0-test9, so without the 
implied "v2 only" of the "tree root". If I take this file up to GPLv3, what 
can you do against it? Nothing. What can Randy Gobbel do about it? Nothing, 
either. If you claim, it's "v2 only", Randy Gobbel could do something (e.g. 
asking puntative damages from you in the US, or denying you the right to 
redistribute Linux further, because you don't fulfill your obligations).

*) e.g. Microsoft lawyers will hunt down all pre-2.4.0-test9 copyright 
holders and pay the one who's willing to stop Linux a billion.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 06:32:57PM +0200, Bernd Paysan wrote:

> BTW: If I grep through Linux, I find two files where you have noted your 
> copyright and the release conditions (GPL v2), and I think last time I did 
> the same thing, I found two GPLv2-files, as well - all other files with "Al 
> Viro" in it apparently have multiple authors. These two files may be the 
> same ones, or maybe there are two other files, making it four in total (or 
> some further I missed, the text of v2 only is not as normed as the text 
> for "v2 or later", but in general it's rare). These files clearly have to 
> be rewritten or premission has to be asked when updating COPYING to GPLv3. 
> But that's not a show-stopper.
 
Rot.  "Multiple authors" doesn't get you out of that.  If you take a code
available under GPLv2 or later and combine it with code under specific
version of GPL, result is under than specific version of GPL.  If you want
to argue against that, make sure to Cc RMS on that, I would really like to
hear his opinion.

Multiple authors == need permission from each author with enough
contributions to that file to make the contributions in question
copyrightable.

And in my case (and case of gregkh, and...) that would be considerably
more than a couple of files.  Really.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Bron Gondwana <brong@fastmail.fm> wrote:

> On Thu, Jun 14, 2007 at 01:58:26AM -0300, Alexandre Oliva wrote:

>> Do we actually get any benefit whatsoever from TiVO's choice of Linux
>> as the kernel for its device?

> Sure, if they make any changes or fixes to Linux.  Other than that,
> only the same benefit that Microsoft get from Windows piracy - TiVo
> employees become familiar with Linux and are more likely to use it
> and maybe contribute more in another job later.

Now, what if TiVO actually permitted all of its customers to make
changes or fixes to Linux, and become familiar with it and become more
likely to use it and maybe contribute more later?

Would we lose more or gain more?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> And? There is *absolutely* *nothing* in any version of the GPL *prior* to 3 
> that says that hardware cannot impose restrictions.

It's not that the hardware is deciding to impose restrictions on its
own.  It's the hardware distributor that is deciding to use the
hardware to impose restrictions on the user.  Seems like a violation
of section 6 of GPLv2 to me.

> What the GPL *does* say is that you can't "add additional
> restrictions to the license"

Not quite.  It's more general than that:

  You may not impose any further restrictions on the recipients'
  exercise of the rights granted herein.

>> > So take another example: I obviously distribute code that is copyrighted
>> > by others under the GPLv2. Do I follow the GPLv2? I sure as hell do! But
>> > do I give you the same rights as I have to modify the copy on
>> > master.kernel.org as I have? I sure as hell DO NOT!

>> That's an interesting argument.

>> People don't get your copy, so they're not entitled to anything about
>> it.

>> When they download the software, they get another copy, and they have
>> a right to modify that copy.

> But you get the TiVO corporations copy of the software?

Yes.  The customer gets the copy that TiVO stored in the hard disk in
the device it sells.  And it's that copy that the customer is entitled
to modify because TiVO is still able to modify it.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Paul Mundt <lethal@linux-sh.org> wrote:

> I don't see how you can claim that the vendor is infringing on your
> freedom, _you_ made the decision to go out and buy the product knowing
> that the vendor wasn't going to go out of their way to help you hack
> the device.

But I also made this decision fully aware that the software included
in the package was published under a license that said I was entitled
to modify it.  More than once I purchased a device that claimed to
have GNU/Linux software on it, only to find out that I couldn't use
the freedoms, because the distributor was infringing the license in
various ways.

> If you don't like what the vendor has done with the product, you have the
> freedom to not support the vendor, and to try and encourage people to
> follow suit.

Sure.  But wouldn't it be nice if the copyright holder could also help
in this effort?  It doesn't mean the copyright holder has to: s/he can
always grant an additional permission, or simply refrain from
enforcing this provision of the license.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Matt Keenan <tank.en.mate@gmail.com> wrote:

> Alexandre Oliva wrote:

>> Err, no.  Software, per legal definitions in Brazil, US and elsewhere,
>> require some physical support.  That's the hard disk in the TiVO DVR,
>> in this case.  I don't see how this matters, though.

> I'm now intrigued, where are these (Brazilian and US) definitions
> stipulated, and under what authority?

http://www.planalto.gov.br/ccIVIL_03/LEIS/L9609.htm

LEI Nº 9.609 , DE 19 DE FEVEREIRO DE 1998.

Art. 1º Programa de computador é a expressão de um conjunto organizado
de instruções em linguagem natural ou codificada, contida em suporte
                                                  ^^^^^^^^^^^^^^^^^^
físico de qualquer natureza, de emprego necessário em máquinas
^^^^^^^^^^^^^^^^^^^^^^^^^^^
automáticas de tratamento da informação, dispositivos, instrumentos ou
equipamentos periféricos, baseados em técnica digital ou análoga, para
fazê-los funcionar de modo e para fins determinados.


-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Al Viro <viro@ftp.linux.org.uk> wrote:
> On Thu, Jun 14, 2007 at 06:32:57PM +0200, Bernd Paysan wrote:
>
> > BTW: If I grep through Linux, I find two files where you have noted your
> > copyright and the release conditions (GPL v2), and I think last time I did
> > the same thing, I found two GPLv2-files, as well - all other files with "Al
> > Viro" in it apparently have multiple authors. These two files may be the
> > same ones, or maybe there are two other files, making it four in total (or
> > some further I missed, the text of v2 only is not as normed as the text
> > for "v2 or later", but in general it's rare). These files clearly have to
> > be rewritten or premission has to be asked when updating COPYING to GPLv3.
> > But that's not a show-stopper.
>
> Rot.  "Multiple authors" doesn't get you out of that.  If you take a code
> available under GPLv2 or later and combine it with code under specific
> version of GPL, result is under than specific version of GPL.  If you want
> to argue against that, make sure to Cc RMS on that, I would really like to
> hear his opinion.
>
> Multiple authors == need permission from each author with enough
> contributions to that file to make the contributions in question
> copyrightable.
>
> And in my case (and case of gregkh, and...) that would be considerably
> more than a couple of files.  Really.

I would expect that if you contribute to a file that explicitely says
"GPL v2 or later" and you do not change that wording then you agree
GPL v2 or later for that particular contribution. So for example
drivers/net/plip.c could be changed to GPL v3 even though you
contributed to it.

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 01:01:20PM -0400, Dmitry Torokhov wrote:
> >Multiple authors == need permission from each author with enough
> >contributions to that file to make the contributions in question
> >copyrightable.
> >
> >And in my case (and case of gregkh, and...) that would be considerably
> >more than a couple of files.  Really.
> 
> I would expect that if you contribute to a file that explicitely says
> "GPL v2 or later" and you do not change that wording then you agree
> GPL v2 or later for that particular contribution. So for example
> drivers/net/plip.c could be changed to GPL v3 even though you
> contributed to it.

After you exclude such cases it's still more than a couple of files...

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Sean]

On Thu, 14 Jun 2007 09:01:32 -0700 (PDT)
Linus Torvalds <torvalds@linux-foundation.org> wrote:

> In other words, we're just *much* better off with a friendly license and 
> not trying to force people to choose sides, than with the rabid idealism 
> that was - and still is - the FSF. The FSF always makes for this horrible 
> "you're with us, or you're against us" black-and-white mentality, where 
> there are "evil" companies (Tivo) and "good" companies (although I dunno 
> if the FSF really sees anybody as truly "good").

Linus,

If you really believe that then why didn't you choose a BSD license
for Linux?  You didn't say "completely free, no restrictions attached,
people will follow because they'll see it's best, we just won't buy
products that use Linux in a way with which we disagree".

Instead you chose a license which enforced the so called tit-for-tat
policy you think is fair.  But people who prefer the BSD license may
think you're a moron for forcing your political agenda (ie. tit-for-tat)
on users of your code.  The point of all that being, you _do_ believe
in enforcing restrictions or you wouldn't like the GPL v2.

So you draw the line of "fairness" and belief that people will
do-the-right-thing somewhere short of the BSD license.  Why is it
so hard then to accept that the FSF draws the line short of the
GPLv2 after having gained practical experience with it
since its release?

You can argue till the cows come home the belief that _your_
restrictions are more fair, moral and reasonable than theirs.
But at the end of the day it's all just a matter of opinion about
what constitutes fair and reasonable.  You think its a fair trade
that you get code back, the FSF think its fair that people can hack
and run the code anywhere its used..  It all comes down to the
author of the code getting to attach whatever restrictions they
choose.

Sean

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> It's not that the hardware is deciding to impose restrictions on its
> own.  It's the hardware distributor that is deciding to use the
> hardware to impose restrictions on the user.  Seems like a violation
> of section 6 of GPLv2 to me.

You *still* haven't figured out the difference between "the software" and 
"a particular copy of the software", have you?

What's your problem?

I doubt you're really stupid, so I think your problem is that if you admit 
that "the software" is something *different* from "a particular copy of 
the software", you realize (perhaps subconsciously) that your arguments do 
not make any sense. So you do not allow yourself to think clearly about 
the matter.

So let's look at that "section 6" that you talk about, and quote the 
relevant parts, will  we:

	You may not impose any further restrictions on the recipients' 
	exercise of the rights granted herein.

and then let's look at Red Hat sending me a CD-ROM or a DVD.

Now, Red Hat clearly *did* "further restrict" my rights as it pertains TO 
THAT COPY ON THE CD-ROM! I cannot change it! Waa waa waa! I'll sue your 
sorry ass off!

See the issue? You are continually making the mistake of thinking that the 
GPLv2 talks about individual copies of software. And I'm continually 
having to point out that that is a MISTAKE.

And it's OBVIOUSLY a mistake, because your reading is nonsensical. If you 
think that Tivo does somethign bad, then hat Red Hat does is the same 
badness, thousads times over! I strongly suspect Red Hat has shipped a lot 
more CD-ROM's than Tivo has shipped boxes!

So let me iterate AGAIN:

 - the rights that the GPLv2 gives *cannot* be about "the particular copy" 
   that you send, since that would be INSANE. Red Hat sends lots of copies 
   of software that are NOT MODIFIABLE!

 - ergo, the rights about "the software" in the GPLv2 must be about 
   something else. 

See? Your argument about "individual copies" simply DOES NOT MAKE SENSE!

Just admit it.

> > What the GPL *does* say is that you can't "add additional
> > restrictions to the license"
> 
> Not quite.  It's more general than that:
> 
>   You may not impose any further restrictions on the recipients'
>   exercise of the rights granted herein.

And your point is? Nothing.

The rights granted are the rights to "distribute and modify the software". 
But by "the software", the license is not talking about a particular 
*copy* of the software, it's talking about the software IN THE ABSTRACT.

In other words, the reason that Red Hat is not violating the GPLv2 is that 
no, I cannot change the copy on the software on the particular CD-ROM or 
DVD, but I can get a copy of the sources other ways, and make my own 
modifications *SOMEWHERE*ELSE*!

The fact that Red Hat made a "restricted copy" is totally irrelevant.

In fact, it's exactly as irrelevant as the fact that Tivo makes a 
"restricted copy". The *software* is still free!

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Adrian Bunk]

On Thu, Jun 14, 2007 at 09:01:32AM -0700, Linus Torvalds wrote:
> On Thu, 14 Jun 2007, Adrian Bunk wrote:
> > 
> > Here in Germany, the rules at court are roughly "the loser pays 
> > everything including the costs of the winner", so if a big company is 
> > sure they will win at court there's no reason not to go there.
> 
> Well, the thing is (and I've said this before), a lawsuit is (and _should_ 
> be) very much a last resort.
> 
> I think that the Open Source community (and the FSF too) is much better 
> off *not* concentrating so much on "legal rules" of what can and cannot be 
> done, and instead spend much more effort on showing people why the whole 
> "Open Source" thing actually works.

I'm wondering more and more why you choose the GPL and not the BSD 
licence for the Linux kernel...

Companies are violating the GPL and this only works as long as noone 
starts taking legal actions against them.

And taking legal actions easily takes all of your fortune at risk.

> And in fact, I think that's _exactly_ what Linux has been doing for the 
> last decade!
> 
> A lot of companies are actually doing the Right Thing (tm).
> 
> Not because of anybody "forcing" them, but because they have literally 
> bought into the whole "Open Source can do things better" mentality. 

And there are some companies for whom it's better if they can take the 
open source code and turn it into some closed thing. There's a reason 
why TiVo warned it's investors that the GPLv3 might hurt their business.

> In fact, the whole "coercive" approach is counter-productive. It makes 
> people dislike you. It makes companies _resist_ open source, rather than 
> see it as a potential ally. 

And what are the risk of your allies actions?

Consider e.g. that your ally AMD offers legally questionable non-GPL 
modules and Debian shipping binaries.

If one of the many copyright holders of the kernel wants to take legal 
actions against this suing AMD might simply be out of reach due to 
financial reasons.

But legal actions against the maintainers of ftp.<your_country>.debian.org 
distributing binaries of these legally questionable modules have a good 
chance of success.

The legal risks might not be a problem for a big company like AMD, but 
anyone seeking an easy legal win will _naturally_ attack mirrors.

And considering the lucratice "cease and desist letter" business in 
Germany, it really seems to be only a matter of time until mirrors will 
have to pay (the lawyer costs of the ones sending the cease and desist 
letters) for the actions of your "allies".

> And no, I'm not speaking out of my *ss. Anybody who goes back fifteen 
> years and looks at how the FSF was acting wrt the GPL (v2, back then), and 
> how many friends - and enemies - they were making, should see that as a 
> big clue. Linux really *did* change the landscape - for the better (*). By 
> being much less contrary.
> 
> So look at Intel in the open source space. They're doing well. Look at 
> Sun. They aren't _forced_ to open-source, they see others open-sourcing, 
> and they see that it works damn well.

Look at AMD.
Look at NVIDIA.

Will they ever switch from providing illegal modules to open source?

> In the "Tivo space", look at Neuros. 
> 
> In other words, we're just *much* better off with a friendly license and 
> not trying to force people to choose sides, than with the rabid idealism 
> that was - and still is - the FSF. The FSF always makes for this horrible 
> "you're with us, or you're against us" black-and-white mentality, where 
> there are "evil" companies (Tivo) and "good" companies (although I dunno 
> if the FSF really sees anybody as truly "good").
> 
> I'd much rather just see "individuals" and "companies". They're not evil 
> or good, they are all in it for their own reasons (and their reasons are 
> *NOT* the same reasons they are for me, you, or anybody else), and we 
> should show them that the whole "Open Source" approach really does work 
> for them.
>...

Until someone starts to attack ftp mirrors and other people distributing 
the questionable code.

There are so many copyright holders of the Linux kernel (including e.g. 
the former Caldera), and if someone wants to spread fear among Linux 
users legal actions against people maintaining ftp mirrors or selling 
their old Linux distribution at ebay will be the cheap and effective way 
to achieve this goal.

> 			Linus
>...

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Al Viro <viro@ftp.linux.org.uk> wrote:
> On Thu, Jun 14, 2007 at 01:01:20PM -0400, Dmitry Torokhov wrote:
> > >Multiple authors == need permission from each author with enough
> > >contributions to that file to make the contributions in question
> > >copyrightable.
> > >
> > >And in my case (and case of gregkh, and...) that would be considerably
> > >more than a couple of files.  Really.
> >
> > I would expect that if you contribute to a file that explicitely says
> > "GPL v2 or later" and you do not change that wording then you agree
> > GPL v2 or later for that particular contribution. So for example
> > drivers/net/plip.c could be changed to GPL v3 even though you
> > contributed to it.
>
> After you exclude such cases it's still more than a couple of files...
>

Undoubtedly. I was just responding to neet to contact multiple authors point.

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Paulo Marques]

Al Viro wrote:
> On Thu, Jun 14, 2007 at 01:01:20PM -0400, Dmitry Torokhov wrote:
>>> Multiple authors == need permission from each author with enough
>>> contributions to that file to make the contributions in question
>>> copyrightable.
>>>
>>> And in my case (and case of gregkh, and...) that would be considerably
>>> more than a couple of files.  Really.
>> I would expect that if you contribute to a file that explicitely says
>> "GPL v2 or later" and you do not change that wording then you agree
>> GPL v2 or later for that particular contribution. So for example
>> drivers/net/plip.c could be changed to GPL v3 even though you
>> contributed to it.
> 
> After you exclude such cases it's still more than a couple of files...

FWIW,

$ find -name "*.c" | xargs grep "any later version" | wc -l
3138
$ find -name "*.c" | wc -l
9482

Watching the output of the first grep without "wc -l" shows that, 
although it is not 100% accurate, it is still ok just to get a rough 
estimate.

So yes, ~6300 files are definitely more than a couple ;)

-- 
Paulo Marques - www.grupopie.com

"God is love. Love is blind. Ray Charles is blind. Ray Charles is God."

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 09:06:31AM -0700, Kevin Fox wrote:
 
> This whole argument is about the spirit of the GPL. Linus and others
> think the spirit is one thing, the FSF guys think its something else.

What the fsck it is, linux-kernel or bleeding Council of Nikea?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Thursday 14 June 2007 03:11:45 Alexandre Oliva wrote:
>> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
>> > Ah, well... In the case of "Windos" and other proprietary OS's I try to
>> > educate people and get them to switch.

>> Good.  So I presume you'd tell them to switch away from a
>> turned-proprietary GNU/Linux operating system as well, right?

> If that happened I'd be lost. I've tried the various BSD's and found they had 
> problems with hardware support and getting a new version of the BSD kernel to 
> compile and boot is something of a black art.

> The point is moot, though. It can never happen.

Look again, it's already happened in the TiVO and other devices.

The software that ships in them is no longer Free Software.


Consider a new microprocessor.

Consider that Linux is ported to it by the microprocessor
manufacturer.

Consider that the manufacturer only sells devices with that
microprocessor with TiVO-like locks.

How exactly can you enjoy the freedoms WRT the GPLed software you got
from the manufacturer?


Now consider that you have a single computer, and that's built by TiVO.

How exactly can you enjoy the freedoms the author meant you to have,
if the TiVO box won't run the program after you modify it?

> If this "run modified copies on the same hardware you received the
> original on" *IS* the "spirit" of the license, then why isn't it
> stated anywhere before GPLv3?

For the same reasons that the pro-DRM laws weren't mentioned before,
and the patent retaliation clauses weren't mention before: these
specific cases hadn't been studied, only the general idea of
respecting users' freedoms was.

> I'll grant you that. But, at this point, where can I find a copy of
> the GPLv1 without having to dig around the net ?

In the program you received under GPLv1.

Hey, you said there was code under GPLv1.1 in the Linux tree.  Then,
there should be a copy of GPLv1.1 in there, otherwise AFAICT the
distribution of that code is copyright infringement.  IANAL.

>> In contrast, your TiVO may get a software upgrade without your
>> permission that will take your rights away from that point on, and
>> there's very little you can do about it, other than unplugging it from
>> the network to avoid the upgrade if it's not too late already.

> And because its a device that connects to their network - and TiVO
> isn't a telecommunications company - they have the right to upgrade
> and configure the software inside however they want. (In the US at
> least)

But do they have the right to not pass this right on, under the GPL?


>> > A lot of them would probably have private modifications that would
>> > never be distributed - and under the GPLv2 it is clear that you can
>> > keep modifications private as long as you don't distribute them.

>> Likewise with GPLv3.

> I can see this, but will a company see this?

In what sense does the GPLv3 make this particular point any less
obscure?

> True. But that doesn't save them from lawsuits trying to force them
> to obey the terms of the new revision even though they received the
> software under an earlier version.

Nothing saves anyone from silly lawsuits.  This one would likely be
laughed out of court in no time.  Anyone worried about this should
also be concerned about their neighbor suing them for copyright
infringment every time they set their stereo loud enough for the
neighbors to listen and be annoyed.  (Hint: only the copyright holder
would stand a chance of winning such a lawsuit)

>> > (and don't try to argue that even though those modifications are
>> > truly private (to the company) they should be released anyway to
>> > comply with the "spirit" of the license. It is made clear that it
>> > isn't by the text of the license itself)
>> 
>> How could you possibly come to the conclusion that forcing anyone to
>> release private modifications would be in compliance with the spirit
>> of the license?  can != must

> I was trying to be sarcastic and inject a little humor here. Guess I should 
> have used the old <sarcasm> tag :)

Aah.  I'm not sure I'd have understood it either.

>> >> > Why should I repeat Linus' explanation of the ways that GPLv3 violates
>> >> > the spirit of GPLv2?
>> >>
>> >> Don't worry about parrotting here, he hasn't provided that explanation
>> >> yet ;-)  Please give it a try.
>> >
>> > But he has. Whether you have accepted that his explanations are
>> > valid or not doesn't change the fact.
>> 
>> His explanation is based on a reading of the license that doesn't
>> match what its authors meant.  I guess the authors know better what
>> they meant the spirit of the license to be than someone else who
>> studied it a lot but that until very recently couldn't even tell the
>> spirit from the legal terms.

> And his interpretation is no less valid than that of anyone else. In
> fact, after a recent conversation with a couple of lawyers that I
> know, I can state that his interpretation isn't that far off from
> theirs.

Interpretation as applied to the legal terms, yes.  As for the spirit
of the license, the authors ought to know better than anyone else what
they meant.  Sure, other interpretations might lead to different
understandings as to what the readers *think* it means, but that
doesn't change what it was *intended* to mean.

> Then you're lucky. I've had a lot of people say something similar to the 
> following: "Oh, I've heard about that. So which version of the GNU-Linux 
> kernel are you running?" 

Oh my.  That's indeed unfortunate and unfair.

> As I've stated before - I can find nothing in the history of the GPL or the 
> FSF that makes the "on the same hardware" requirement clear and part of 
> the "spirit" of "Free Software".

Put the considerations above, about a single computer or a
uniformly-limited computing platform, and you'll see that this "on the
same hardware" argument is just a means to deny people freedom.  If I
could stop you from running modified versions on one piece of
hardware, then I could on two, and 3, and then soon it's all of them,
and we're back to square zero in terms of freedom.

>> > What I won't do is release whatever tools and such that are needed to
>> > make the hardware run a different version of the kernel. Why? Because:
>> > the hardware was designed so that a specific version of the kernel runs
>> > without problems, there is hardware that is very picky and running a
>> > customized kernel could cause that hardware to fail, etc...

>> Why do you care?  It's no longer your hardware, it's theirs.

> Legal requirements in some countries that require manufacturers to provide 
> support for their product for a period of time after it has been purchased.

If you replace a component in the hardware, are you still required to
provide support or offer warranty?  Why should this be different just
because it's a software component?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Florin Malita]

Alexandre Oliva wrote:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
>   
>>> When they download the software, they get another copy, and they have
>>> a right to modify that copy.
>>>       
>> But you get the TiVO corporations copy of the software?
>>     
>
> Yes.  The customer gets the copy that TiVO stored in the hard disk in
> the device it sells.  And it's that copy that the customer is entitled
> to modify because TiVO is still able to modify it.
>   

No, by this twisted logic Tivo *cannot* modify that particular copy any 
more than you can. They can modify *another* copy (just like you) and 
they can *replace* the copy in your device with the new version (unlike 
you).

So your entire logical construct does not stand because this is not 
(cannot be) about modifying a particular copy (how would you do that 
anyway? hexedit the binary blob in place?) but about the ability to 
deploy the software on a particular platform.

---
fm

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Sean wrote:
> 
> If you really believe that then why didn't you choose a BSD license
> for Linux? 

Because I think the GPLv2 is a great license.

And I don't like the FSF's radical world-view, but I am able to separate 
the license (the GPLv2) from the author and source of the license (rms and 
the FSF).

Why do people always confuse the two? The GPLv2 stands on its own. The 
fact that I disagree with the FSF on how to act has _zero_ relevance for 
my choice of license. 

The BSD license, as far as I'm concerned, is _horrible_ for any project I 
would use. I have actually released code under it, but never a "project". 
I've given some code of mine that I don't care about that much to the BSD 
projects, just because I didn't think that code really mattered, and I 
thought it would be stupid and small-minded not to let the BSD's use it.

But for a project I actually care about, I would never choose the BSD 
license. The license doesn't encode my fundamental beliefs of "fairness". 
I think the BSD license encourages a "everybody for himself" mentality, 
and doesn't encourage people to work together, and to merge.

Let me put this in source management terms, since I've also been working 
on a source control management project for the last few years: the BSD 
license encourages "branching", but the fact is, branching is not really 
all that interesting. What's interesting is "merging": the branching is 
just a largely irrelevant prerequisite to be able to merge.

The GPLv2 encourages *merging*. Again, the right to "branch" needs to be 
there in order for merges to be possible, but the right to branch is 
actually much less important than the right to "merge".

See? 

So I'm a *big* believer in the GPLv2. I think the GPLv2 is an almost 
perfect license. That doesn't mean that I have to agree with the FSF on 
everything else.

> Instead you chose a license which enforced the so called tit-for-tat
> policy you think is fair.  But people who prefer the BSD license may
> think you're a moron for forcing your political agenda (ie. tit-for-tat)
> on users of your code.

Oh, and some people did and do.

And you know what? That's PERFECTLY OK!

I think that the BSD license is wrong for me. Does that mean that people 
who choose the BSD license are wrong to do so? No. For *them* the choices 
that the BSD license makes may be the right ones!

> The point of all that being, you _do_ believe in enforcing restrictions 
> or you wouldn't like the GPL v2.

..  but I think that the software license I choose should be about the 
software, and about giving back in kind.

And the GPLv2 is _perfect_ for that.

And the GPLv3 is horrible.

And you know what? YOU can choose the GPLv3 for your projects. I'm not 
saying anything else. I'm saying that no, I was _not_ confused when I 
chose the GPLv2. I thought it was a good license 15 years ago. I thought 
it was a good license 10 years ago. I thought it was a good license five 
years ago. And I think it's a good license today.

Because it fundamnetally does what I think is fair. 

In a way that the GPLv3 DOES NOT.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, "Alan Milnes" <alan@itfoundation.org.uk> wrote:

> Agreed - if you want to take my work you are welcome as long as you
> contribute back your changes.  That's the deal that GPL2 enforces and
> why it has been so successful.

Where did you get this impression that GPLv2 enforces this deal?

It doesn't, and this is *exactly* why I dispute the claim that GPLv2
is tit-for-tat.

> GPL3 is a very different beast with a much wider agenda,

The agenda is *precisely* the same: ensure that all users are free to
modify and share the licensed software.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Ingo Molnar <mingo@elte.hu> wrote:

> I think the proper limit is the boundary where the limit of the
> software is - because that's the only sane and globally workable way
> to stop the power-hungry.

But see, I'm not talking about getting permission to hack the
hardware.  I'm only talking about getting permission to hack the Free
Software in it.

It's your position that mingles the issues and permits people to use
the hardware to deprive users of freedom over the software that
they're entitled to have.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Adrian Bunk wrote:
> 
> I'm wondering more and more why you choose the GPL and not the BSD 
> licence for the Linux kernel...

Why do people confuse "anti-GPLv3" with "pro-BSD"?

What's the logic?

The BSD license is not doing tit-for-tat. It doesn't give me anything 
back. I don't believe in that kind of model. So I'd not use it for my 
projects.

The GPLv2 has a good balance. It encourages tit-for-tat, and it makes sure 
that the software is kept free. And it doesn't try to force anything else, 
or play politics. The only thing you have to believe in is "tit-for-tat".

The GPLv3 goes too far. It's no longer "tit-for-tat", it's "our software 
is worth _soo_ much, that we want to force you to behave well, or you 
cannot use it".

I think one of the above licenses are good. The fact that I reject the 
GPLv3 in _no_ way implies that I should like the BSD license. Both the BSD 
license and the GPLv3 are flawed - they are just flawed in fundamentally 
different ways.

So the whole question of "why don't you use he BSD license then" is just 
fundamentally bogus. A license is about a *balance* of things. "Fairness" 
is not about laissez-faire (BSD) or about total-control (GPLv3). To me, 
It's about something in the middle, where people give back in kind.

And btw, that "to me" is important. 

Different people have different opinions. That's _fine_. Use the GPLv3 for 
your projects. Go wild. Use the BSD license. It's your choice. 

But by the same token, it was _my_ choice (and it was an informed choice) 
to use the GPLv2. 

And to then come in fifteen years later and call me "confused" about a 
license I've chosen is a damn affront to me. I'm not confused. Somebody 
else may be, but it's not me.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Robin Getz <rgetz@blackfin.uclinux.org> wrote:

> On Thu 14 Jun 2007 01:07, Alexandre Oliva pondered:

>> then maybe the small 
>> company could have been more careful about the regulations.  There are
>> various ways to prevent these changes that don't involve imposing
>> restrictions of modification on any software in the device, all the
>> way from hardware-constrained output power to hardware-verified
>> authorized configuration parameters.

> As a person pretty familiar with the hardware in these types of
> devices - this just isn't practical.

I actually left out the most obvious one: store the program in ROM.
Is that not practical?

You're claiming that adding hardware locks and chains and bolts,
implemented with help from the loader software, is simpler than just
using ROM?

Well, then, ok: do all that loader and hardware signature-checking
dancing, sign the image, store it in the machine, and throw the
signing key away.  This should be good for the highly-regulated areas
you're talking about.  And then, since you can no longer modify the
program, you don't have to let the user do that any more.  Problem
solved.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> But see, I'm not talking about getting permission to hack the
> hardware.  I'm only talking about getting permission to hack the Free
> Software in it.

You have that permission. You can hack on it all you want.

Oh, but you want to hack the hardware to accept it? That's a totally 
different issue. If so, buy a Neuros OSD box.

Really. Go to google, and type in "Neuros OSD".

Do it *now*, and then stop wasting our time. You *can* do what you want to 
do.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Rene Herman]

On 06/14/2007 06:01 PM, Linus Torvalds wrote:

> It's totally pointless to try to "force" people to be good. That's like 
> "curing" gay people. Not going to happen. 

Tangent, but that could in fact quite easily be construed as saying that gay 
people aren't good which I hope is not the point you are making :-/

Rene.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> In fact, the whole "coercive" approach is counter-productive.

Let me see if I got your position right: when TiVO imposes
restrictions, that's ok, but when others want to find ways to stop it,
then it's not.  *Now* I'm confused ;-)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 665 bytes --]

On Thu, 14 Jun 2007 03:03:40 -0300, Alexandre Oliva said:
> On Jun 14, 2007, Valdis.Kletnieks@vt.edu wrote:
> 
> > If a company sells you hardware that includes a ROM that contains GPL'ed
> > software, are they in violation of the GPL if they don't include a ROM burner
> > in the hardware?  Or are ROM burners like compilers, where you have to supply
> > your own?
> 
>   this requirement does not apply if neither you nor any third party
>   retains the ability to install modified object code on the User
>   Product (for example, the work has been installed in ROM).

Do they have to provide a ROM burner if the ROM is socketed rather than
soldered into place?

[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Chris Friesen]

Alexandre Oliva wrote:

> But see, I'm not talking about getting permission to hack the
> hardware.  I'm only talking about getting permission to hack the Free
> Software in it.

No you're not...you're talking about being able to hack the software 
*and load it back onto the original hardware*.

> It's your position that mingles the issues and permits people to use
> the hardware to deprive users of freedom over the software that
> they're entitled to have.

The software license controls the software.  If the hardware has 
restrictions on it that limit what software it will run, then that is 
unrelated to the software license.

There is nothing stopping you from taking the code for the tivo, 
modifying it, distributing it, or even running it on other hardware.

Suppose I had some machine that will only run microsoft-signed binaries. 
  Would it be at all related to any software license that this machine 
won't let me run linux?

Chris

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Lennart Sorensen]

On Wed, Jun 13, 2007 at 07:38:14PM -0300, Alexandre Oliva wrote:
> So let's go back to the preamble, that provides motivations and some
> guidance as to the interpretation of the legal text (i.e., the spirit
> of the license):
> 
>   [...] the GNU General Public License is intended to guarantee your
>   freedom to share and change free software--to make sure the software
>   is free for all its users. [...]
> 
>   [...] Our General Public Licenses are designed to make sure that you have
>   the freedom to distribute copies of free software (and charge for
>   this service if you wish), that you receive source code or can get
>   it if you want it, that you can change the software or use pieces of
>   it in new free programs; and that you know you can do these things.
> 
>   To protect your rights, we need to make restrictions that forbid
>   anyone to deny you these rights or to ask you to surrender the
>   rights.  These restrictions translate to certain responsibilities
>   for you if you distribute copies of the software, or if you modify
>   it.
> 
>   [...] if you distribute copies of such a program, whether gratis or
>   for a fee, you must give the recipients all the rights that you have
> 
> 
> Can anyone show me how any of the provisions of GPLv3 fails to meet
> this spirit?

Well much as I don't like what Tivo did with only allowing signed
kernels to run, I don't see anything in the above that says they can't
do that.  They let you have the code and make changes to it, they just
don't let you put that changed stuff on the device they build.  The
software is free, even though the hardware is locked down.  The GPL v3
really seems to change the spirit to try and cover usage and hardware
behaviour, while the spirit of the GPL v2 seemed to me at least to
simply be to allow people to copy and change and use the code, and pass
that on to people.  It didn't have anything to do with what they did
with it on hardware.  Nothing prevents you from taking tivos kernel
changes and building your own hardware to run that code on, and as such
the spirit of the GPL v2 seems fulfilled.  It covers freedom of the
source code and resulting binaries, not of the platform you run it on.
The GPL v3 has a much broader coverage of what it wants to control,
which to me means the spirit is different.

I don't have a tivo, I use mythtv on my own PC.  Tivo doesn't force you
to buy their hardware after all.

--
Len Sorensen

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Alexandre Oliva wrote:
>> 
>> People don't get your copy, so they're not entitled to anything about
>> it.
>> 
>> When they download the software, they get another copy, and they have
>> a right to modify that copy.

> Umm. I notice how you must have known how *idiotic* your response was, 
> because you snipped away the part where I talked about Red Haty 
> distributing CD-ROM's.

> In other words, Red Hat distributes copies (and yes, you *get* that copy), 
> and you cannot modify that copy that you got.

And Red Hat can't either.  I thought that was quite obvious.

>> > See any parallels here? Any parallel to a CD-ROM distribution, or a Tivo 
>> > distribution?

>> Yes.  You see how TiVO is different?  It is modifyable, and I actually
>> receive the copy that TiVO can still modify, but I can't.

> You keep on harping on that "modifyable", but no-where in the GPLv2 is 
> that an issue. I claim that it *cannot* be an issue, since CD's are 
> obviously ok.

The 'passing on the rights you have' makes it an issue.

> You cannot use that as an argument that the GPLv3 didn't change things, 

Compare the preambles of v2 and v3 and you'll understand why the
argument is sound, and not circular.

> Those people who have argued for using the BSD license, btw, argued so in 
> the name of "freedom".

But individual freedom, rather than community freedom.

Think local vs global optimization.

> If so, why the hell do you think _you_ are right?

Because, like you, I'm always right, even though not everyone agrees
with that assessment? ;-P :-D

>  - do you admit that the GPLv3 is a new license that does new and 
>    different things?

Yes, of course.  The new legal terms are answers to new threats to the
freedoms depicted in the preamble, that didn't exist or hadn't been
thought of by the time GPLv2 was published.

>  - do you admit that I chose the GPLv2, and have argued that I chose it 
>    because I understood what it said?

Yes.

>  - do you admit that authors have the right to choose their own licenses?

Within the boundaries of ethics and morals, yes.

>  - if you answered "yes" to all the above questions, HOW THE HELL can you 
>    call me confused, and argue against me when I say that the GPLv2 is a 
>    better license? It wasn't your choice. 

The thing is I'm not arguing that point.

I'm disputing that there was a change in the spirit of the license
between v2 and v3.  Heck, a mere 48 hours ago you couldn't even tell
the spirit from the legal terms.


I still think v3 will serve better any Free Software community,
because it will push away the abusers that contribute little, or turn
them into cooperative or at least harmless participants, that further
enable the active participation of their downstream users.  This would
enable wider participation under the same 'tit-for-tat' conditions
that you attribute to GPLv2.


It appears to me that the only significant point of contention
remaining is the issue of Tivoization.  If you feel so strongly about
permitting Tivoization, even though it denies the freedoms that the
original spirit of the license you chose says they are entitled to
have, you can make this provision by means of an additional permission
for that, on top of GPLv3, and be done with it.

> It really is that easy. 

Yes, indeed.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> Let me see if I got your position right: when TiVO imposes
> restrictions, that's ok

Sure. I think it's ok that Microsoft imposes restrictions too on the 
software they create. It's *their* choice.

And I think it's ok for you to impose any restrictions (including the ones 
in GPLv3) on the software *you* produce. It's your choice.

> but when others want to find ways to stop it, then it's not.  *Now* I'm 
> confused ;-)

You are indeed totally confused.

It's *ok* to impose restrictions on the stuff you create. Everybody has a 
different world-view, and for some it's about making money, for some it's 
about something else, and some don't want any restrictions at all.

For me, the GPLv2 was the license I liked. I didn't like the BSD license, 
so I didn't choose it. I don't like a license that restricts hardware, so 
I didn't choose that.

And I *still* don't choose that.

See? I think the GPLv3 is a *much* inferior license to the GPLv2. It's 
better than its drafts were, but it's still doing things I disagree with. 

So tell me, why do you think I'm confused about the GPLv3? Why do you 
think I should have said "GPLv2 or anything else the FSF comes up with"? 

So the only thing I want you to say is:

 (a) Linus knows what he is doing, and isn't actually confused.

and

 (b) It was my right to use the license of my choice for a project that I 
     started.

and

 (c) I have the right to see the difference between the GPLv2 and v3, and 
     think that the GPLv3 is the inferior license.

Comprende? MY CHOICE. Not the FSF's. Not yours. Not anybody elses.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, "Dmitry Torokhov" <dmitry.torokhov@gmail.com> wrote:

> On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> On Jun 14, 2007, Dmitry Torokhov <dtor@insightbb.com> wrote:
>> 
>> > On Wednesday 13 June 2007 21:59, Alexandre Oliva wrote:
>> >> For example, if you distribute copies of such a program, whether
>> >> gratis or for a fee, you must give the recipients
>> >> all the rights that you have.
>> >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> 
>> > So if I am a sole author of a program and I chose to distribute it under
>> > GPL
>> 
>> then you're not a licensee, you're a licensor, and these terms don't
>> apply to you.

> Heh. When you change a GPLed program and pass your changes you are the
> licensor for the new code. You still have a right and license pieces
> of the code you wrote under different license but you do not pass that
> right to recepient of modified work.

You are the author of the change, and you can license them however you
like.  Your change itself is not bound by the terms of the GPL, it is
only if it is a derived work of the GPLed work.

If your change is not a derived work, you're not bound by the terms of
the GPL as far as the change is concerned, so the GPL has no say
whatsoever as to how you must release it.  If you choose the GPL, then
you're a licensor, and the requirements to pass on all the rights you
have do not apply.

If it *is* a derived work, then you're constrained by the terms of the
license, and you can only distribute it under the same license.  You
don't have a right to offer it under a different license in the first
place, so you can't pass this right on.

Derived work or not, when you combine that change with the program,
then you're bound by the terms of the license, and then you cannot
change the licensing terms of the whole program, so you can't pass
this right on either.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> > In other words, Red Hat distributes copies (and yes, you *get* that copy), 
> > and you cannot modify that copy that you got.
> 
> And Red Hat can't either.  I thought that was quite obvious.

That's TOTALLY IRRELEVANT!

There is no language in the GPLv2 (only in the GPLv3 drafts) about "same 
upgradability as third parties".

You're arguing a point that DOES NOT EXIST in the GPLv2.

The GPLv2 talks about specific rights, like the ability to make changes 
and distribute things, and says that you have to give downstream all those 
same rights.

And I've pointed out to you (now about five times) that those rights 
CANNOT be able "in-place", since even Red Hat does not actually give you 
the right to do in-place modification of the software they sell.

> The 'passing on the rights you have' makes it an issue.

No. It does not.

I have extra rights as a copyright holder, and that "the rights you have" 
are as they pertain to the software under the GPLv2, not as it pertains to 
the physical device, or outside the GPLv2.

For example, for any code that I have full copyright over, I have rights 
that you DO NOT HAVE! I have the right to re-license it under some other 
license. The fact that I pass on a copy of the software to you under the 
GPLv2 does *not* give you those rights, but that's not even what the GPLv2 
asks for!

The GPLv2, when it talks about "passing on the rights", talks about the 
rights you got *per*the*GPLv2*. 

Any other reading is nonsensical, since the copyrigth owner *always* has 
more rights than a licensee! I legally literally *couldn't* pass over all 
the rights I have to my software! If you read the GPLv2 as meaning that I 
have to, you are mis-reading it. It's that simple.

Anyway, I'm not interested in continuing this flame war.

The fact is, the license for the kernel is the GPLv2. And I think it's a 
superior license. As such, I'd be a total moron to relicense the kernel 
under what I believe is a worse license.

So if you want to argue that I should re-license, you should argue that 
the GPLv3 is better. And quite frankly, you haven't.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bill Nottingham]

Alexandre Oliva (aoliva@redhat.com) said: 
> > Sure, if they make any changes or fixes to Linux.  Other than that,
> > only the same benefit that Microsoft get from Windows piracy - TiVo
> > employees become familiar with Linux and are more likely to use it
> > and maybe contribute more in another job later.
> 
> Now, what if TiVO actually permitted all of its customers to make
> changes or fixes to Linux, and become familiar with it and become more
> likely to use it and maybe contribute more later?

a) there's nothing that prevents a Tivo user from changing or fixing
Linux completely outside of the Tivo
b) the 'interesting' bits that someone would modify the Tivo to change
*aren't actually the bits that everyone is kvetching about here*

Bill

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Thu, Jun 14, 2007 at 10:46:55AM -0700, Linus Torvalds wrote:
> 
> Oh, but you want to hack the hardware to accept it? That's a totally 
> different issue. If so, buy a Neuros OSD box.
> 
> Really. Go to google, and type in "Neuros OSD".
> 
> Do it *now*, and then stop wasting our time. You *can* do what you want to 
> do.

Unfortunatly that device has some closed source kernel drivers, so you
don't really have "full" control over the system :(

thanks,

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Bill Nottingham <notting@redhat.com> wrote:

> Alexandre Oliva (aoliva@redhat.com) said: 
>> > Sure, if they make any changes or fixes to Linux.  Other than that,
>> > only the same benefit that Microsoft get from Windows piracy - TiVo
>> > employees become familiar with Linux and are more likely to use it
>> > and maybe contribute more in another job later.
>> 
>> Now, what if TiVO actually permitted all of its customers to make
>> changes or fixes to Linux, and become familiar with it and become more
>> likely to use it and maybe contribute more later?

> a) there's nothing that prevents a Tivo user from changing or fixing
> Linux completely outside of the Tivo

But how about inside the TiVO, so as to use Linux and the rest of the
GNU/Linux distro put in there for an even better DVR experience?

Sure, this might still be accomplished on another hardware platform.
But the TiVO already has all the hardware there, and you already have
all the software ready to work on it.  Except that you can't change
it.  You'd have to waste time and money just to get to the same status
on another hardware platform.

What do we gain?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Florin Malita <fmalita@gmail.com> wrote:

> Alexandre Oliva wrote:
>> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
>> 
>>>> When they download the software, they get another copy, and they have
>>>> a right to modify that copy.

>>> But you get the TiVO corporations copy of the software?

>> Yes.  The customer gets the copy that TiVO stored in the hard disk in
>> the device it sells.  And it's that copy that the customer is entitled
>> to modify because TiVO is still able to modify it.

> No, by this twisted logic Tivo *cannot* modify that particular copy
> any more than you can. They can modify *another* copy (just like you)
> and they can *replace* the copy in your device with the new version
> (unlike you).

Again, replacing is one form of modification.

What do you think you do when you save a modified source file in your
editor?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Lennart Sorensen]

On Thu, Jun 14, 2007 at 02:26:30PM -0300, Alexandre Oliva wrote:
> In the program you received under GPLv1.
> 
> Hey, you said there was code under GPLv1.1 in the Linux tree.  Then,
> there should be a copy of GPLv1.1 in there, otherwise AFAICT the
> distribution of that code is copyright infringement.  IANAL.

So now the copy of the GPL v2 isn't good enough for the GPLv1.1 code?
Maybe that code said 'or later' in the license and hence someone added
it to a GPL v2 project since that sounds perfectly OK.

--
Len Sorensen

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Alexandre Oliva wrote:
>> 
>> It's not that the hardware is deciding to impose restrictions on its
>> own.  It's the hardware distributor that is deciding to use the
>> hardware to impose restrictions on the user.  Seems like a violation
>> of section 6 of GPLv2 to me.

> You *still* haven't figured out the difference between "the software" and 
> "a particular copy of the software", have you?

I have.  And so has GPLv2, look:

  2. You may modify your copy or copies of the Program or any portion
                         ^^^^

> What's your problem?

Trying to get you to see what is so obvious to me.

> So let's look at that "section 6" that you talk about, and quote the 
> relevant parts, will  we:

> 	You may not impose any further restrictions on the recipients' 
> 	exercise of the rights granted herein.

> and then let's look at Red Hat sending me a CD-ROM or a DVD.

> Now, Red Hat clearly *did* "further restrict" my rights as it pertains TO 
> THAT COPY ON THE CD-ROM! I cannot change it! Waa waa waa! I'll sue your 
> sorry ass off!

Red Hat is not stopping you from making changes.  The media is, and
that's not something Red Hat can control.

Compare this with the TiVO.  TiVO *designs* the thing such that it can
still make changes, but customers can't.

That's the difference.

TiVO is using hardware to "impose further restrictions on the
recipients' exercise of the rights granted herein", and this violates
section 6 of GPLv2.

> See the issue? You are continually making the mistake of thinking that the 
> GPLv2 talks about individual copies of software.

It does.  You're making the mistake of thinking that it doens't.  And
even in the legal terms that you claimed to have understood so
thoroughly.

> The rights granted are the rights to "distribute and modify the software". 

More specifically, some of the rights are:

  copy and distribute verbatim copies of the Program's source code as
  you receive it

  modify your copy or copies of the Program or any portion of it, thus
  forming a work based on the Program, and copy and distribute such
  modifications or work

> But by "the software", the license is not talking about a particular 
> *copy* of the software, it's talking about the software IN THE ABSTRACT.

Please read it again.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Paulo Marques <pmarques@grupopie.com> wrote:

> $ find -name "*.c" | xargs grep "any later version" | wc -l
> 3138
> $ find -name "*.c" | wc -l
> 9482

How many of these don't mention version 2?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 1160 bytes --]

On Thu, 14 Jun 2007 06:44:25 +0200, Michael Gerdau said:

> > > TiVo retains the right to modify that copy of Linux as it sees fit.
> > >
> > > It doesn't give the recipients the same right.
> > It does, can't you modify their kernel source? Where does it say you should be
> > able to run you modifications on the same hardware?
> 
> Come on! The whole idea of software is to have it run on some HW.
> Why would I want to change it in the first place if I can't run it ?

Maybe this quote will summarize the situation:

Judith: [on Stan's desire to be a mother] Here! I've got an idea: Suppose you
	agree that he can't actually have babies, not having a womb - which is
	nobody's fault, not even the Romans' - but that he can have the *right*
	to have babies.
Francis: Good idea, Judith. We shall fight the oppressors for your right to
	have babies, brother... sister, sorry.
Reg: What's the *point*?
Francis: What?
Reg: What's the point of fighting for his right to have babies, when he can't
	have babies?
Francis: It is symbolic of our struggle against oppression.
Reg: It's symbolic of his struggle against reality.

		-- Monty Python's "Life of Brian"



[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Neshama Parhoti]

Forgive me for a little off-topic question but I have a difficulty
to understand a technical issue about this all.

The Linux Kernel cannot easily switch licenses because of the
large amount of people involved in it (i.e. contributed code on which
they have copyright).

But many of FSF's GNU projects are similar - for example GCC has contributors
from many many companies and individuals, from which I presume there
are who might object to GPLv3.

So how come they can so easily move to GPLv3 ?
Don't they have to have permission from all of those contributors (many
of which are Linux companies and distributors who might prefer staying
at GPLv2) ?

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> ..  but I think that the software license I choose should be about the 
> software, and about giving back in kind.

> And the GPLv2 is _perfect_ for that.

> And the GPLv3 is horrible.

Is there anything other than TiVOization to justify these statements?


Also, can you elaborate on what you mean about 'giving back in kind'?
(I suspect this is related with the tit-for-tat reasoning, that you've
failed to elaborate on before)


The only thing the GPL demands is respect for others' freedoms, as in,
"I, the author, respect your freedoms, so you, the licensee, must
respect others' freedoms as well".  Is this the "in kind" you're
talking about?  Or are you mistaken about the actual meaning of even
GPLv2?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Diego Calleja]

El Thu, 14 Jun 2007 14:49:19 -0300, Alexandre Oliva <aoliva@redhat.com> escribió:

> Let me see if I got your position right: when TiVO imposes
> restrictions, that's ok, but when others want to find ways to stop it,
> then it's not.  *Now* I'm confused ;-)


Me, I agree that hardware shouldn't lock users. And since I'm one of those
evil european socialdemocrats, I may go as far as to think that there should
be laws that *forbid* selling such hardware.

But I think that all this iss a *hardware* issue. It seems to me that lot of
people at the FSF wants to regulate the hardware industry using the influence
of free/open software in the computing industry and the "V2 or later" phrase
from the GPL. 

But the fact is that free/open source runs on _top_ of hardware. You don't
control hardware, you only control the things that are built on top of your
software, not the parts you use to build your software.

And the FSF is trying to control the design and licensing of hardware throught
the influence of their software. And I think it's wrong. I'm all to forbid hardware
that imposes restrictions on hardware, but software licenses are NOT the way
to make it. That's a task for a "Free Hardware Foundation", not the FSF.

What the FSF is trying to do is EVIL. It's not about free software, it's
not about freedom, it's about the FSF trying to have to much control over
things that they shouldn't even try to control. I think that the FSF can do 
a terrible damage to free/open source with such stupid ideas. I wouldn't
even be surprised that some jugde rules that a software license that tries
to 'control' hardware is invalid

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Lennart Sorensen wrote:
> 
> So now the copy of the GPL v2 isn't good enough for the GPLv1.1 code?
> Maybe that code said 'or later' in the license and hence someone added
> it to a GPL v2 project since that sounds perfectly OK.

Where did that GPLv1.1 nonsense come from?

There is no GPLv1.1 code in the tree. By the time I selected the GPL for 
the kernel license, the GPLv1.1 had long since been discontinued. The 
kernel was *never* GPLv1.1-only compatible. That's just total nonsense.

There was indeed a kernel license before the GPLv2, but it wasn't the GPL 
at all, it was my own made-up thing. Appended here, for those who are too 
lazy to actually look up and check the original Linux-0.01 announcement.

		Linus

---
This kernel is (C) 1991 Linus Torvalds, but all or part of it may be
redistributed provided you do the following:

	- Full source must be available (and free), if not with the
	  distribution then at least on asking for it.

	- Copyright notices must be intact. (In fact, if you distribute
	  only parts of it you may have to add copyrights, as there aren't
	  (C)'s in all files.) Small partial excerpts may be copied
	  without bothering with copyrights.

	- You may not distibute this for a fee, not even "handling"
	  costs.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 14, 2007, Paulo Marques <pmarques@grupopie.com> wrote:
>
> > $ find -name "*.c" | xargs grep "any later version" | wc -l
> > 3138
> > $ find -name "*.c" | wc -l
> > 9482
>
> How many of these don't mention version 2?
>

It does not matter. GPL v2 and later can be reduced to v2 by
recepient. Linus did just that so unless individual source file
explicitely carries "and later" it is v2.

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 14, 2007, "Dmitry Torokhov" <dmitry.torokhov@gmail.com> wrote:
>
> > On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> >> On Jun 14, 2007, Dmitry Torokhov <dtor@insightbb.com> wrote:
> >>
> >> > On Wednesday 13 June 2007 21:59, Alexandre Oliva wrote:
> >> >> For example, if you distribute copies of such a program, whether
> >> >> gratis or for a fee, you must give the recipients
> >> >> all the rights that you have.
> >> >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> >>
> >> > So if I am a sole author of a program and I chose to distribute it under
> >> > GPL
> >>
> >> then you're not a licensee, you're a licensor, and these terms don't
> >> apply to you.
>
> > Heh. When you change a GPLed program and pass your changes you are the
> > licensor for the new code. You still have a right and license pieces
> > of the code you wrote under different license but you do not pass that
> > right to recepient of modified work.
>
> You are the author of the change, and you can license them however you
> like.
[... skip...]
>
> Derived work or not, when you combine that change with the program,
> then you're bound by the terms of the license, and then you cannot
> change the licensing terms of the whole program, so you can't pass
> this right on either.
>

Ok, consider non-derived work. Because I am distributing whole program
I have to do it under GPL. However I still have the right to
distribute just the portion that is written by me under whatevel
license I want but you as a recepient of GPLed whole do not get this
right. IOW I am not passing all the rights _I have_.

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> Is there anything other than TiVOization to justify these statements?

Do you need anything else?

But if by the question you mean "would you think the GPLv3 is fine without 
the new language in section 6 about the 'consumer devices'", then the 
answer is that yes, I think that the current GPLv3 draft looks fine apart 
from that.

> Also, can you elaborate on what you mean about 'giving back in kind'?
> (I suspect this is related with the tit-for-tat reasoning, that you've
> failed to elaborate on before)

I've *not* failed to elaborate on that before. Not at all. 

Just google for

	torvalds tit-for-tat

and you'll see a lot of my previous postings. Trying to claim that this is 
somehow "new" is ludicrous. In fact, some of the google hits you find are 
from 2004, *loong* before the current GPLv3 discussion.

So your "failed to elaborate" is not a failure on my side. 

Giving back "in kind" is obvious. I give you source code to do with as you 
see fit. I just expect you to give back in kind: source code for me to do 
with as I see fit, under the same license I gave you source code.

How hard is that to accept?

I don't ask for money. I don't ask for sexual favors. I don't ask for 
access to the hardware you design and sell. I just ask for the thing I 
gave you: source code that I can use myself.

I really don't think my "tit-for-tat" or "give back in kind" is that hard 
to understand, is it?

And no, it's not a new concept. Neither is the fact that I've never agreed 
with the FSF's agenda about "freedom" (as defined by _them_ - I have a 
notion of "freedom" myself, and the FSF doesn't get to define it for me).

I don't call Linux "Free Software". I haven't called it that for close to 
ten years! Because I think the term "Open Source" is a lot better.

> The only thing the GPL demands is respect for others' freedoms, as in,
> "I, the author, respect your freedoms, so you, the licensee, must
> respect others' freedoms as well".  Is this the "in kind" you're
> talking about?  Or are you mistaken about the actual meaning of even
> GPLv2?

I respect your freedom to design products around Linux. You can do 
whatever you damn well please - I just ask that you give the software back 
in a usable form. That's all I ask for.

And that's all the GPLv2 asks for. 

Which is why I selected the GPLv2 in the first place, and why I *still* 
think the GPLv2 is a wonderful license!

So I claim that the "freedoms" that the GPLv2 embodies are *greater* than 
the "freedoms" embodied in the GPLv3.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Greg KH]

On Thu, Jun 14, 2007 at 09:42:16PM +0300, Neshama Parhoti wrote:
>  Forgive me for a little off-topic question but I have a difficulty
>  to understand a technical issue about this all.
> 
>  The Linux Kernel cannot easily switch licenses because of the
>  large amount of people involved in it (i.e. contributed code on which
>  they have copyright).
> 
>  But many of FSF's GNU projects are similar - for example GCC has 
>  contributors
>  from many many companies and individuals, from which I presume there
>  are who might object to GPLv3.
> 
>  So how come they can so easily move to GPLv3 ?

The FSF required copyright assignment to themselves in order to accept
the changes from the developers.  So the FSF owns the whole copyright
and can change things whenever they want, to whatever license they want.

This is the exact opposite of the kernel in which all of the original
contributors own the copyright.

Hope this helps,

greg k-h

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Bill Nottingham]

Neshama Parhoti (pneshama@gmail.com) said: 
> But many of FSF's GNU projects are similar - for example GCC has 
> contributors
> from many many companies and individuals, from which I presume there
> are who might object to GPLv3.

FSF requires copyright assignment to the FSF on things like the compiler.

Bill

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Kevin Fox wrote:
>> 
>> The hardware isn't directly covered by the GPL, correct. But, if they
>> want to use the software on the hardware, they have to comply with the
>> GPL.

> Only with the GPLv3.

This is not true.  The terms of the GPLv2 that say you can't impose
further restrictions on the exercise of the freedoms apply to the
software under GPLv2 and GPLv3 just the same way.

> Do you like licenses that force the licensee to give money back?

> So why do you like licenses that force the licensee to give access to 
> hardware back?

I don't know where the 'back' in the second question amounts to, but
it definitely isn't about GPLv3.

In fact, the GPL isn't about giving anything back.  It's about passing
on.

So both requirements, as you phrased them, would be equally wrong.

So let's change the question to turn them into forms of passing on:

  Do you like licenses that force the licensee to pass money on?

  Do you like licenses that force the licensee to pass on access to
  hardware?

This is still bad.  This is still not what the GPLv3 is about.
There's no requirement to let the user go wild and do whatever she
likes on the hardware.

The only requirement is the one that was always there: to respect the
freedoms of the users of the software, i.e., let them modify and share
the software, not imposing any further restrictions, by whatever
means.

So the second question would be correctly phrased as 

  Do you like licenses that force the licensee to pass on the right to
  modify the software in the hardware containing it?

Or, reframing it:

  Do you like licenses that permit the licensee to deny others the
  right to modify the software in the hardware containing it?

> It's a form of "extra compensation" that the GPLv2 never had.

No, sir, it's still respect for the freedoms.  The same "in kind"
contribution as always.

> The GPLv2 talks about giving access to the *source* code.

It does.  But that's not all.  Even GPLv1 went further than that.

> Can people really not see the difference, and why I might think it's a 
> fundamental difference, and why I might choose to say that the GPLv3 is a 
> worse license?

Since someone brought liberal (Original BSD, Modified BSD, MIT, etc)
licenses into the picture, and you expressed dislike for them, let me
pick that up for a moment.

> The license doesn't encode my fundamental beliefs of "fairness".  I
> think the BSD license encourages a "everybody for himself"
> mentality, and doesn't encourage people to work together, and to
> merge.

And then you say what TiVO does is ok, saying:

> Oh, but you want to hack the hardware to accept it? That's a totally
> different issue. If so, buy a Neuros OSD box.

Sounds a lot like the very "everybody for himself" attitude you
dislike.

So can you please explain to me how enabling TiVO to deny others the
freedom that it received "in kind", failing to keep with the "in kind"
spirit of the GPL, encourage people to work together, and to merge?

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Diego Calleja wrote:
> 
> And the FSF is trying to control the design and licensing of hardware throught
> the influence of their software. And I think it's wrong. I'm all to forbid hardware
> that imposes restrictions on hardware, but software licenses are NOT the way
> to make it. That's a task for a "Free Hardware Foundation", not the FSF.

Amen. And btw, opencores.org does actually exist.

I don't even think open hardware is a big issue: the worry-warts about 
hardware are likely wrong, and hardware today is a lot more open than it 
used to be even just a decade ago. You can much more easily design your 
own (FPGA's are cheap and powerful), and yes, it's more complex today, but 
that's actually an argument _for_ openness rather than against it (open 
processes work better in complex environments!).

The real issue is "open content", and we do actually have various 
organizations that support that in particular. I would heartily encourage 
people to get involved with the Creative Commons, and the EFF, and I think 
Larry Lessig is a really smart and articulate person, who you should 
listen to.

> What the FSF is trying to do is EVIL.

I wouldn't go that far (although, in the heat of the moment I probably 
_have_ gone that far. Oops ;).

I don't think the FSF is evil. They're just too single-minded, and look 
too much at one issue, and only care about the one thing they care about, 
and in the process, they tend to have a really hard time seeing the other 
side of the coin.

They define "freedom" one way, and by defining it in a very particular 
way, they miss the fact that what is "freedom" to them is not "freedom" to 
somebody else.

They have a very particular agenda, and in having that agenda and a very 
strict view of how the world should look (according to the FSF), they 
dismiss the fact that other people have _other_ agenda's, and see the same 
world totally differently.

And I think that kind of single-mindedness is silly and 
counter-productive. 

I literally think that the GPLv2 has worked so well exactly because you 
can strip it of its high-falutin' morality and the FSF Kool-Aid, and just 
see it as a "tit-for-tat" license. It allows everybody to see that the 
work they put in (into the _software_) is protected, and people cannot 
make improved versions of that software and distribute those improved 
versions without giving you the right back to use those improvements (to 
the _software_).

So the GPLv2 may have come out of a very single-minded endeavor, but I 
think it ended up being capable of so much more than rms really even 
envisioned, exactly because you don't have to _view_ it in that manner.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, "Chris Friesen" <cfriesen@nortel.com> wrote:

> Alexandre Oliva wrote:
>> But see, I'm not talking about getting permission to hack the
>> hardware.  I'm only talking about getting permission to hack the Free
>> Software in it.

> No you're not...you're talking about being able to hack the software
> *and load it back onto the original hardware*.

Yes.  You wouldn't impose restrictions on modifying the software like
that, now would you?  Even though the GPL says you can't impose
further restrictions on modification and distribution.

>> It's your position that mingles the issues and permits people to use
>> the hardware to deprive users of freedom over the software that
>> they're entitled to have.

> The software license controls the software.  If the hardware has
> restrictions on it that limit what software it will run, then that is
> unrelated to the software license.

As in, the license controls the software.  If a patent creates
restrictions that limit what you can do with the software, then that
is unrelated to the software license.

As in, the license controls the software.  If a discriminatory
contract limits what you can do with the software, then that is
unrelated to the software license.

As in, the license controls the software.  If I send you the source
code, but it happens to be protected by a key that only the hardware
can decode, and it won't decode for you, then that is unrelated to the
software license.

Is that so, really?

> There is nothing stopping you from taking the code for the tivo,
> modifying it, distributing it, or even running it on other hardware.

True.  But TiVO is still imposing further restrictions on how I can
modify the software stored in their device, while reserving that
ability to itself.  This is wrong.  This is not "in kind".  This is
not "tit-for-tat".  Tit-for-tat is: if they can, then I can too, and
if I can't, then they can't either.

> Suppose I had some machine that will only run microsoft-signed
> binaries. Would it be at all related to any software license that this
> machine won't let me run linux?

That would be an unfortunate machine to have, but if Linux or some
other GPLed software was not shipped in it, then I don't see how this
is relevant to this discussion.  It's not about the hardware, it's
about the software in it, and about passing on the freedoms related
with it.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Alexandre Oliva wrote:
>> 
>> Let me see if I got your position right: when TiVO imposes
>> restrictions, that's ok

> Sure. I think it's ok that Microsoft imposes restrictions too on the 
> software they create. It's *their* choice.

Last I looked, TiVO was not the author of Linux.  Did you sell out or
something? ;-P :-D

> So tell me, why do you think I'm confused about the GPLv3?

I think you're confused about the spirit of the GPL, that applies
equally to v1, v2 and v3.

I think you're confused because you claim the GPL is tit-for-tat, that
it encourages/requires (you haven't been consistent) contributions in
kind, but the only contribution in kind is respect for the freedoms of
others.  But then, when measures are introduced to ensure compliance
with this twisted tit-for-tat notion, you claim they're wrong, that
they escape the spirit of the license.

This is why I think you're confused.


That said, it is possible that you disregarded the spirit of the GPL
entirely, focused on some of the legal terms and decided that was
something you wanted for your project.  And that it models what you
want for your project better than GPLv3 does, because GPLv3 takes
the spirit that you disregarded even more seriously than GPLv2.


I still fail to see why what it is in GPLv2 that makes it better to
satisfy your intentions WRT Linux than GPLv3.  I must assume that,
when you say "tit-for-tat", you mean something else, and not respect
for others' freedoms.  If you take the time to explain what it is,
then perhaps it will become clear why you consider the GPLv2 a better
license to achieve your goals, or perhaps it will show that you're
indeed confused about what GPLv2 and GPLv3 mean.

>  (a) Linus knows what he is doing, and isn't actually confused.

I can't say that yet.  Maybe after the points above are sufficiently
explored I will be able to say that.

>  (b) It was my right to use the license of my choice for a project that I 
>      started.

No doubt about it.

>  (c) I have the right to see the difference between the GPLv2 and v3, and 
>      think that the GPLv3 is the inferior license.

You sure do have that right.

> Comprende? MY CHOICE. Not the FSF's. Not yours. Not anybody elses.

Until you started accepting contributions from others, yes.


BTW, in Portuguese the correct spelling would be "compreende", with a
double 'e'.  "Comprende" is Spanish, and in Brazil, where I live, we
speak Portuguese.  But thanks for trying, that's appreciated ;-)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Alexandre Oliva wrote:
>> 
>> > In other words, Red Hat distributes copies (and yes, you *get* that copy), 
>> > and you cannot modify that copy that you got.
>> 
>> And Red Hat can't either.  I thought that was quite obvious.

> The GPLv2 talks about specific rights, like the ability to make changes 
> and distribute things, and says that you have to give downstream all those 
> same rights.

The spirit gives the intuition of "passing on all the rights".  The
legal terms have to be more careful about that, to avoid the very
situation you're debating, so they state "you can't impose further
restrictions on the exercise of the rights".

Do you understand the difference?

> For example, for any code that I have full copyright over, I have rights 
> that you DO NOT HAVE!

No dispute about that, and this is irrelevant to this point.  I've
already responded and clarified this point 2 or 3 times in this
thread.  Do you need me to find a URL for you?  It was in respose to
Dmitri Torokhov.

> So if you want to argue that I should re-license, you should argue that 
> the GPLv3 is better. And quite frankly, you haven't.

In fact, I haven't even tried.  So far, I've merely been trying to
show that it still follows the same spirit, dispelling the muth that
it doesn't, and trying to understand why you think GPLv2 is so much
better, which I think is related with tit-for-tat and retribution in
kind.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

RE: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[David Schwartz]

> The GPL applies to "the Program" which in this case is the Linux kernel
> as a whole and it in fact does indicate a specific version. All code
> submitted and included in this program has has been submitted with the
> understanding that the work as a whole is specifically licensed as
> GPLv2. Some authors have granted additional rights, such as dual BSD/GPL
> or GPLv2 and later and explicitly added such a notice.

Since the Linux kernel as a whole does not have a single author, it is
impossible to license it as a whole. Nobody has the authority to do that.
(The GPL is not a copyright assignment type license.)

Fortunately, the GPL clears this up:

"Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions.  You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
You are not responsible for enforcing compliance by third parties to
this License."

Linus cannot impose any further restrictions on the recipients' exercise of
the rights granted.

When you download a copy of the Linux kernel, you do not receive one license
because nobody could grant you one license. You receive a logically separate
license from each original licensor. You receive from Linus only a license
to his contributions.

Note that you cannot take a GPLv2+ work and redistribute it as GPLv3 only.
You can license your contributions as GPLv3 only of course. However, each
recipient still receives a GPLv2+ license to the parts that were originally
licensed that way. The people you distribute the work from receive licenses
from the original licensors to those parts, and you have no right to modify
that license. (See GPL section 6, quoted above.)

DS

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Valdis.Kletnieks@vt.edu wrote:

> On Thu, 14 Jun 2007 03:03:40 -0300, Alexandre Oliva said:
>> On Jun 14, 2007, Valdis.Kletnieks@vt.edu wrote:
>> 
>> > If a company sells you hardware that includes a ROM that contains GPL'ed
>> > software, are they in violation of the GPL if they don't include a ROM burner
>> > in the hardware?  Or are ROM burners like compilers, where you have to supply
>> > your own?
>> 
>> this requirement does not apply if neither you nor any third party
>> retains the ability to install modified object code on the User
>> Product (for example, the work has been installed in ROM).

> Do they have to provide a ROM burner if the ROM is socketed rather than
> soldered into place?

Of course not.  They just can't impose restrictions on your obtaining
a ROM burner and doing the work yourself.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Lennart Sorensen]

On Thu, Jun 14, 2007 at 07:48:03PM +0200, Rene Herman wrote:
> On 06/14/2007 06:01 PM, Linus Torvalds wrote:
> 
> >It's totally pointless to try to "force" people to be good. That's like 
> >"curing" gay people. Not going to happen. 
> 
> Tangent, but that could in fact quite easily be construed as saying that 
> gay people aren't good which I hope is not the point you are making :-/

I certainly read that as 'trying to force people to be good is just as
crazy as trying to force people to not be gay'.  Some people are good,
and some aren't (no idea why), and similarly some people are gay and
some aren't (again, no idea why).  Neither can be changed by declaring
that it must be changed.

I always love Linus' analogies. :)

--
Len Sorensen

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> They let you have the code and make changes to it,

Not to the software installed in the device.

What they do is like an author A who distributes a program to user B
under a non-Free Software license, and to user C under a Free Software
license.

C passes the program on to B under the same license.  Now B has two
copies of the program.  One is free, the other is not.

Except that TiVO had no right to distribute the program under non-Free
terms in the first place, because it was not the author, and the
license it had explicitly said it couldn't impose further
restrictions.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Chris Friesen]

Alexandre Oliva wrote:

> So can you please explain to me how enabling TiVO to deny others the
> freedom that it received "in kind", failing to keep with the "in kind"
> spirit of the GPL, encourage people to work together, and to merge?

They're not denying others the freedom that they themselves received. 
Tivo took GPL'd software, modified it, and distributed it with their own 
custom hardware.  You have the right to take their changes, possibly 
modify them further, and distribute them (possibly with your own hardware).

The fact that you can't modify the software and load it back onto the 
tivo is irrelevent.  They are not restricting your distribution of the 
software in any way.  Rather, they're restricting the *running* of the 
software on their proprietary hardware platform.

Chris

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Alexandre Oliva wrote:
>> 
>> Is there anything other than TiVOization to justify these statements?

> Do you need anything else?

No, I'm quite happy that this is all.

> But if by the question you mean "would you think the GPLv3 is fine without 
> the new language in section 6 about the 'consumer devices'", then the 
> answer is that yes, I think that the current GPLv3 draft looks fine apart 
> from that.

Then would you consider relicensing Linux under GPLv3 + additional
permission for Tivoization?

>> Also, can you elaborate on what you mean about 'giving back in kind'?
>> (I suspect this is related with the tit-for-tat reasoning, that you've
>> failed to elaborate on before)

> I've *not* failed to elaborate on that before. Not at all. 

> Just google for

> 	torvalds tit-for-tat

> and you'll see a lot of my previous postings. Trying to claim that this is 
> somehow "new" is ludicrous.

I didn't.  But I've provided evidence that your prior musings on this
topic were wrong.  I wanted to give you an opportunity to review your
position under this new light.  I see you haven't changed it at all.

> Giving back "in kind" is obvious. I give you source code to do with as you 
> see fit. I just expect you to give back in kind: source code for me to do 
> with as I see fit, under the same license I gave you source code.

> How hard is that to accept?

Forgive me if I find this a bit hard, because that's *not* what the
GPL says.

Where do you think the GPL say that you get the source code back?

> I don't ask for money. I don't ask for sexual favors. I don't ask for 
> access to the hardware you design and sell. I just ask for the thing I 
> gave you: source code that I can use myself.

See, that's not what the license says.

The license says what you ask for is respect for other users'
freedoms.  Nothing whatsoever for you.  Only for users.

Freedom is in "in kind" payment, and it's not even a retribution, a
payback: it's payforward, or paysideways.

Do you understand why I find your reasoning hard to accept?

> And no, it's not a new concept. Neither is the fact that I've never agreed 
> with the FSF's agenda about "freedom" (as defined by _them_ - I have a 
> notion of "freedom" myself, and the FSF doesn't get to define it for me).

We don't have to agree on our individual definitions of freedom.  But
we're talking about a specific license that assigns a specific meaning
to the term "freedoms", and that's all this is about.

> I don't call Linux "Free Software". I haven't called it that for close to 
> ten years! Because I think the term "Open Source" is a lot better.

I can appreciate that you think it's better, but unfortunately it
appears to be playing a significant role in confusing your
interpretation of the GPL.  The GPL is not just about making the
source code visible, or even modifyable by others.  It's about
respecting others' freedoms.  No matter how badly you prefer Open
Source over Free Software, how badly you'd rather disregard the
freedoms in the spirit and in the legal terms of the GPL, you chose a
license designed to protect those freedoms, not only the ability to
see and modify source code.

>> The only thing the GPL demands is respect for others' freedoms, as in,
>> "I, the author, respect your freedoms, so you, the licensee, must
>> respect others' freedoms as well".  Is this the "in kind" you're
>> talking about?  Or are you mistaken about the actual meaning of even
>> GPLv2?

> I just ask that you give the software back in a usable form. That's
> all I ask for.

I'm afraid that's not what the GPLv2 says.  There's no provision
whatsoever about giving anything back.  Not in the spirit, not in the
legal terms.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> On Thu, Jun 14, 2007 at 02:26:30PM -0300, Alexandre Oliva wrote:
>> In the program you received under GPLv1.
>> 
>> Hey, you said there was code under GPLv1.1 in the Linux tree.  Then,
>> there should be a copy of GPLv1.1 in there, otherwise AFAICT the
>> distribution of that code is copyright infringement.  IANAL.

> So now the copy of the GPL v2 isn't good enough for the GPLv1.1 code?

If it exists and it's 1.1-only, I believe it wouldn't, but IANAL.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 11:20:34 Adrian Bunk wrote:
> On Thu, Jun 14, 2007 at 12:00:17AM -0400, Valdis.Kletnieks@vt.edu wrote:
> > On Thu, 14 Jun 2007 04:56:40 +0200, Adrian Bunk said:
> > > Reality check:
> > >
> > > Harald convinced companies that they have to provide the private keys
> > > required to run the Linux kernel they ship on their hardware.
> >
> > No, the *real* reality check:
> >
> > The operative words here are "convinced companies" - as opposed to
> > "convinced a judge to rule that private keys are required to be
> > disclosed". (I just checked around on gpl-violations.org, and I don't see
> > any news items that say they actually generated citable case law on the
> > topic of keys...)
> >
> > Harald convinced companies that it was easier/cheaper/faster to provide
> > the private keys than to continue in a long legal battle with an
> > uncertain outcome. If the company estimates the total loss due to keys
> > being released is US$100K, but the costs of taking it to court are
> > estimated at US$200K, it's obviously a win (lesser loss, actually) for
> > the company to just fold.
> >...
>
> Here in Germany, the rules at court are roughly "the loser pays
> everything including the costs of the winner", so if a big company is
> sure they will win at court there's no reason not to go there.
>
> And if they did the effort of using private keys to only allow running
> an official firmware, they must have seen an advantage from doing so.
>
> I'm not saying it legally clear the other way round, my statement was
> an answer to Daniel's emails claiming it was clear what such companies
> do was legal.

I'm sorry if I gave anyone that impression. My point was that it would be 
pointless to argue the case in the US because here it really is, 
usually , "buy the best justice for the money".

DRH

> cu
> Adrian



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Greg KH <greg@kroah.com> wrote:

> The FSF required copyright assignment to themselves in order to accept
> the changes from the developers.

For many strategic projects, but not all of them.

> So the FSF owns the whole copyright and can change things whenever
> they want, to whatever license they want.

This is not true.  Have you ever read the copyright assignment
contract?  It very clearly constrains the ways the FSF can release the
code.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Sam Ravnborg]

On Thu, Jun 14, 2007 at 04:46:36PM -0300, Alexandre Oliva wrote:

> > Giving back "in kind" is obvious. I give you source code to do with as you 
> > see fit. I just expect you to give back in kind: source code for me to do 
> > with as I see fit, under the same license I gave you source code.
> 
> > How hard is that to accept?
> 
> Forgive me if I find this a bit hard, because that's *not* what the
> GPL says.

What part of the word "expect" did you not understand?
And whats your point here anyway?

	Sam

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Thu, 14 Jun 2007, Diego Calleja wrote:

>> And the FSF is trying to control the design and licensing of
>> hardware throught the influence of their software.

It's not.  It's only working to ensure recipients of the Free Software
can modify and share the software.

>> What the FSF is trying to do is EVIL.

> I wouldn't go that far (although, in the heat of the moment I probably 
> _have_ gone that far. Oops ;).

:-)

> I literally think that the GPLv2 has worked so well exactly because you 
> can strip it of its high-falutin' morality and the FSF Kool-Aid, and just 
> see it as a "tit-for-tat" license. It allows everybody to see that the 
> work they put in (into the _software_) is protected, and people cannot 
> make improved versions of that software and distribute those improved 
> versions without giving you the right back to use those improvements (to 
> the _software_).

Can you explain to me how it is that the Tivoization provisions (the
only objection you have to GPLv3) conflict with this?

(nevermind our disagreement as to whether "tit-for-tat" applies to
 either GPLv2 or GPLv3)

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Ingo Molnar]

* Alexandre Oliva <aoliva@redhat.com> wrote:

> On Jun 14, 2007, Ingo Molnar <mingo@elte.hu> wrote:
> 
> > I think the proper limit is the boundary where the limit of the 
> > software is - because that's the only sane and globally workable way 
> > to stop the power-hungry.
> 
> But see, I'm not talking about getting permission to hack the 
> hardware.  I'm only talking about getting permission to hack the Free 
> Software in it.
> 
> It's your position that mingles the issues and permits people to use 
> the hardware to deprive users of freedom over the software that 
> they're entitled to have.

where does this false sense of entitlement come from? The hardware maker
ows you nothing but what is written into the GPLv2. Not more, not less.

(In fact, most hardware makers that utilize free software today give
back _substantially more_ to the community than the license requires!
For example they are currently the largest employers of free software
developers - although nothing in the license forces them to do so. Why?
Because the economic rules that the GPLv2 creates are healthy.)

you are not "entitled" to dictate the hardware's design (or any other 
copyrighted work's design), even if the license gives you the power to 
do so. By your argument we'd have to put the following items into the 
license too:

 - free on-site training for free software developers about the 
   hardware's inner workings. (It is justified to teach free software
   the same know-how as in-house engineers of the hardware maker. 
   Without this, users are hindered in their freedom to use and 
   effectively modify (fix) the software.)

 - free access to all the hardware diagnostics tools that the hardware 
   maker has. (Without that it might be impossible to modify the 
   software as efficiently as the hardware maker's own engineers can do 
   it.)

 - free samples of the hardware to be sent to free software developers,
   upon request. (The hardware maker's own engineers have free access to 
   samples. Otherwise free software users might not get the same level 
   of driver support as the hardware maker can achieve.)

 - free access to the hardware manufacturing equipment. (If i wish to 
   modify the free software in a way that requires more RAM than the 
   hardware has, i need access to the manufacturing equipment to produce 
   a new version of the hardware that can run that free software. The 
   hardware maker has this right and flexibility to modify the software, 
   so i should have that same right too.)

see how quickly your argument becomes totally ludicrous, if brought to 
its logical conclusion?

This "right to modify" and "have the same rights as the hardware maker" 
arguments are _totally_ bogus, they were made up after the fact, just 
because quite apparently RMS had a fit over Tivo and started this verbal 
(and legal) vendetta. The FSF is now attempting to rewrite history and 
pretends that this "always was in the GPLv2" and applies this newly 
thought up concept to the GPLv3 in a way that substantially departs from 
the spirit of the GPLv2. Which spirit the GPLv2 explicitly promised to 
uphold in Section 9. Which could make any contrary section of the GPLv3 
unenforceable, when applied to "GPLv2 or later" licensed software.

	Ingo

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 12:06:31 Kevin Fox wrote:
> On Wed, 2007-06-13 at 20:42 -0400, Daniel Hazelton wrote:
> <SNIP>
>
> > > Do you deny that TiVo prevents you (or at least a random customer)
> > > from modifying the copy of Linux that they ship in their DVR?
> >
> > Exactly. They don't. What TiVO prevents is using that modified version on
> > their hardware. And they have that right, because the Hardware *ISN'T*
> > covered by the GPL.
>
> The hardware isn't directly covered by the GPL, correct. But, if they
> want to use the software on the hardware, they have to comply with the
> GPL. The software license can then influence hardware IF they want to
> use it badly enough.

Until GPLv3 there was no requirement that the modified code be able to operate 
on any given device - even the one its designed for. Claiming otherwise seems 
stupid to me, almost ridiculous.

> For example, the hardware is perfectly capable of being used to break
> the terms of the GPL by being used to distribute a modified binary
> without releasing the source. But the hardware's behavior is restricted
> by the software for the betterment of all.

But this can be done *NOW* - has been done by at least one company, IIRC. 
(and, IIRC again, they didn't so much as "not release the modified version" 
as "not release all tools/scripts/installation/build instructions")

> This whole argument is about the spirit of the GPL. Linus and others
> think the spirit is one thing, the FSF guys think its something else.
> Since the license is clearly owned by the FSF, I think they get the
> final vote on what they "intended" it to be when they wrote it, no? If
> they say they intended it to not allow Tivoization then believe them,
> because they are the only ones that know what they were thinking when
> they wrote it! The GPLv2 seems to allow it though. If Linus and friends
> want to allow it, then they can stay with the GPLv2. For those who want
> to disallow Tivoization, choose v3. No worries guys.
>
> > Do you understand that, or do I need to break out the finger-puppets next
> > ?
>
> Guys, we are all friends here. No reason to be so insulting. Its just a
> difference of opinion. People seem to be talking past each other instead
> of to one another. This usually happens when people are basing their
> underlying assumptions on different things and not listening to the
> other. Please take a step back and think about it.

I noticed that ten or twenty messages after I made that comment. In truth, the 
reason I made it was, and is, because I am tired of explaining the fact that 
there is no "one" interpretation of the GPLv2 - or any license - *UNLESS* it 
has been ruled on by a court. And even then, the courts ruling only applies 
to the parts of the license that were in contention before it.

DRH

> <SNIP>



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, "Dmitry Torokhov" <dmitry.torokhov@gmail.com> wrote:

> Ok, consider non-derived work.

I did, you snipped it out:

>> If your change is not a derived work, you're not bound by the terms
>> of the GPL as far as the change is concerned, so the GPL has no say
>> whatsoever as to how you must release it.  If you choose the GPL,
>> then you're a licensor, and the requirements to pass on all the
>> rights you have do not apply.

> Because I am distributing whole program
> I have to do it under GPL. However I still have the right to
> distribute just the portion that is written by me under whatevel
> license I want but you as a recepient of GPLed whole do not get this
> right. IOW I am not passing all the rights _I have_.

I see what you mean.  IANAL, but I don't think that's how it works.

When your work is not a derived work, the GPL that applies to the rest
of the program does not make you a licensee, and it only covers your
work if you choose to license it that way.  And then, you're the sole
licensor of that piece of the work.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> > Only with the GPLv3.
> 
> This is not true.  The terms of the GPLv2 that say you can't impose
> further restrictions on the exercise of the freedoms apply to the
> software under GPLv2 and GPLv3 just the same way.

The GPLv2 talks *only* about the software. You're making everything else 
up, and when I point out that your reading of the GPLv2 is insane, you 
just ignore my proofs of your internal inconsistency.

> So can you please explain to me how enabling TiVO to deny others the
> freedom that it received "in kind", failing to keep with the "in kind"
> spirit of the GPL, encourage people to work together, and to merge?

Because Tivo *IS NOT DENYING* those freedoms.

Tivo *respected* the freedoms, and gave source back, and gave you all the 
same rights you had to Linux originally, and to their modifications.

How stupid are you to not acknowledge that?

Tivo limited their *hardware*, not the software.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> Last I looked, TiVO was not the author of Linux.  Did you sell out or
> something? ;-P :-D

You're a moron.

I'm the original author, and I selected the GPLv2 for Linux.

Tivo accepted that, and followed the GPLv2. Even the FSF lawyers agreed 
that they followed it.

But Tivo *is* the "author" of their own proprietary applications, and it 
*is* the designer of their hardware.

And exactly like I had the right to the choice of license when it comes to 
Linux, they have the right of choice to license and behaviour when it 
comes to *their* software and hardware (that is not a derived work of 
Linux).

But you cannot follow a coherent argument, because you dont' *want* to 
follow it. Because following the logical argument would take you to a 
place where you don't want to be.

I'm not going to bother discussing this any more. You don't seem to 
respect my right to choose the license for my own code.

So one final time:

 - I chose the GPLv2, fully understanding that the Tivo kind of situation 
   is ok.

 - the FSF lawyers too have acknowledged that what Tivo did was not a 
   license violation, so I obviously am not confused about the issue: YOU 
   are.

 - I think that what Tivo did was not only "technically valid" by the 
   license, it was what I *intended* by my choice of license!

And you are apparently totally unable to understand - or respect - that I 
actually made an informed decision that happens to be different from what 
you *wish* it were.

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> Then would you consider relicensing Linux under GPLv3 + additional
> permission for Tivoization?

No. I'm not stupid.

The GPLv3 explicitly allows removing additional permissions.

So anybody who does "GPLv3 + additional permissions" is basically setting 
himself up for people taking those permissions away.

Since the Tivo kind of permission is in my opinion a *fundamental right* 
(or call if "freedom" if you want), then "GPLv3 + additional permissions" 
simply is not a viable alternative, since anybody could just decide to 
make improvements and strip those permissions.

The whole notion of "additional permissions" in the GPLv3 is totally 
pointless, since it's legally *exactly* the same as allowing dual 
licensing (which a license doesn't even have to spell out: you can 
dual-license *regardless* of the license!).

The reason for the "additional permissions" is just to make the LGPL go 
away, and become a sub-clause of the GPLv3.

If you really thought anything else, you're just uninformed and stupid, 
and didn't think things through.

			Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dmitry Torokhov]

On 6/14/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 14, 2007, "Dmitry Torokhov" <dmitry.torokhov@gmail.com> wrote:
>
> > Ok, consider non-derived work.
>
> I did, you snipped it out:
>
> >> If your change is not a derived work, you're not bound by the terms
> >> of the GPL as far as the change is concerned, so the GPL has no say
> >> whatsoever as to how you must release it.  If you choose the GPL,
> >> then you're a licensor, and the requirements to pass on all the
> >> rights you have do not apply.
>

Yes, I did, thank you for putting the text back in.

> > Because I am distributing whole program
> > I have to do it under GPL.

Please notice this sentence. GPL still influences the way I release
stuff (if I want to release the work as whole) but it does not mean
passing all rigths I could possibly have.

> > However I still have the right to
> > distribute just the portion that is written by me under whatevel
> > license I want but you as a recepient of GPLed whole do not get this
> > right. IOW I am not passing all the rights _I have_.
>
> I see what you mean.  IANAL, but I don't think that's how it works.
>
> When your work is not a derived work, the GPL that applies to the rest
> of the program does not make you a licensee, and it only covers your
> work if you choose to license it that way.  And then, you're the sole
> licensor of that piece of the work.

So, with regard to TIVO, why are you saying that GPL shoudl affect
their hardware (I assume that key check/enforce is done in firmware
taht is separate from kernel image)?

-- 
Dmitry

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Linus Torvalds]

On Thu, 14 Jun 2007, Alexandre Oliva wrote:
> 
> I see what you mean.  IANAL, but I don't think that's how it works.

Why the hell do you keep saying that?

There *are* lawyers who have said that what Tivo did was legal. They were 
the FSF's own lawyers. So now you're saying "I am not a lawyer, but that's 
not right".

So you're trying to state some legal point, admitting that you're not a 
lawyer, and admitting that actual real-life lawyers disagree with you?

So please explain why the *hell* you would expect us to take your points 
seriously?

		Linus

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Dave Neuer]

On 6/14/07, Lennart Sorensen <lsorense@csclub.uwaterloo.ca> wrote:
> Nothing prevents you from taking tivos kernel
> changes and building your own hardware to run that code on, and as such
> the spirit of the GPL v2 seems fulfilled.

Oh, come on: you're not serious, right? Something indeed prevents me
-- the fact that I'm not a hardware manufacturer, I don't have fabs,
outsource vendors to provide me w/ designs, ASICs, etc. Nor to I have
the money to pay one-off prices for various components if they're even
available in batches that small.

This argument seems totally disingenuous to me. The GPLv<3 was written
in a time when the majority of sotware to which the license was
applied was written for general purpose computers. The "user" was the
owner of the computer, and Freedom 0 was about letting that user RUN
modified copies of the software.

Things have changed a lot; we're surrounded by embedded computers, and
Freedom 0 seems to strongly imply I should have the right to run
modified versions of the Free Software I own on the hardware I OWN. Or
is the future of Open Source that you'll be able to hack on free
software as long as you work for Intel, Red Hat, TiVO, Google or OSDL?
Or own many-thousand-$$ fab printer?

Look, I totally respect Linus' and others' position that the license
is an inappropriate way to enforce what they feel are hardware design
decisions, but can we dispense w/ the silly argument that the intent
of the GPL is fullfilled as long as the user is allowed to modify the
software where modify means "imagine a world where they'd be able to
run" it?

Dave

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Florin Malita]

On 06/14/2007 02:27 PM, Alexandre Oliva wrote:
>> No, by this twisted logic Tivo *cannot* modify that particular copy
>> any more than you can. They can modify *another* copy (just like you)
>> and they can *replace* the copy in your device with the new version
>> (unlike you).
>>     
>
> Again, replacing is one form of modification.
>   

No, it's not: replacing does not create derivative work. Modification does.

You've chosen to attach a physical dimension to "program copy" and I'm 
arguing that even under this distorted line of reasoning you can't 
support your position:

> The customer gets the copy that TiVO stored in the hard disk in
> the device it sells.  And it's that copy that the customer is entitled
> to modify because TiVO is still able to modify it.

* Tivo takes public sources, modifies them and builds a brand new blob
* Tivo installs this new copy on the device, most likely side-by-side 
with the old one - notice how the new copy is derived from public 
sources and has absolutely nothing to do with the old version (heck, it 
can be a totally different kernel for what it's worth)
* Tivo deletes the old copy from the device

It seems pretty obvious that the only right Tivo is withholding is the 
right to install new versions on the device - they never do (and really 
never could) "modify" the physical copy on your device (which is your 
main argument).


> What do you think you do when you save a modified source file in your
> editor?

Don't skip the part where the in-memory version started as an exact copy 
of the original being replaced. Notice the difference? ;)

---
fm

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Al Viro]

On Thu, Jun 14, 2007 at 09:55:17PM +0200, Ingo Molnar wrote:
> This "right to modify" and "have the same rights as the hardware maker" 
> arguments are _totally_ bogus, they were made up after the fact, just 
> because quite apparently RMS had a fit over Tivo and started this verbal 
> (and legal) vendetta. The FSF is now attempting to rewrite history and 
> pretends that this "always was in the GPLv2" and applies this newly 
> thought up concept to the GPLv3 in a way that substantially departs from 
> the spirit of the GPLv2. Which spirit the GPLv2 explicitly promised to 
> uphold in Section 9. Which could make any contrary section of the GPLv3 
> unenforceable, when applied to "GPLv2 or later" licensed software.

That, BTW, is perhaps the worst problem with v2 (inherited by v3).
WTF _is_ "the spirit of the license" and who gets to decide if two
licenses are in the same spirit?  As soon as we get to "well, original
authors of the license are the final authority on that", we are
in the "I've always said ..." country.

Look, humans _suck_ at revision control, especially that of our
intentions and opinions.  It doesn't even require malice, all
ancedotes about spouses/mothers-in-law/etc. nonwithstanding.

We all easily fall into belief that we had always meant what we mean
now; that even if we said something different, it was just a poor
wording; that if we had known what we know now, we would certainly
had come to the same conclusions we have come to now.

"In the same spirit" is just about the weakest requirement in that
area.  I.e. the most prone to drift, especially when one is an ideologist
and thus has severely decayed integrity to start with.  Call it a
professional disease of crystal ball users - or a prerequisite for
playing a visionary, if you will ;-/

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Daniel Hazelton]

On Thursday 14 June 2007 13:26:30 Alexandre Oliva wrote:
> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Thursday 14 June 2007 03:11:45 Alexandre Oliva wrote:
> >> On Jun 14, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> >> > Ah, well... In the case of "Windos" and other proprietary OS's I try
> >> > to educate people and get them to switch.
> >>
> >> Good.  So I presume you'd tell them to switch away from a
> >> turned-proprietary GNU/Linux operating system as well, right?
> >
> > If that happened I'd be lost. I've tried the various BSD's and found they
> > had problems with hardware support and getting a new version of the BSD
> > kernel to compile and boot is something of a black art.
> >
> > The point is moot, though. It can never happen.
>
> Look again, it's already happened in the TiVO and other devices.
>
> The software that ships in them is no longer Free Software.
>

In *YOUR* opinion and by *YOUR* definition of the term. Yes, I have seen some 
evidence that TiVO hasn't made some of the modifications they made public - 
doesn't mean that they won't, just it hasn't *YET* been done. (Not that I'm 
so omniscient I can say, definitively, whether they will or won't - or even 
that they haven't done it already).

By my own definition replacement != modification.

>
> Consider a new microprocessor.
>
> Consider that Linux is ported to it by the microprocessor
> manufacturer.
>
> Consider that the manufacturer only sells devices with that
> microprocessor with TiVO-like locks.
>
> How exactly can you enjoy the freedoms WRT the GPLed software you got
> from the manufacturer?

The same as I would with a TiVO. I have the right to copy, modify, distribute 
and run the code - even if I can't do any of those things on the hardware the 
original binary operates on.

>
>
> Now consider that you have a single computer, and that's built by TiVO.
>
> How exactly can you enjoy the freedoms the author meant you to have,
> if the TiVO box won't run the program after you modify it?

Simple: I don't buy it. Each and every piece of hardware I buy has a rather 
laborious research process before I actually spend the money on it. This 
makes it a certainty that I can use the hardware in the manner I want without 
problems like your hypothetical.

Whats worse - forcing your morals and ideals on someone or giving them the 
same freedom of choice you had?

Before you answer remember that that is *EXACTLY* what is being done with 
GPLv3. With GPLv2 and prior there was a simple guarantee that 
every "Licensee" had exactly the same rights. With GPLv3 you are forcing your 
ethics and morals on people - and isn't this exactly what the Roman Catholic 
church did during the Spanish Inquisition?

> > If this "run modified copies on the same hardware you received the
> > original on" *IS* the "spirit" of the license, then why isn't it
> > stated anywhere before GPLv3?
>
> For the same reasons that the pro-DRM laws weren't mentioned before,
> and the patent retaliation clauses weren't mention before: these
> specific cases hadn't been studied, only the general idea of
> respecting users' freedoms was.

Bzzt! Wrong! The reason is that it wasn't necessary - at all. It still isn't, 
but a group that feels modification == replacement wants it to be, so it has 
suddenly become necessary. (Note that anti-DRM stuff *IS* good - DRM is part 
of an attempt by failing business models to stop the failure)

> > I'll grant you that. But, at this point, where can I find a copy of
> > the GPLv1 without having to dig around the net ?
>
> In the program you received under GPLv1.
>
> Hey, you said there was code under GPLv1.1 in the Linux tree.  Then,
> there should be a copy of GPLv1.1 in there, otherwise AFAICT the
> distribution of that code is copyright infringement.  IANAL.

Ah, but I never said I had a GPLv1 program. If GPLv1 is still valid and 
available I should be able to find a copy of it *RIGHT* *NOW* to license a 
new project if I want to use GPLv1 as its license. So your logic is again 
flawed.

> >> In contrast, your TiVO may get a software upgrade without your
> >> permission that will take your rights away from that point on, and
> >> there's very little you can do about it, other than unplugging it from
> >> the network to avoid the upgrade if it's not too late already.
> >
> > And because its a device that connects to their network - and TiVO
> > isn't a telecommunications company - they have the right to upgrade
> > and configure the software inside however they want. (In the US at
> > least)
>
> But do they have the right to not pass this right on, under the GPL?

Yes, they do. It isn't a right they have as "copyright holders" - in fact, it 
isn't a part of their rights under the copyright at all. It's a part of their 
rights as the owners of the network. 

> >> > A lot of them would probably have private modifications that would
> >> > never be distributed - and under the GPLv2 it is clear that you can
> >> > keep modifications private as long as you don't distribute them.
> >>
> >> Likewise with GPLv3.
> >
> > I can see this, but will a company see this?
>
> In what sense does the GPLv3 make this particular point any less
> obscure?

Never claimed it was less obscure, just that you've usually got a board-room 
filled with middle-aged men that might have problems agreeing that it is a 
clear-cut case.

> > True. But that doesn't save them from lawsuits trying to force them
> > to obey the terms of the new revision even though they received the
> > software under an earlier version.
>
> Nothing saves anyone from silly lawsuits.  This one would likely be
> laughed out of court in no time.  Anyone worried about this should
> also be concerned about their neighbor suing them for copyright
> infringment every time they set their stereo loud enough for the
> neighbors to listen and be annoyed.  (Hint: only the copyright holder
> would stand a chance of winning such a lawsuit)

Yes, but the fact that it would cost money to get the suit dropped is a 
problem. 

> >> > (and don't try to argue that even though those modifications are
> >> > truly private (to the company) they should be released anyway to
> >> > comply with the "spirit" of the license. It is made clear that it
> >> > isn't by the text of the license itself)
> >>
> >> How could you possibly come to the conclusion that forcing anyone to
> >> release private modifications would be in compliance with the spirit
> >> of the license?  can != must
> >
> > I was trying to be sarcastic and inject a little humor here. Guess I
> > should have used the old <sarcasm> tag :)
>
> Aah.  I'm not sure I'd have understood it either.
>
> >> >> > Why should I repeat Linus' explanation of the ways that GPLv3
> >> >> > violates the spirit of GPLv2?
> >> >>
> >> >> Don't worry about parrotting here, he hasn't provided that
> >> >> explanation yet ;-)  Please give it a try.
> >> >
> >> > But he has. Whether you have accepted that his explanations are
> >> > valid or not doesn't change the fact.
> >>
> >> His explanation is based on a reading of the license that doesn't
> >> match what its authors meant.  I guess the authors know better what
> >> they meant the spirit of the license to be than someone else who
> >> studied it a lot but that until very recently couldn't even tell the
> >> spirit from the legal terms.
> >
> > And his interpretation is no less valid than that of anyone else. In
> > fact, after a recent conversation with a couple of lawyers that I
> > know, I can state that his interpretation isn't that far off from
> > theirs.
>
> Interpretation as applied to the legal terms, yes.  As for the spirit
> of the license, the authors ought to know better than anyone else what
> they meant.  Sure, other interpretations might lead to different
> understandings as to what the readers *think* it means, but that
> doesn't change what it was *intended* to mean.

Doesn't matter what the author intended it to mean - at all. What matters is 
how its interpreted when/if it shows up in court. This can be seen *ALL* the 
time with laws across the globe. In the US there is this thing called 
the "RICO" Laws- "Racketeering Influenced Criminal Organization" - that give 
the government the ability to seize anything that is deemed a "profit of drug 
sales" or of a "Criminal Organization". It has been used for that purpose, 
but its interpretation has caused it to be used to seize money that has had 
no source in either.

> > Then you're lucky. I've had a lot of people say something similar to the
> > following: "Oh, I've heard about that. So which version of the GNU-Linux
> > kernel are you running?"
>
> Oh my.  That's indeed unfortunate and unfair.
>
> > As I've stated before - I can find nothing in the history of the GPL or
> > the FSF that makes the "on the same hardware" requirement clear and part
> > of the "spirit" of "Free Software".
>
> Put the considerations above, about a single computer or a
> uniformly-limited computing platform, and you'll see that this "on the
> same hardware" argument is just a means to deny people freedom.  If I
> could stop you from running modified versions on one piece of
> hardware, then I could on two, and 3, and then soon it's all of them,
> and we're back to square zero in terms of freedom.
>
> >> > What I won't do is release whatever tools and such that are needed to
> >> > make the hardware run a different version of the kernel. Why? Because:
> >> > the hardware was designed so that a specific version of the kernel
> >> > runs without problems, there is hardware that is very picky and
> >> > running a customized kernel could cause that hardware to fail, etc...
> >>
> >> Why do you care?  It's no longer your hardware, it's theirs.
> >
> > Legal requirements in some countries that require manufacturers to
> > provide support for their product for a period of time after it has been
> > purchased.
>
> If you replace a component in the hardware, are you still required to
> provide support or offer warranty?  Why should this be different just
> because it's a software component?

Artificial distinctions in the law

DRH

-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Alexandre Oliva]

On Jun 14, 2007, Sam Ravnborg <sam@ravnborg.org> wrote:

> On Thu, Jun 14, 2007 at 04:46:36PM -0300, Alexandre Oliva wrote:
>> > Giving back "in kind" is obvious. I give you source code to do with as you 
>> > see fit. I just expect you to give back in kind: source code for me to do 
>> > with as I see fit, under the same license I gave you source code.
>> 
>> > How hard is that to accept?
>> 
>> Forgive me if I find this a bit hard, because that's *not* what the
>> GPL says.

> What part of the word "expect" did you not understand?

http://lkml.org/lkml/2006/9/24/246

  It asks everybody - regardless of circumstance - for the same thing.
  It asks for the effort that was put into improving the software to
  be given back to the common good.  You can use the end result any
  way you want (and if you want to use it for "bad" things, be my
  guest), but we ask the same exact thing of everybody - give your
  modifications back.

> And whats your point here anyway?

The the GPL doesn't do that.  It encourages that.  But what it asks
for is respect for the freedoms it defends WRT the software licensed
under it.

-- 
Alexandre Oliva         http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member         http://www.fsfla.org/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}

Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[Rene Herman]

On 06/14/2007 09:29 PM, Lennart Sorensen wrote:

> On Thu, Jun 14, 2007 at 07:48:03PM +0200, Rene Herman wrote:
>> On 06/14/2007 06:01 PM, Linus Torvalds wrote:
>>
>>> It's totally pointless to try to "force" people to be good. That's like 
>>> "curing" gay people. Not going to happen.
>> 
>> Tangent, but that could in fact quite easily be construed as saying
>> that gay people aren't good which I hope is not the point you are
>> making :-/
> 
> I certainly read that as 'trying to force people to be good is just as
> crazy as trying to force people to not be gay'.  Some people are good,
> and some aren't (no idea why), and similarly some people are gay and
> some aren't (again, no idea why).  Neither can be changed by declaring
> that it must be changed.

Yes, just my sense of humour, I'm afraid... ;-)

Rene.

RE: Dual-Licensing Linux Kernel with GPL V2 and GPL V3

[David Schwartz]

> What about if your GPL program ends up in a piece of hardware
> (e.g. a ROM,
> or an embedded ROM, or if it's some GPL code from OpenCores, as gate
> netlist in silicon)? My interpretation is that you need a permission from
> the author for doing that, unless there's an easy way to replace
> it with a
> modified copy (e.g. if you put the OpenCores stuff into an FPGA,
> replacing
> the configuration PROM would do it).

The GPL does not require it to be easy in fact to modify the piece of
software. It just requires that you have the right to modify it, that is,
that there be no legal obstacles in your way. You are entitled to the source
code in modifiable, understandable form. There are no legal restrictions,
other than those in the GPL and in the law, on what you can do with it.

What you are actually *able* to do, however, depends upon a wide variety of
factors way outside the scope of the GPL.

By the way, I have a lot of sympathy for the argument that *if* you provide
me a binary made from GPL'd code that required a key to produce that binary,
I am entitled to that key. The key is precisely analogous to any other piece
of source code -- it is mathematically 'combined' and 'processed' by tools
to produce the final, distributed executable. If there's some rational basis
for a legal difference between a signing key and a header file, I don't know
what it is.

DS