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

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

[Al Boldi]

Alexandre Oliva wrote:
> On Jun 17, 2007, "Gabor Czigola" <czigola@gmail.com> wrote:
> > I wonder why the linux kernel development community couldn't propose
> > an own GPL draft (say v2.2) that is "as free as v2" and that includes
> > some ideas (from v3) that are considered as good (free, innovative, in
> > the spirit of whatever etc.) by the majority of the kernel developers.
>
> For one, because the text of the GPL is copyrighted by the FSF, and
> licensed without permission for modification.  And that's as it should
> be, you don't want others to modify the terms of the license you chose
> for your code, do you?

Wow!

Under what circumstances would it be possible to receive permission for 
modification?


Thanks for being GPL!

--
Al

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

[Alexandre Oliva]

On Jun 17, 2007, Al Boldi <a1426z@gawab.com> wrote:

> Under what circumstances would it be possible to receive permission for 
> modification?

You have to ask the copyright holder.

Affero did just that, and so the Affero GPL was born.

Just don't assume the FSF will grant such permissions lightly.

-- 
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 Boldi]

Alexandre Oliva wrote:
> On Jun 17, 2007, Al Boldi <a1426z@gawab.com> wrote:
> > Under what circumstances would it be possible to receive permission for
> > modification?
>
> You have to ask the copyright holder.
>
> Affero did just that, and so the Affero GPL was born.
>
> Just don't assume the FSF will grant such permissions lightly.

But wouldn't this be against the spirit of the GPL?


Thanks!

--
Al

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

[Alexandre Oliva]

On Jun 17, 2007, Al Boldi <a1426z@gawab.com> wrote:

> Alexandre Oliva wrote:
>> On Jun 17, 2007, Al Boldi <a1426z@gawab.com> wrote:
>> > Under what circumstances would it be possible to receive permission for
>> > modification?
>> 
>> You have to ask the copyright holder.
>> 
>> Affero did just that, and so the Affero GPL was born.
>> 
>> Just don't assume the FSF will grant such permissions lightly.

> But wouldn't this be against the spirit of the GPL?

I'm not sure what you mean.

The moment you modified it, it would no longer be the GPL and, as long
as it's not the FSF publishing it as later version of the GPL, it
wouldn't be bound by the public commitment that all versions of the
GPL would have the same spirit, and as long as it's not the FSF
publishing it, it wouldn't be bound by the FSF's public mission and
the copyright assignments it received.

-- 
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

[Tim Post]

On Sun, 2007-06-17 at 23:19 +0300, Al Boldi wrote:

> 
> Wow!
> 
> Under what circumstances would it be possible to receive permission for 
> modification?
> 
> 
> Thanks for being GPL!
> 
> --
> Al
> 

If the GPL2 were itself modifiable, there would be so many GPL licenses
that the term "Relased under the GPL 2" would only indicate you chose
one of a million licenses or wrote your own.

That may not be a bad thing, not for me to debate. But I think that's
why the FSF imposes the restriction. I don't feel that is unreasonable,
but you might, and its your right to do so and be vocal about it.

You could, of course take what language that you agree with, re-write it
and make it your own and add a clause that allows people to modify the
license provided that they do not call it the "XYZ" license any longer.

But you wouldn't be making a license - you'd be making the raw material
for people to make their own license. 

I don't think that's what the FSF wanted to accomplish. Again, just my
opinion. I didn't make the GPL so there's no way I could know.

Best,
--Tim

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

[Tim Post]

On Mon, 2007-06-18 at 01:08 +0300, Al Boldi wrote:
> Alexandre Oliva wrote:
> > On Jun 17, 2007, Al Boldi <a1426z@gawab.com> wrote:
> > > Under what circumstances would it be possible to receive permission for
> > > modification?
> >
> > You have to ask the copyright holder.
> >
> > Affero did just that, and so the Affero GPL was born.
> >
> > Just don't assume the FSF will grant such permissions lightly.
> 
> But wouldn't this be against the spirit of the GPL?

This may be helpful to you :

http://www.gnu.org/licenses/gpl-faq.html#ModifyGPL

It describes the scenario and how to go about doing it.

Best,
--Tim

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

[Al Boldi]

Alexandre Oliva wrote:
> Consider this scenario: vendor tivoizes Linux in the device, and
> includes the corresponding sources only in a partition that is
> theoretically accessible using the shipped kernel, but that nothing in
> the software available in the machine will let you get to.  Further,
> sources (like everything else on disk) are encrypted, and you can only
> decrypt it with hardware crypto that is disabled if the boot loader
> doesn't find a correct signature for the boot partition, or maybe the
> signature is irrelevant, given that everything on disk is encrypted in
> such a way that, if you don't have the keys, you can't update the
> kernel properly anyway.  The vendor refuses to give customers other
> copies of the sources.  To add insult to the injury, the vendor
> configures the computer to set up the encrypted disk partition
> containing the sources as a swap device, such that the shared-secret
> key used to access that entire filesystem is overwritten upon the
> first boot, rendering the entire previous contents of the partition
> holding the source code into an incomprehensible stream of bits.
>
> Does anyone think this is permitted by the letter of GPLv2?

Yes.

> Is it in the spirit of GPLv2?

No, but that's besides the point.

You can only hold people responsible for the letter, lest there be chaos.

If there is a specific usage spirit you want to protect, then you must 
formulate it in letter.

> How are the sources passed on in this way going to benefit the user or the
> community?

They still have to provide the source by other GPL means of their choosing.

> Is this still desirable by the Linux developers?

Looks undesirable to me, but still valid.


Thanks!

--
Al

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

[Alexandre Oliva]

On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:

>> Is it in the spirit of GPLv2?

> No, but that's besides the point.

Thanks for informing me about the point *I*'m trying to make ;-)

> You can only hold people responsible for the letter, lest there be chaos.

That's not *quite* how it works, but that's a general idea, yes.

>> How are the sources passed on in this way going to benefit the user or the
>> community?

> They still have to provide the source by other GPL means of their choosing.

This is contradictory.  You said the scenario I described was
permitted, and the scenario included the vendor's refusal to give
customers other copies of the sources.

Which is 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 Boldi]

Alexandre Oliva wrote:
> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:
> >> Is it in the spirit of GPLv2?
> >
> > No, but that's besides the point.
>
> Thanks for informing me about the point *I*'m trying to make ;-)
>
> > You can only hold people responsible for the letter, lest there be
> > chaos.
>
> That's not *quite* how it works, but that's a general idea, yes.
>
> >> How are the sources passed on in this way going to benefit the user or
> >> the community?
> >
> > They still have to provide the source by other GPL means of their
> > choosing.
>
> This is contradictory.  You said the scenario I described was
> permitted, and the scenario included the vendor's refusal to give
> customers other copies of the sources.
>
> Which is it?

I read your scenario of the vendor not giving you the source to mean: not 
directly; i.e.  they could give you a third-party download link.


Thanks!

--
Al

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

[Alexandre Oliva]

On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:

> I read your scenario of the vendor not giving you the source to mean: not 
> directly; i.e.  they could give you a third-party download link.

This has never been enough to comply with GPLv2.

FWIW, it is one of the improvements in 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

[david]

On Tue, 26 Jun 2007, Alexandre Oliva wrote:

> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:
>
>> I read your scenario of the vendor not giving you the source to mean: not
>> directly; i.e.  they could give you a third-party download link.
>
> This has never been enough to comply with GPLv2.
>
> FWIW, it is one of the improvements in GPLv3.

either it's an improvement in the GPLv3 or it's a violation of GPLv2.

you can't say that the GPLv2 prohibits it _and_ it's an improvement in the 
GPLv3.

unless you are saying that the GPLv3 is saying that a third party link now 
_is_ sufficiant. this seems to be counter to what the FSF is claiming 
(with good reasoning. after all, you don't control the third party site, 
so it could change or go away and now the people who got the binaries from 
you can't get the sources)

David Lang

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

[David Schwartz]

Alexandre Oliva:

> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:
>
> > I read your scenario of the vendor not giving you the source to
> > mean: not
> > directly; i.e.  they could give you a third-party download link.
>
> This has never been enough to comply with GPLv2.

A lot of people seem to say this, but I don't think it's true. Section 3b
says:

    Accompany it with a written offer, valid for at least three
    years, to give any third party, for a charge no more than your
    cost of physically performing source distribution, a complete
    machine-readable copy of the corresponding source code, to be
    distributed under the terms of Sections 1 and 2 above on a medium
    customarily used for software interchange; or,

A web page with a download URL is just such an offer. The Internet is a
medium customarily used for software interchange. I do not see why the
following statement doesn't meet the requirements above:

"The source code for this product is available under the terms of the GPL
from the following web page http://www.mycompanyname.com/gpl"

This assumes that no special steps are needed to obtain the software from
that web page.

DS

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

[Tomas Neme]

> "The source code for this product is available under the terms of the GPL
> from the following web page http://www.mycompanyname.com/gpl"
>
> This assumes that no special steps are needed to obtain the software from
> that web page.

But thats YOURcompanyname.com. Not a third party. If you gave as a
link somebodyelsescompany.com/gpl then somebodyelse could get rid of
the link, and your offer wouldn't be valid for "at least three years"

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Alexandre Oliva]

On Jun 26, 2007, david@lang.hm wrote:

> unless you are saying that the GPLv3 is saying that a third party link
> now _is_ sufficiant.

Yup.  The improvement in GPLv3 is to relax the requirement of
providing source code in physical medium if you choose to not
distribute it along with the binaries.  It's recognizing that internet
access is no longer a barrier that could stop someone from obtaining
the sources they're entitled to.  Even someone who doesn't have
regular or fast internet access can hire a third party who does to
perform the download and record it.

I.e., with GPLv3, you *can* point at the sources you used, even in a
site that you don't control.

However, if the site takes the sources out, you're still responsible
for providing sources to those who received the sources from you from
that point on.  Or something like that, 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

[david]

On Tue, 26 Jun 2007, Alexandre Oliva wrote:

> On Jun 26, 2007, david@lang.hm wrote:
>
>> unless you are saying that the GPLv3 is saying that a third party link
>> now _is_ sufficiant.
>
> Yup.  The improvement in GPLv3 is to relax the requirement of
> providing source code in physical medium if you choose to not
> distribute it along with the binaries.  It's recognizing that internet
> access is no longer a barrier that could stop someone from obtaining
> the sources they're entitled to.  Even someone who doesn't have
> regular or fast internet access can hire a third party who does to
> perform the download and record it.
>
> I.e., with GPLv3, you *can* point at the sources you used, even in a
> site that you don't control.
>
> However, if the site takes the sources out, you're still responsible
> for providing sources to those who received the sources from you from
> that point on.  Or something like that, IANAL ;-)

this sounds like a step backwards, you may not have the sources at that 
point if you were relying on the other site to host them.

and by the way, internet access never was a barrier that could stop 
someone from obtaining them, the only issue was you hosting the source vs 
someone else hosting the source.

David Lang

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

[Alexandre Oliva]

On Jun 26, 2007, "David Schwartz" <davids@webmaster.com> wrote:

> Alexandre Oliva:

>> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:

>> > I read your scenario of the vendor not giving you the source to
>> > mean: not directly; i.e.  they could give you a third-party
>> > download link.

>> This has never been enough to comply with GPLv2.

> A lot of people seem to say this, but I don't think it's true.

http://www.gnu.org/licenses/gpl-faq.html#TOCUnchangedJustBinary and
the 3 questions after that should be enlightening as to why people say
this ;-)

    cost of physically performing source distribution, a complete
            ^^^^^^^^^^

Why would 'physically' be there if it didn't mean anything?  When
interpreting legal texts, that's one sort of question you should ask
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

[David Schwartz]

> But thats YOURcompanyname.com. Not a third party. If you gave as a
> link somebodyelsescompany.com/gpl then somebodyelse could get rid of
> the link, and your offer wouldn't be valid for "at least three years"
>
> T

You mean it might not be valid for at least three years. It also might be.
You also might go out of business in less than three years.

You're not violating the GPL simply because you might not be able to honor
your offer in less than three years. You're violating the GPL when you in
fact fail to honor it.

I do agree that providing a link to a third party URL is risky, but I do not
agree that it doesn't comply with the GPL.

DS

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

[David Schwartz]

> On Jun 26, 2007, "David Schwartz" <davids@webmaster.com> wrote:
>
> > Alexandre Oliva:
>
> >> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:
>
> >> > I read your scenario of the vendor not giving you the source to
> >> > mean: not directly; i.e.  they could give you a third-party
> >> > download link.
>
> >> This has never been enough to comply with GPLv2.
>
> > A lot of people seem to say this, but I don't think it's true.
>
> http://www.gnu.org/licenses/gpl-faq.html#TOCUnchangedJustBinary

This consists of entirely unsupported statements.

> and
> the 3 questions after that should be enlightening as to why people say
> this ;-)
>
>     cost of physically performing source distribution, a complete
>             ^^^^^^^^^^
>
> Why would 'physically' be there if it didn't mean anything?

It does. It means you can't charge for adminsitrative or other costs
associated with performing the source distribution. It's necessary because
otherwise someone could claim that it costs them $10,000 to give you the
source code because they need to purchase a license from someone else. This
limits what you can charge for but does not specify what you have to do.

> When
> interpreting legal texts, that's one sort of question you should ask
> yourself.

It's obvious why it's there. If you're going to charge for the distribution,
the charge must be nominal and justified by actual distribution cost.

Note that even a distribution over the Internet must be physically performed
in this sense (actual physical activity by a human being is required to
perform this type of distribution, both in setup and in maintenance). I
would argue that the GPL allows you to charge these costs if you really
wanted to, though it's hard to imagine why anyone would bother.

DS

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

[David Schwartz]

david@lang.hm wrote:

> this sounds like a step backwards, you may not have the sources at that
> point if you were relying on the other site to host them.

You would then be violating the GPL, under any version. The GPL is quite
clear that being unable to comply with it means you do not get the benefits
it offers rather than excusing you from meeting its requirements.

You *MUST* have the source code in order to distribute it on request. You
cannot ship GPL'd works without offering source code just by arranging it
(deliberately or accidentally) so that you don't have the source code.

> and by the way, internet access never was a barrier that could stop
> someone from obtaining them, the only issue was you hosting the source vs
> someone else hosting the source.

The GPLv2 never specified one way or the other.

If you do allow someone else to host them, you are responsible for making
sure they remain available for at least three years from the last time you
used them as an offer. Should they stop distributing, you would be violating
the GPL. Nothing in the GPL says you can't rely on third parties for your
GPL compliance. Of course, this could be a very risky thing to do. However,
there is no GPL violation so long as they do in fact remain operational for
three years from the last time you distributed.

In fact, a third party is no more risky than any other setup. Any company
can go out of business within the three-year period after distribution.
There are many real-world cases where a third party having the source is
actually more likely to result in actual GPL compliance than the
distributor. (Consider a fly-by-night company selling Fedora binary
distributions burnt to CDROM on the stop for $1 on a street corner.)

One way to avoid this problem is to maintain your own web page that links to
the third party's download. You would have to host the sources yourself if
you couldn't make other arrangements at any point during the three year
period.

This is no different from any other case where the offer is not honored. If
the offer is not honored in a case where the GPL requires that it be, the
GPL is being violated.

DS

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

[Alexandre Oliva]

On Jun 26, 2007, david@lang.hm wrote:

>> However, if the site takes the sources out, you're still responsible
>> for providing sources to those who received the sources from you from
>> that point on.  Or something like that, IANAL ;-)

> this sounds like a step backwards, you may not have the sources at
> that point if you were relying on the other site to host them.

You should have them.  This provision is not an excuse from your
obligations, it's just a pragmatic concession.

> and by the way, internet access never was a barrier that could stop
> someone from obtaining them

Back when GPLv2 was written, it really was.

-- 
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 Boldi]

Alexandre Oliva wrote:
> On Jun 26, 2007, Al Boldi <a1426z@gawab.com> wrote:
> > I read your scenario of the vendor not giving you the source to mean:
> > not directly; i.e.  they could give you a third-party download link.
>
> This has never been enough to comply with GPLv2.

Section 3a of the GPLv2 mentions "a medium customarily used for software 
interchange".  I would think the Internet is a medium customarily used for 
software interchange, is it not?


Thanks!

--
Al

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

[Tomas Neme]

> In the sense that he can decide to remove all contributions from
> dissenting authors, yes, he does.  But he can't impose his more lax
> interpretation upon other authors.  Under copyright, it's the more

yes, I saw my argument going weak as I wrote it, but what I said later:

> So if you own a part of the kernel, then you can pursuit TiVo on your
> own, if they did direct use of that part especifically and break (in
> your opinion) what you feel GPLv2 means. You can form the CATV2
> (CodersAgainstTiVo v2) and try to get TiVo to stop using the Linux
> Kernel for their product (because, believe me, they WON'T release the
> keys). Yeay, we lost Tivo's improvements on the kernel, and the
> posibility of having a working kernel if anyone feels like
> back-ingeneering TiVo for their own amusement.

is still right. What I meant, at least by the end of that email, was
that he has the last word on trying to stop TiVo from using The Linux
Kernel. Each author can still go and stop them from using his part,
and the derivative work that is The Linux Kernel.

But that brings another question: what if TiVo decided to remove all
code from the complaining parts and rewrite them? that wouldn't be The
Linux Kernel anymore, but it would be a derivative work of all the
parts that don't disagree with Tivoization, but is that legal?

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Alexandre Oliva]

On Jun 22, 2007, "Tomas Neme" <lacrymology@gmail.com> wrote:

> The thing is, what matters in copyright and licencing matters is what
> the author of the code understands, no the licence's author, if
> ambiguous. And the kernel's rights holder is Linus.

Since he didn't get copyright assignments, each contributor is the
copyright holder of her/his own contribution.  And this means each
holder gets a say on how s/he understood GPLv2.

IANAL, but I think if Linus' intended interpretation had been
clarified all the way from the beginning, he could have grounds to
claim that everyone else had implicitly accepted that reading by
contributing to the project.

But since it was a decision made many years later, his clarification
on his reading of the license is in a way an additional permission
that affects only his own contributions; other authors are still
entitled to try to enforce their understanding of the legal terms of
the license, and they would have the spirit of the GPL and its
preamble on their side to guide the interpretation.  Even if contract
law states something like, in adhesion contracts, the party who writes
the contract gives the other party the benefits of any ambiguity in
the writing, the GPL is not a contract, it's a license, and per
copyright law, licenses are to be interpreted restrictively.

> Linus has the last word on it.

In the sense that he can decide to remove all contributions from
dissenting authors, yes, he does.  But he can't impose his more lax
interpretation upon other authors.  Under copyright, it's the more
restrictive reading that prevails, in that any holder who understands
his rights are being trampled can enforce them.  And since at least
one such author is vocal in his dissent, not even estoppel defenses
would apply.  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

[Tomas Neme]

> powerful.  It is pretty obvious that when Linus adopted GPLv2 he
> didn't realize it reached that point.  That when Tivo invented
> Tivoization, he decided he wanted to permit this, and thus grants an
> implicit additional permission for anyone to do it with his code,
> doesn't mean other participants in the Linux community feel the same
> way, or read the GPLv2 the same way, and could be somehow stopped from
> enforcing the license the way they meant it.

The thing is, what matters in copyright and licencing matters is what
the author of the code understands, no the licence's author, if
ambiguous. And the kernel's rights holder is Linus. The authors of the
particular bits of code can complain about what tivo's doing, but
since TiVo's using the linux kernel, and GPLv2 says

"These requirements apply to the modified work as a whole.  If
identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.  But when you
distribute the same sections as part of a whole which is a work based
on the Program, the distribution of the whole must be on the terms of
this License, whose permissions for other licensees extend to the
entire whole, and thus to each and every part regardless of who wrote it."

Linus has the last word on it. IANAL, but as far as I can tell this
reads as "this licence applies to the whole, not to the parts by
themselves", and so does who the "holder" is, I'd believe.

So if you own a part of the kernel, then you can pursuit TiVo on your
own, if they did direct use of that part especifically and break (in
your opinion) what you feel GPLv2 means. You can form the CATV2
(CodersAgainstTiVo v2) and try to get TiVo to stop using the Linux
Kernel for their product (because, believe me, they WON'T release the
keys). Yeay, we lost Tivo's improvements on the kernel, and the
posibility of having a working kernel if anyone feels like
back-ingeneering TiVo for their own amusement.

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Kyle Moffett]

On Jun 21, 2007, at 15:19:35, Stephen Clark wrote:
> David Schwartz wrote:
>>> On Wed, 20 Jun 2007 12:55:10 -0700 "David Schwartz"  
>>> <davids@webmaster.com> wrote:
>>>> A key is a number. A signature is a number. They are neither  
>>>> statements nor instructions. The argument that GPLv2 prohibits  
>>>> Tivoization is really and truly absurd. It has neither a legal  
>>>> nor a moral leg to stand on.
>>>
>>> A computer program is a number too.
>>
>> No, it's not. It can be expressed as a number, but it is not a  
>> number.
>>
> ??? can be expressed as a number, but it is not a number ??? sure  
> its a number.
>
>> Keys are purely numbers, they are nothing else. Signatures are  
>> pure primitive facts encoded as numbers (authority X blessed  
>> object Y).
>>
>> A computer program is a set of instructions to accomplish a  
>> particular result. It can be expressed as a number, but that  
>> doesn't mean it is a number.
>>
>> It might be true in principle to develop a scheme whereby every  
>> physical object uniquely corresponds to an extremely large number.  
>> That doesn't turn physical objects into numbers.

Both of you lose this argument.  All irrational numbers, for example,  
"break" every copyright that could possibly exist.  For example, you  
can find any arbitrary sequence of Base-N digits when you express PI  
in base-N form.  I can simultaneously express both the laws of  
physics (not copyrightable) and the latest episode of the TV show  
"Numbers" (thoroughly copyrighted) as numbers.  In fact, we do both  
all the time (you can express both the latest equations for  
theoretical physics and a TV show as bits (IE: numbers) on an HDD.   
Ergo "$FOO is a number" says *NOTHING* about whether or not copyright  
applies to $FOO.  In case you haven't noticed, the whole damn point  
of math is that you can express *EVERYTHING* as numbers, albeit maybe  
horribly unbelievably complex ones.

Now, back to actual legal issues: Since most copyright laws  
explicitly prevent copyrighting of pure math, the only actual  
protection you have for some collection of so-called numbers is  
whether or not the numbers *REPRESENT* something which may be  
copyrighted.  Furthermore, copyright has _always_ been independent of  
representation; a person owns copyright on a book regardless of  
whether it's hardback, softcover, digital, memorized, etc.  The  
person who owns the copyright on a book is able to prevent someone  
who has memorized the book from giving public recitals of said book,  
and the neuron-linkage-based storage the brain uses is about as far  
as you can possibly get from twiddling magnetic bits on a disk drive  
or dumping carbon-based inks on a page made of plant cellulose.

Cheers,
Kyle Moffett

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

[David Schwartz]

On Jun 21, 2007, aoliva@redhat.com wrote:

> On Jun 21, 2007, david@lang.hm wrote:

> >> For the record, GPLv2 is already meant to accomplish this.  I don't
> >> understand why people who disagree with this stance chose GPLv2.
> >> Isn't "no further restrictions" clear enough?

> > everyone else is reading this as 'no further license restrictions'

> I didn't see anyone else add "license" where you did.  "No further
> restrictions on the rights granted herein" is very powerful and
> extensive, and that's how it was meant to be.

I agree. For example, a patent might impose a further restriction. The GPL
was clearly meant to foreclose that. There are any number of other ways of
nasty, subtle ways to impose further restrictions, and the GPLv2 meant to
foreclose all of them.

> > not no hardware restrictions' becouse GPLv2 explicitly says that it
> > has nothing to do with running the software, only with distributing
> > it.

> It also says that running the software is not restricted, and since
> copyright law in the US doesn't regulate execution, receiving the
> software does grant the recipient the right to run the software.  So
> the distributor can't impose restrictions on it.

Right.

The response to this argument is that it is mind-bogglingly obvious that the
GPL doesn't mean that no *authorization* decisions can stand in your way. It
didn't mean that I couldn't keep the root password to my server secret even
though that denies you the "right" to modify the Linux kernel running on it.

Some entity has to decide what software runs on any particular piece of
hardware, and it was never the intent of the GPL to specify or limit who
that person was. This has been discussed many times over many years, longer
before Tivoization was even thought of, and it was agreed that the GPL
didn't foreclose authorization obstacles to software modification.

Why doesn't Linux allow a non-root user to install a module or change which
kernel the system runs? Doesn't that limit the GPL rights of all non-root
users to modify the GPL'd kernel software on that machine? OF COURSE NOT.

DS

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>> On Jun 21, 2007, david@lang.hm wrote:
>> 
>>> this is your right with your code. please stop browbeating people who
>>> disagree with you.
>> 
>> For the record, GPLv2 is already meant to accomplish this.  I don't
>> understand why people who disagree with this stance chose GPLv2.
>> Isn't "no further restrictions" clear enough?

> everyone else is reading this as 'no further license restrictions'

I didn't see anyone else add "license" where you did.  "No further
restrictions on the rights granted herein" is very powerful and
extensive, and that's how it was meant to be.

> not no hardware restrictions' becouse GPLv2 explicitly says that it
> has nothing to do with running the software, only with distributing
> it.

It also says that running the software is not restricted, and since
copyright law in the US doesn't regulate execution, receiving the
software does grant the recipient the right to run the software.  So
the distributor can't impose restrictions on 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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> this is your right with your code. please stop browbeating people who
>> disagree with you.
>
> For the record, GPLv2 is already meant to accomplish this.  I don't
> understand why people who disagree with this stance chose GPLv2.
> Isn't "no further restrictions" clear enough?

everyone else is reading this as 'no further license restrictions' not 'no 
hardware restrictions' becouse GPLv2 explicitly says that it has nothing 
to do with running the software, only with distributing it.

David Lang

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> this is your right with your code. please stop browbeating people who
> disagree with you.

For the record, GPLv2 is already meant to accomplish this.  I don't
understand why people who disagree with this stance chose GPLv2.
Isn't "no further restrictions" clear enough?

-- 
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 21, 2007, Andrew McKay <amckay@iders.ca> wrote:

> Alexandre Oliva wrote:
>> On Jun 21, 2007, Andrew McKay <amckay@iders.ca> wrote:
>> 
>>> A balance of freedom to the licensee and the licenser.  It's my
>>> opinion that GPLv3 potentially shifts the balance too far to the
>>> licensee.
>> 
>> It's more of a balance of freedom between licensee and licensee,
>> actually.  It's a lot about making sure no one can acquire a
>> privileged position, such that every licensee plays under the same
>> rules.  (The copyright holder is not *acquiring* a privileged
>> position, copyright law had already granted him/her that position.)

> I do see what you're saying here.  But it does take the away the
> ability of a licensee to protect themselves from another malicious
> licensee.

Sorry, I don't follow what the "it" refers to in your sentence.

> If the ultimate goal of the Free Software community is to get source
> code out to the public, I think that was captured in GPLv2.

That's a correct logical inference, but since the premise is false,
the conclusion is garbage.

GPLv2 goes far beyond getting source code out to the public.  It
contains the "no further restrictions" language, which is very
powerful.  It is pretty obvious that when Linus adopted GPLv2 he
didn't realize it reached that point.  That when Tivo invented
Tivoization, he decided he wanted to permit this, and thus grants an
implicit additional permission for anyone to do it with his code,
doesn't mean other participants in the Linux community feel the same
way, or read the GPLv2 the same way, and could be somehow stopped from
enforcing the license the way they meant it.

Ultimately, the current situation is that we have two
mutually-incompatible license intents being used in Linux, and no
matter how much those who want to grant the permission say so, this
doesn't trample other contributor's rights to enforce the license they
chose for their code.  Especially those who started contributing long
before the decision that "what TiVo does is good" was announced.

Now, since these two license intents are expressed by the same
license, and what the license demands is that derived works must be
under the same license, they are compatible, but since the intents are
distinct, what prevails is, as in any case of combination of different
licensing provisions, is the most restrictive provision.

So Linux does not permit tivoization today.  Linus does, Linux
doesn't.

All this fuss about the anti-tivoization provisions in GPLv3 is just a
consequence of reading the GPLv2 without fully understanding its
intended consequences, not having foresight to clarify the intent to
constrain the "no further restrictions" provisions to match the
alternate interpretation, and opposing the removal of the ambiguity
because it doesn't match the choice that *some* of the developers
would like it to go.

Who's the ambiguity good for?

-- 
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]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>>> If it's input-only, then you can't possibly harm the operation of the
>>> network by only listening in, can you?
>
>> Ok, so you consider any anti-piracy measures to be something that
>> GPLv3 should prohibit.
>
> In general, I object to the use of my code in ways that don't permit
> users to run it for any purpose, study it, adapt it to suit their own
> needs, modify the code and distribute it, modified or not.
>
> If this means my code can't be used to implement DRM, that's a good
> thing.
>
> All anti-piracy measures I've ever known deprive users of legitimate
> rights too.  So, yes, I don't agree with these measures.
>
> I believe in punishing the guilty, not in punishing innocent
> bystanders just such that the guilty have to find work arounds to
> become guilty.

this is your right with your code. please stop browbeating people who 
disagree with you.

David Lang

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>> If it's input-only, then you can't possibly harm the operation of the
>> network by only listening in, can you?

> Ok, so you consider any anti-piracy measures to be something that
> GPLv3 should prohibit.

In general, I object to the use of my code in ways that don't permit
users to run it for any purpose, study it, adapt it to suit their own
needs, modify the code and distribute it, modified or not.

If this means my code can't be used to implement DRM, that's a good
thing.

All anti-piracy measures I've ever known deprive users of legitimate
rights too.  So, yes, I don't agree with these measures.

I believe in punishing the guilty, not in punishing innocent
bystanders just such that the guilty have to find work arounds to
become guilty.

-- 
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

[Andrew McKay]

Alexandre Oliva wrote:
> On Jun 21, 2007, Andrew McKay <amckay@iders.ca> wrote:
> 
>> A balance of freedom to the licensee and the licenser.  It's my
>> opinion that GPLv3 potentially shifts the balance too far to the
>> licensee.
> 
> It's more of a balance of freedom between licensee and licensee,
> actually.  It's a lot about making sure no one can acquire a
> privileged position, such that every licensee plays under the same
> rules.  (The copyright holder is not *acquiring* a privileged
> position, copyright law had already granted him/her that position.)
> 

I do see what you're saying here.  But it does take the away the ability of a 
licensee to protect themselves from another malicious licensee.  If the ultimate 
goal of the Free Software community is to get source code out to the public, I 
think that was captured in GPLv2.  GPLv3 oversteps its bounds.

Anyways I think this topic has been quite covered.

Andrew

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

[Andrew McKay]

Alexandre Oliva wrote:
> On Jun 21, 2007, Andrew McKay <amckay@iders.ca> wrote:
> 
>> david@lang.hm wrote:
>>> how can the server tell if it's been tampered with?
> 
>> I agree with this statement.
> 
> Err...  That's a question, not a statement ;-)
> 

Sorry, that's what happens when one types before getting a cup of coffee in the 
morning.

Andrew

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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>>> On Jun 21, 2007, david@lang.hm wrote:
>>>
>>>> no, one of the rules for the network is that the software must be
>>>> certified,
>>>
>>> In this case you might have grounds to enforce this restriction of the
>>> network on the network controller itself, I suppose.
>
>> how would the network controller know if the software has been modified?
>
> The loader could check that and set a flag in the controller.
>
>> what sort of signal can the network controller send that couldn't be
>> forged by the OS?
>
> Whatever the network controller designer created to enable it to do
> so.
>
>> how would you do this where the device is a receiver on the netwoek
>> (such as a satellite receiver)
>
> If it's input-only, then you can't possibly harm the operation of the
> network by only listening in, can you?

Ok, so you consider any anti-piracy measures to be something that GPLv3 
should prohibit.

thanks for finally takeing a position.

David Lang

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> On Thu, 21 Jun 2007, Lennart Sorensen wrote:

>> You can't use this code if you cooporate with anyone that requires
>> DRM systems.

> I think their earlier versions did say this.

Show me a GPLv3 draft that did it?

Start here, section 3:
http://gplv3.fsf.org/gpl-draft-2006-01-16.html

> DRM does have some legitimate uses, for example redhat installations
> not installing unsigned software is a form of DRM

Doh.  Then chmod og-r is DRM too.  And stronger DRM while at that,
since the user denied permission to read the file cannot take it back,
whereas the verification of unsigned software is just a warning, that
you can often bypass by telling the software to go ahead and install
it regardless of signatures.

-- 
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 21, 2007, Andrew McKay <amckay@iders.ca> wrote:

> A balance of freedom to the licensee and the licenser.  It's my
> opinion that GPLv3 potentially shifts the balance too far to the
> licensee.

It's more of a balance of freedom between licensee and licensee,
actually.  It's a lot about making sure no one can acquire a
privileged position, such that every licensee plays under the same
rules.  (The copyright holder is not *acquiring* a privileged
position, copyright law had already granted him/her that position.)

-- 
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 21, 2007, Andrew McKay <amckay@iders.ca> wrote:

> david@lang.hm wrote:
>> how can the server tell if it's been tampered with?

> I agree with this statement.

Err...  That's a question, not a statement ;-)

-- 
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 21, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> Apparently the only restrictions ever permitted are the ones the FSF
> thinks of.

Where does this nonsensical idea come from?  How does it follow that,
from FSF offering a licensing option to authors, you conclude that
nobody could ever establish whatever other restrictions they liked?

> So really what the GPL v3 wants to have is to make sure that the user
> can reproduce from the sources a bit for bit identical copy of the
> binaries?

No, this is not enough to enable someone to adapt the software to
one's own needs.

> Too bad compilers that put time stamps and such into the
> binary would make that imposible.

This would be the copyright author imposing such a restriction, not
the software distributor.

> I don't think there is any way that can be written into the GPL that
> can prevent all loop holes for how to make signed binaries.

Which is one possible reason to explain why the FSF switched to the
'Installation Information' approach.

> There doesn't have to be an agreement.  The software company could just
> release specs for a hardware design and let others freely go and build
> them from that design.

Aah, so the software company has designed a mechanism to restrict
users' freedoms, and is just leaving it up to third parties to
complete the implementation?  I think these design documents could be
used in a court to prove intent to impose restrictions on the users,
but IANAL.

>> However, if there's no such agreement, if the copyright holder has no
>> copyright claims over the hardware or works shipped in it, there's
>> nothing the copyright holder can do about it, and that's probably how
>> it should be, since a copyright license (!= contract) can't possibly
>> prohibit people from creating hardware limited in function, it can
>> only tell people that, in order for them to have permission to modify
>> or distribute the covered work, they must abide by certain conditions.
>> And if they don't want to abide by the conditions, and they don't
>> manage to obtain a license from the copyright holders that doesn't
>> impose conditions they can't accept, they just can't modify or
>> distribute the work.

> But if the hardware ships with only code that simply waits for the user
> to provide some code for it to isntall (which has to be signed in a way
> the hardware likes), then the hardware has nothing to do with the
> license of the software.

Correct.  That's pretty much what I said, isn't it?

> I hope no one does this, but I still don't see how the GPLv3 draft deals
> with this case, or even how it could deal with it.

It doesn't, and it probably shouldn't.

-- 
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 21, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> On Wed, 20 Jun 2007 12:55:10 -0700
>> "David Schwartz" <davids@webmaster.com> wrote:

>> > A key is a number. A signature is a number. They are neither
>> > statements nor
>> > instructions. The argument that GPLv2 prohibits Tivoization is
>> > really and
>> > truly absurd. It has neither a legal nor a moral leg to stand on.

>> A computer program is a number too.

> No, it's not. It can be expressed as a number, but it is not a number.

By this logic, then a key is a key, and a signature is a signature.
They can be expressed as numbers, sure.

> A computer program is a set of instructions to accomplish a particular
> result. It can be expressed as a number, but that doesn't mean it is a
> number.

A key is an input to a cryptographical algorithm, and a signature is
an output.  I could try to come up with more creative definitions, but
you get the idea already.

-- 
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 21, 2007, Bernd Schmidt <bernds_cb1@t-online.de> wrote:

> I went and made some comments on the draft, and they appear to no
> longer be there a few days later.

This would be very bad.  Please let me know what they were about and
I'll try to figure out what happened.

Did you by any chance file them against an earlier draft?  Those (for
obvious reasons) no longer appear against the current draft, but
they're still accessible by other means.

-- 
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 21, 2007, david@lang.hm wrote:

> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>> On Jun 21, 2007, david@lang.hm wrote:
>> 
>>> no, one of the rules for the network is that the software must be
>>> certified,
>> 
>> In this case you might have grounds to enforce this restriction of the
>> network on the network controller itself, I suppose.

> how would the network controller know if the software has been modified?

The loader could check that and set a flag in the controller.

> what sort of signal can the network controller send that couldn't be
> forged by the OS?

Whatever the network controller designer created to enable it to do
so.

> how would you do this where the device is a receiver on the netwoek
> (such as a satellite receiver)

If it's input-only, then you can't possibly harm the operation of the
network by only listening in, can you?

-- 
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

[Stephen Clark]

David Schwartz wrote:

>>On Wed, 20 Jun 2007 12:55:10 -0700
>>    
>>
>
>  
>
>>"David Schwartz" <davids@webmaster.com> wrote:
>>    
>>
>
>  
>
>>>A key is a number. A signature is a number. They are neither
>>>statements nor
>>>instructions. The argument that GPLv2 prohibits Tivoization is
>>>really and
>>>truly absurd. It has neither a legal nor a moral leg to stand on.
>>>      
>>>
>
>  
>
>>A computer program is a number too.
>>    
>>
>
>No, it's not. It can be expressed as a number, but it is not a number.
>  
>
??? can be expressed as a number, but it is not a number ???
sure its a number.

>Keys are purely numbers, they are nothing else. Signatures are pure
>primitive facts encoded as numbers (authority X blessed object Y).
>
>A computer program is a set of instructions to accomplish a particular
>result. It can be expressed as a number, but that doesn't mean it is a
>number.
>
>It might be true in principle to develop a scheme whereby every physical
>object uniquely corresponds to an extremely large number. That doesn't turn
>physical objects into numbers.
>
>DS
>
>
>-
>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/
>
>  
>


-- 

"They that give up essential liberty to obtain temporary safety, 
deserve neither liberty nor safety."  (Ben Franklin)

"The course of history shows that as a government grows, liberty 
decreases."  (Thomas Jefferson)

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

[Alan Cox]

> So much for "Land of the free". :(

That was always just a typo. Its the Land of the Fee

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

[david]

On Thu, 21 Jun 2007, Lennart Sorensen wrote:

> On Thu, Jun 21, 2007 at 10:51:06AM -0700, david@lang.hm wrote:
>> you snippede the bit about not knowing how to stop it
>
> I did?  As far as I can tell I quoted it all.  What did I miss?
>
>> they call the section the anti-tivoization, how much more explicit can
>> they get?
>
> They could be as explicit as:
> You can't use this code if you cooporate with anyone that requires
> DRM systems.

I think their earlier versions did say this.

> All their attempts to define user devices and such is just going to
> screw up and miss some things they wanted covered, and disallow things
> they didn't intend to disallow (assuming there is any such thing).
>
>> by the way, just in case anyone is misunderstanding me. I don't believe
>> for a moment that all these anti-tamper features actually work in the real
>> world (the PS3 hacking kits are proof of the lengths people will go to to
>> make the 'hard' hardware-level hacking trivial to do) but the approach
>> needs to be at secure modulo hardware tampering or software bugs.
>
> DRM is completely pointless.  It only stops casual end users from doing
> things.  It doesn't stop anyone with any technical clue from doing
> things.  I keep hoping one day the people in charge at the big media
> companies will understand this, and stop asking for people to implement
> it.  Of course in the mean time there are companies perfectly willing to
> claim to have unbreakable DRM for sale, while knowing full well (if they
> are competent) that it is a lie.  So as long as the people in charge at
> big media are clueless about technology, and as long as there are
> companies willing to lie to them for money, then we will probably
> continue to have DRM crap to deal with.

DRM does have some legitimate uses, for example redhat installations not 
installing unsigned software is a form of DRM

> I don't think the GPLv3 is the place to try to remove DRM.  What the FSF
> should be doing is try to educate the people who are advocating the use
> of DRM about the fact that it can't ever work.  You can make more and
> more stupid laws about how people can't remove the DRM, but people who
> break copyright obviously already are breaking the law, so what is the
> point in having more lows for them to break.  That is where this problem
> should be fought, not in the GPLv3.  The GPLv3 is never going to solve
> the problem, only educating people can do that.

this is exactly what most of the people who are arguing against this 
provision are saying.

David Lang

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

[David Schwartz]

> On Wed, 20 Jun 2007 12:55:10 -0700

> "David Schwartz" <davids@webmaster.com> wrote:

> > A key is a number. A signature is a number. They are neither
> > statements nor
> > instructions. The argument that GPLv2 prohibits Tivoization is
> > really and
> > truly absurd. It has neither a legal nor a moral leg to stand on.

> A computer program is a number too.

No, it's not. It can be expressed as a number, but it is not a number.

Keys are purely numbers, they are nothing else. Signatures are pure
primitive facts encoded as numbers (authority X blessed object Y).

A computer program is a set of instructions to accomplish a particular
result. It can be expressed as a number, but that doesn't mean it is a
number.

It might be true in principle to develop a scheme whereby every physical
object uniquely corresponds to an extremely large number. That doesn't turn
physical objects into numbers.

DS

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

[Lennart Sorensen]

On Thu, Jun 21, 2007 at 10:51:06AM -0700, david@lang.hm wrote:
> you snippede the bit about not knowing how to stop it

I did?  As far as I can tell I quoted it all.  What did I miss?

> they call the section the anti-tivoization, how much more explicit can 
> they get?

They could be as explicit as: 
You can't use this code if you cooporate with anyone that requires
DRM systems.

All their attempts to define user devices and such is just going to
screw up and miss some things they wanted covered, and disallow things
they didn't intend to disallow (assuming there is any such thing).

> by the way, just in case anyone is misunderstanding me. I don't believe 
> for a moment that all these anti-tamper features actually work in the real 
> world (the PS3 hacking kits are proof of the lengths people will go to to 
> make the 'hard' hardware-level hacking trivial to do) but the approach 
> needs to be at secure modulo hardware tampering or software bugs.

DRM is completely pointless.  It only stops casual end users from doing
things.  It doesn't stop anyone with any technical clue from doing
things.  I keep hoping one day the people in charge at the big media
companies will understand this, and stop asking for people to implement
it.  Of course in the mean time there are companies perfectly willing to
claim to have unbreakable DRM for sale, while knowing full well (if they
are competent) that it is a lie.  So as long as the people in charge at
big media are clueless about technology, and as long as there are
companies willing to lie to them for money, then we will probably
continue to have DRM crap to deal with.

I don't think the GPLv3 is the place to try to remove DRM.  What the FSF
should be doing is try to educate the people who are advocating the use
of DRM about the fact that it can't ever work.  You can make more and
more stupid laws about how people can't remove the DRM, but people who
break copyright obviously already are breaking the law, so what is the
point in having more lows for them to break.  That is where this problem
should be fought, not in the GPLv3.  The GPLv3 is never going to solve
the problem, only educating people can do that.

--
Len Sorensen

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

[david]

On Thu, 21 Jun 2007, Lennart Sorensen wrote:

> On Thu, Jun 21, 2007 at 10:26:04AM -0700, david@lang.hm wrote:
>> the bios doesn't have enough capability to talk to the outside world for
>> updates.
>
> Of course, although perhaps it could.  More likely my thought was that
> the service when it decides to download an update, would include the
> updated bios image and put it on the boot drive where the existing bios
> can find it.  No signature needs to be added to the boot drive or
> kernel, just checksums in the bios image.
>
>> what tivo actually does is very similar to this
>>
>> they encode into the bios the ability to check a checksum/signature for
>> the kernel+boot filesystem and if they don't match look to see if there is
>> another kernel+boot filesystem available
>>
>> then software on the boot filesystem checks to see if the rest of the
>> system has been tampered with before it mounts /

you snippede the bit about not knowing how to stop it

>> the GPLv3 is trying to do this.
>
> Perhaps they should just explicitly say that then.

they call the section the anti-tivoization, how much more explicit can 
they get?

David Lang

by the way, just in case anyone is misunderstanding me. I don't believe 
for a moment that all these anti-tamper features actually work in the real 
world (the PS3 hacking kits are proof of the lengths people will go to to 
make the 'hard' hardware-level hacking trivial to do) but the approach 
needs to be at secure modulo hardware tampering or software bugs.

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

[Lennart Sorensen]

On Thu, Jun 21, 2007 at 10:26:04AM -0700, david@lang.hm wrote:
> the bios doesn't have enough capability to talk to the outside world for 
> updates.

Of course, although perhaps it could.  More likely my thought was that
the service when it decides to download an update, would include the
updated bios image and put it on the boot drive where the existing bios
can find it.  No signature needs to be added to the boot drive or
kernel, just checksums in the bios image.

> what tivo actually does is very similar to this
> 
> they encode into the bios the ability to check a checksum/signature for 
> the kernel+boot filesystem and if they don't match look to see if there is 
> another kernel+boot filesystem available
> 
> then software on the boot filesystem checks to see if the rest of the 
> system has been tampered with before it mounts /
> 
> the GPLv3 is trying to do this.

Perhaps they should just explicitly say that then.

--
Len Sorensen

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

[david]

On Thu, 21 Jun 2007, Lennart Sorensen wrote:

> 
> On Wed, Jun 20, 2007 at 04:07:57PM -0400, Michael Poole wrote:
>> I do not say that the BIOS is doing anything (legally) wrong.  The
>> wrong act is distributing the binary kernel image without distributing
>> complete source code for it.
>
> So how about this idea then:
>
> Tivo builds a kernel for their box, and release all the sources for how
> to build exactly that kernel.
>
> Tivo builds a bios image for their box, and encodes into it the checksum
> of the kernel, or at least parts of it that they want to ensure are
> present and unmodified.
>
> Everytime the device boots, it checks the kernel image, and it the
> checksums match, it loads and runs the kernel, and otherwise it checks
> if there is a new bios image with a proper signature, updates itself and
> reboots and tries again.

the bios doesn't have enough capability to talk to the outside world for 
updates.

what tivo actually does is very similar to this

they encode into the bios the ability to check a checksum/signature for 
the kernel+boot filesystem and if they don't match look to see if there is 
another kernel+boot filesystem available

then software on the boot filesystem checks to see if the rest of the 
system has been tampered with before it mounts /

> Preventing people from doing things with their own hardware certainly
> seems morally wrong, but legally, I don't see any way to prevent it.
>
> I suppose you could say in the license: You may not use this code in any
> way if you do what the RIPP/MPAA/etc want you to do.

the GPLv3 is trying to do this.

David Lang

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

[Lennart Sorensen]

On Wed, Jun 20, 2007 at 04:07:57PM -0400, Michael Poole wrote:
> I do not say that the BIOS is doing anything (legally) wrong.  The
> wrong act is distributing the binary kernel image without distributing
> complete source code for it.

So how about this idea then:

Tivo builds a kernel for their box, and release all the sources for how
to build exactly that kernel.

Tivo builds a bios image for their box, and encodes into it the checksum
of the kernel, or at least parts of it that they want to ensure are
present and unmodified.

Everytime the device boots, it checks the kernel image, and it the
checksums match, it loads and runs the kernel, and otherwise it checks
if there is a new bios image with a proper signature, updates itself and
reboots and tries again.

Preventing people from doing things with their own hardware certainly
seems morally wrong, but legally, I don't see any way to prevent it.

I suppose you could say in the license: You may not use this code in any
way if you do what the RIPP/MPAA/etc want you to do.

--
Len Sorensen

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

[Lennart Sorensen]

On Thu, Jun 21, 2007 at 01:23:01AM -0300, Alexandre Oliva wrote:
> And then the user who uses such features in ways not permitted by the
> copyright holders are committing a crime.  They can be prosecuted by
> the copyright holders and convicted of the crime.

Well we already clearly know the content providers in the US don't trust
anyone else, and certainly don't think just using copyright laws to sue
people who violate copyright is enough.  No they want to put all sorts
of things in place to try to prevent it from even being possible to
violate copyright in the first place.  They don't believe in innocent
until proven guilty at all.

> That TiVo can somehow become liable for this just shows how broken the
> legal system in the US is.  It's like making a knife manufacturer
> liable for a killing using a knife they made, just because the knife
> didn't have technical measures intended to prevent the knife from
> being used to kill people.

So much for "Land of the free". :(

--
Len Sorensen

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

[david]

On Thu, 21 Jun 2007, Tomas Neme wrote:

>> >  as long as this right is not used by the software distributor to
>> >  impose restrictions on the user's ability to adapt the software to
>> >  their own needs.  The GPLv3 paragraph above makes a fair concession in
>> >  this regard, don't you agree?
>>
>>  no, one of the rules for the network is that the software must be
>>  certified, you are requireing the device to permit the software to be
>>  changed to an uncertified version.(to store credit card numbers and send
>>  them to a third party for example)
>
> Also another way of doing this is having every network ask the kernel
> for its key, and checking it. If it doesn't match a certified key,
> then not allowing you to access the network.

no, this doesn't work becouse if the software has been altered you don't 
know if the key it's giving you matches that software. it could be giving 
you the key from the unmodified software.

David Lang

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

[Andrew McKay]

Alan Cox wrote:
>>> You've made an important mistake. You said "their system". Now its "our
>>> code" and "whoever bought the units' hardware" so it isn't their anything.
>> Yes, the hardware belongs to the user, and the software belongs to the Linux 
>> community.   However I think I wasn't 100% clear, I also mean keeping companies 
>> networks and content secured.  Credit card companies insuring the software 
>> hasn't been modified to skim cards (not that it's the only way to skim a card), 
> 
> If credit card companies are doing this they are failing badly and it
> clearly isn't working. Also lets be clear about this - I don't need any
> credit card company network access to skim older cards, and the newer
> ones have been broken by various non software schemes.

Agreed.  Credit card companies are failing very badly at preventing skimming. 
They definitely need to rethink their model of how the credit card system should 
work.

> 
>>   or Tivo making sure that their content providers are protected.  Lets look at 
>> the credit card example.  Sure the user could modify the system and boot their 
>> own kernel, but it doesn't have to play nice with Mastercard's network anymore. 
>>   Or better yet, would actually report that a certain business's card reader had 
>> been tampered with.
> 
> That to me is a fair comment. I should IMHO be able to load my code and
> my keys on my Tivo. And Walt Disney in return probably should be quite
> free not to trust my keys. You need some fairly strong competition law
> enforcement to make all that work right in the marketplace but as a
> philosophical basis it seems fine. In practice it is likely to lead to
> serious monopoly abuse problems and all sorts of ugly tying of goods that
> you don't want in a free market. Given the completely ineffectual way the
> US enforces its anti-monopoly law, and the slowness of the EU at it the
> results might well be bad - but for other reasons.

I agree as well, the model could be heavily abused.  I'm not sure what the 
solution is as far as fighting monopolies and corporate greed.  But I'd rather 
that it was fought through education of consumers and not with a license 
agreement that should be giving people freedom.  A balance of freedom to the 
licensee and the licenser.  It's my opinion that GPLv3 potentially shifts the 
balance too far to the licensee.

Andrew McKay

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

[Andrew McKay]

david@lang.hm wrote:
> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
> 
>> On Jun 21, 2007, david@lang.hm wrote:
>>
>>> how exactly can they prevent a system that's been tampered with from
>>> accessing their network?
>>
>> By denying access to their servers?  By not granting whatever is
>> needed to initiate network sessions?
>>
>> And note, "it's been tampered with" is not necessarily enough of a
>> reason to cut someone off, it has to meet these requirements:
> 
> how can the server tell if it's been tampered with?
> 

I agree with this statement.  Imagine a proprietary private network where a 
device has been modified to run in an invisible promiscuous mode.  The device 
looks as if it isn't doing anything wrong, but is forwarding the network out 
another interface.  The only way to prevent that type of attack is to not allow 
unauthorized signed Kernels onto that network.

Andrew McKay

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

[Lennart Sorensen]

On Thu, Jun 21, 2007 at 10:56:33AM +0400, Manu Abraham wrote:
> Providing the changes back itself is a great thing altogether.

It also makes sense.  If the changes are accepted back, the community at
large will keep the changes maintained.  Less work for me to do when
going to newer code versions later.  And even better, it may help
someone else out too.

A company is likely to like the reduced maintenance burden part, but I
think the other part is even better.  After all we saved time not having
to write everything our selves, so helping others save time seems only
fair.  The GPL may only require giving the sources to the people who
buys the product, but there isn't really any benefit to us in doing only
that.

--
Len Sorensen

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

[Lennart Sorensen]

On Wed, Jun 20, 2007 at 05:52:40PM -0300, Alexandre Oliva wrote:
> On Jun 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:
> > A patent prevents you from using the software in any way at all,
> > while a hardware restriction prevents you from using the software on
> > that particular hardware, but not on lots of other hardware.  Very
> > big difference.
> 
> So, one disrespects a lot, the other disrespects a little.  Is that
> relevant, when the requirement is "no further restrictions"?

What about the freedom to buy devices with certified code on it, while
still being able to look through the source code and verify for yourself
that it is correct and not full of bugs?  Would it be better if the
devices that have to be certified and locked down used secret code so
that the purchaser can't verify the code?

Apparently the only restrictions ever permitted are the ones the FSF
thinks of.

> > So what would happen if some company was to make software for a tivo and
> > released their binaries signed with some specific key, and they released
> > information on how to check this was signed with their key, and then
> > some other companies went and made tivo hardware and decided that they
> > would only allow code signed by the first companies key to run on it,
> 
> I was pretty sure this had been covered in the section about technical
> barriers to modification in the third draft's rationale, but I can't
> find it right now.  http://gplv3.fsf.org/gpl3-dd3-rationale.pdf

So really what the GPL v3 wants to have is to make sure that the user
can reproduce from the sources a bit for bit identical copy of the
binaries?  Too bad compilers that put time stamps and such into the
binary would make that imposible.  I don't think there is any way that
can be written into the GPL that can prevent all loop holes for how to
make signed binaries.

> Anyhow, the argument I read went like: if there's an agreement between
> the parties to do this, then the copyright holder can probably enforce
> the license regardless of the software and hardware distributor being
> different parties, since the software is being distributed with
> information whose purpose is to enable the hardware to deny the user
> the freedom to run modified versions of the software.

There doesn't have to be an agreement.  The software company could just
release specs for a hardware design and let others freely go and build
them from that design.

> However, if there's no such agreement, if the copyright holder has no
> copyright claims over the hardware or works shipped in it, there's
> nothing the copyright holder can do about it, and that's probably how
> it should be, since a copyright license (!= contract) can't possibly
> prohibit people from creating hardware limited in function, it can
> only tell people that, in order for them to have permission to modify
> or distribute the covered work, they must abide by certain conditions.
> And if they don't want to abide by the conditions, and they don't
> manage to obtain a license from the copyright holders that doesn't
> impose conditions they can't accept, they just can't modify or
> distribute the work.

But if the hardware ships with only code that simply waits for the user
to provide some code for it to isntall (which has to be signed in a way
the hardware likes), then the hardware has nothing to do with the
license of the software.

The signed binaries from the service provider/software developer on the
other hand is GPL and the sources are released with changes.  They just
happen to sign their binaries in a way that allows them to install on
the hardware in question.  It could also install on hardware that
doesn't check the signature as long as it is functionally identical
otherwise.

I hope no one does this, but I still don't see how the GPLv3 draft deals
with this case, or even how it could deal with it.

--
Len Sorensen

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

[Tomas Neme]

> > as long as this right is not used by the software distributor to
> > impose restrictions on the user's ability to adapt the software to
> > their own needs.  The GPLv3 paragraph above makes a fair concession in
> > this regard, don't you agree?
>
> no, one of the rules for the network is that the software must be
> certified, you are requireing the device to permit the software to be
> changed to an uncertified version.(to store credit card numbers and send
> them to a third party for example)

Also another way of doing this is having every network ask the kernel
for its key, and checking it. If it doesn't match a certified key,
then not allowing you to access the network.

Besides the fact that this would be a very costly approach, having
every network needing to update their certified keys list every time
TiVo and every other DVR vendor updates their kernels, it would also
prevent any form of modified software to give you any of the TiVo's
expected functionality: it would load, you would be able to play pong
on it, but not watch or record TV, and they can't be blamed for it,
because if the kernel's been tampered with, it might have been made so
it saves the video unencrypted on the Harddrive, and it certainly *is*
the network's right to stop you from doing so. So what the fuck do you
want from them?

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Alan Cox]

> > You've made an important mistake. You said "their system". Now its "our
> > code" and "whoever bought the units' hardware" so it isn't their anything.
> 
> Yes, the hardware belongs to the user, and the software belongs to the Linux 
> community.   However I think I wasn't 100% clear, I also mean keeping companies 
> networks and content secured.  Credit card companies insuring the software 
> hasn't been modified to skim cards (not that it's the only way to skim a card), 

If credit card companies are doing this they are failing badly and it
clearly isn't working. Also lets be clear about this - I don't need any
credit card company network access to skim older cards, and the newer
ones have been broken by various non software schemes.

>   or Tivo making sure that their content providers are protected.  Lets look at 
> the credit card example.  Sure the user could modify the system and boot their 
> own kernel, but it doesn't have to play nice with Mastercard's network anymore. 
>   Or better yet, would actually report that a certain business's card reader had 
> been tampered with.

That to me is a fair comment. I should IMHO be able to load my code and
my keys on my Tivo. And Walt Disney in return probably should be quite
free not to trust my keys. You need some fairly strong competition law
enforcement to make all that work right in the marketplace but as a
philosophical basis it seems fine. In practice it is likely to lead to
serious monopoly abuse problems and all sorts of ugly tying of goods that
you don't want in a free market. Given the completely ineffectual way the
US enforces its anti-monopoly law, and the slowness of the EU at it the
results might well be bad - but for other reasons.

Alan

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

[Alan Cox]

On Wed, 20 Jun 2007 12:55:10 -0700
"David Schwartz" <davids@webmaster.com> wrote:

> 
> > The kernel you build from the source code that Tivo distributes must
> > be accepted by Tivo's hardware without making other modifications (to
> > Tivo's hardware or bootloader).  If that is possible, I will retract
> > what I said.  If it is not possible, they are omitting part of the
> > program's source code:
> >
> >     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.
> >     -- US Code, Title 17, Section 101
> 
> A key is a number. A signature is a number. They are neither statements nor
> instructions. The argument that GPLv2 prohibits Tivoization is really and
> truly absurd. It has neither a legal nor a moral leg to stand on.

A computer program is a number too.

Alan

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

[Bernd Schmidt]

Greg KH wrote:
> On Sun, Jun 17, 2007 at 02:56:24AM -0300, Alexandre Oliva wrote:
>> If you want your opinions to stand a chance to make a difference, the
>> right place to provide them is gplv3.fsf.org/comments, and time is
>> running short.
[...]
> So, why would we want to waste our time filling out web forms after
> that?

In case anyone was wondering if the FSF is genuinely interested in 
feedback - I went and made some comments on the draft, and they appear 
to no longer be there a few days later.  Thanks for the invitation Alex.


Bernd

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

[Tim Post]

On Wed, 2007-06-20 at 22:30 -0700, david@lang.hm wrote:

> 
> asking a device that's running software that you haven't verified to give 
> you a checksum of itself isn't going to work becouse the software can just 
> lie to you.
> 

I don't think there is any way I _could_ make a device if it had to be
tamper proof and use free software if that was the case. 

I'd need to make some kind of proprietary network connection back to my
company that used its own network device. I could not trust it if the
free kernel could touch it, if I wanted to allow a modified in place
kernel.

If I hope for that device to use the internet to talk to me (i.e. just a
secondary nic), I'd have to write my own kernel to power this second
network device that was capable of encrypting and validating traffic
over tcp-ip. Or I have to pull my own copper to every location where my
device is used.

So either way, I'm writing my own kernel if I want to do that, because I
could not POSSIBLY allow the kernel talking to my private connection to
the device to be modified. 

What a nasty, vicious cycle that would be. Yikes!
 

--Tim

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

[Manu Abraham]

Bernd Petrovitsch wrote:
> On Wed, 2007-06-20 at 18:14 -0300, Tomas Neme wrote:
> [....]
>> Why, if you let user-compiled kernels to run in a TiVo, it might be
>> modified so the TiVo can be used to pirate-copy protected content,
> 
> Or it might be modified to fix a bug - either a technical one or a legal
> one as described below.
> 
>> which is a serious security hole. TiVo would need to read, approve of,
> 
> "Pirate copying" is forbidden anyways in almost every jurisdiction
> AFAIK. Perhaps we should disallow cars on the streets since one could
> drive too fast with them.
> 

Pirate copying should not be a reason to keep things closed as there are
better methods to keep things open, yet provide a _not_ free service.
But that would be upto the vendor how/what they wish to do rather than
we talking about it. Which would be of no use.

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

[Bernd Petrovitsch]

On Wed, 2007-06-20 at 18:14 -0300, Tomas Neme wrote:
[....]
> Why, if you let user-compiled kernels to run in a TiVo, it might be
> modified so the TiVo can be used to pirate-copy protected content,

Or it might be modified to fix a bug - either a technical one or a legal
one as described below.

> which is a serious security hole. TiVo would need to read, approve of,

"Pirate copying" is forbidden anyways in almost every jurisdiction
AFAIK. Perhaps we should disallow cars on the streets since one could
drive too fast with them.

And it is not a security hole for the owner of the hardware (I consider
secret keys somewhere else a much greater security threat) to the Linux
community or a lot of other entities. Probably just music industry
thinks like above.

And there are legislations were it is *legal* to make private copies
(for sure as long as you don't pass them on and somewhere even giving
away for nothing is legal). So I actually have a *right* (which also
can't be killed by contracts) to copy that movie for my private use. And
up to now it is actually legal in .at to do (more or less) everything to
get this right (and that may include hacking the device).

> and sign any modified kernels the users intend to use on their
> hardware. If GPLv3 allows for this, it'd be doing exactly that

	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

[Manu Abraham]

david@lang.hm wrote:

> what sort of signal can the network controller send that couldn't be
> forged by the OS?
> 
> how would you do this where the device is a receiver on the netwoek
> (such as a satellite receiver)

just for the question on the HOWTO (not on anything else)

You can easily have scrambled streams with regards to DVB, eg: using
EN50221. Some (like NDS for example in _some_ instances) use proprietary
schemes, but there are standards based methods also. eg: Common
Scrambling Algorithm (CSA)

But in any case, this can be broken down too .. :-)

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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> no, one of the rules for the network is that the software must be
>> certified,
>
> In this case you might have grounds to enforce this restriction of the
> network on the network controller itself, I suppose.

how would the network controller know if the software has been modified?

> Not that you should disable the network controller entirely (this
> would render the computer useless for many other purpose unrelated
> with blocking connections to your network).
>
> You could instead arrange for the network controller to send some
> signal that enables the network to recognize that the device is
> running certified software.

what sort of signal can the network controller send that couldn't be 
forged by the OS?

how would you do this where the device is a receiver on the netwoek (such 
as a satellite receiver)

David Lang

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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> frankly, I haven't checked the licenses on the software. I'd suggest
>> going to www.tivo.com/linux and download all the source for all the
>> different versions there.
>
> Yeah, thanks, I remembered someone had posted that URL the second
> after a hit Send :-(
>
> I've already got cmd.tar.gz and I'm looking at it now.
>
> Thanks again,
>
>> by the way, it looks like there is one wireless driver that they ship
>> in some releases but don't provide the source for. but if you plan on
>> going after them for that you better go after everyone who ships
>> binary kernel modules.
>
> Only copyright holders of Linux can go after them on matters of kernel
> drivers.  Or is this driver derived from any software copyrighted by
> myself?  Or did you mean the FSF, with whom I'm not associated in any
> way other than ideologically?

from the page listed above

NOTE: 4.0.1a and later releases contain Atmel WLAN drivers in addition to 
the other network adapter drivers that are posted here. These WLAN drivers 
were received by TiVo directly from Atmel and are under the terms of a 
formal non-disclosure agreement, so we cannot publish them under the terms 
of the GPL version 2. Atmel has since released a different version of the 
drivers under the GPL version 2; however, that release does not change our 
prior obligation with Atmel. At this time, we have no plans to use Atmel's 
publicly released drivers in our development.

when I talked about 'going after tivo' I wasn't just refering to doing so 
with lawsuits. the way some people in this thread have acted I could see 
people trying to use this as 'the smoking gun' that proves that the evil 
tivo people didn't release everything they were supposed to.

modules are a grey area, some are clearly derived from the kernel and some 
are just as clearly not derived from the kernel, most are not as clear. I 
don't know the situation for this particular module, but that has nothing 
to do with this topic.

David Lang

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> no, one of the rules for the network is that the software must be
> certified,

In this case you might have grounds to enforce this restriction of the
network on the network controller itself, I suppose.

Not that you should disable the network controller entirely (this
would render the computer useless for many other purpose unrelated
with blocking connections to your network).

You could instead arrange for the network controller to send some
signal that enables the network to recognize that the device is
running certified 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 21, 2007, david@lang.hm wrote:

> frankly, I haven't checked the licenses on the software. I'd suggest
> going to www.tivo.com/linux and download all the source for all the
> different versions there.

Yeah, thanks, I remembered someone had posted that URL the second
after a hit Send :-(

I've already got cmd.tar.gz and I'm looking at it now.

Thanks again,

> by the way, it looks like there is one wireless driver that they ship
> in some releases but don't provide the source for. but if you plan on
> going after them for that you better go after everyone who ships
> binary kernel modules.

Only copyright holders of Linux can go after them on matters of kernel
drivers.  Or is this driver derived from any software copyrighted by
myself?  Or did you mean the FSF, with whom I'm not associated in any
way other than ideologically?

-- 
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

[Manu Abraham]

Lennart Sorensen wrote:
> On Tue, Jun 19, 2007 at 07:28:22PM +0400, Manu Abraham wrote:
>> Well, it is not Tivo alone -- look at http://aminocom.com/ for an
>> example. If you want the kernel sources pay USD 50k and we will provide
>> the kernel sources, was their attitude.
> 
> Hmm, set top boxes are often rented from the cable company rather than
> sold.  Stupid grey area for sure.  At least tivo does give you the
> sources, without demanding more money.  Rather big difference.

I am not talking about the rented aspect, since these STB's are usually
sold rather than rented.

>> Well, it is not Tivo alone, a large chunk of the vendors do that. The
>> vendors who actually do it the clean way are just few and can be counted
>> very easily.
> 
> Well at least where I work we don't try to lock down the hardware, we do
> contribute our changes and bug fixes to upstream when it makes sense
> (and where our changes wouldn't make sense for upstream, they are still
> clearly included with the sources we have.)  If a customer wants a copy
> of the sources, they will get a nice DVD, although strangely none have
> asked for one yet.


Providing the changes back itself is a great thing altogether.

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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> how exactly can they prevent a system that's been tampered with from
>> accessing their network?
>
> By denying access to their servers?  By not granting whatever is
> needed to initiate network sessions?
>
> And note, "it's been tampered with" is not necessarily enough of a
> reason to cut someone off, it has to meet these requirements:

how can the server tell if it's been tampered with?

>>   when the modification itself materially and adversely affects the
>>   operation of the network or violates the rules and protocols for
>>   communication across the network.
>
>> (something even you say they have a right to do)
>
> as long as this right is not used by the software distributor to
> impose restrictions on the user's ability to adapt the software to
> their own needs.  The GPLv3 paragraph above makes a fair concession in
> this regard, don't you agree?

no, one of the rules for the network is that the software must be 
certified, you are requireing the device to permit the software to be 
changed to an uncertified version.(to store credit card numbers and send 
them to a third party for example)

David Lang

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

[david]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 21, 2007, david@lang.hm wrote:
>
>> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>>> On Jun 20, 2007, david@lang.hm wrote:
>>>
>>>> but the signature isn't part of the kernel, and the code that checks
>>>> the signature is completely independant.
>>>
>>> Well, then remove or otherwise mangle the signature in the disk of
>>> your TiVo DVR and see at what point the boot-up process halts.
>
>> I have actually done exactly that. what happens is that the bootloader
>> detects a problem and switches to the other active partition ane
>> reboots. ir neither of the boot partitions meet the requirements the
>> cycle will continue forever.
>
> Oh, too bad.  That must be a bug in the boot loader, right? :-)
>
>
> BTW, since you got a TiVo...  I'm writing an article on Tivoization,
> could you (or anyone else) please help me get information as to what
> other GPLed programs or libraries TiVo includes in their devices?
>
> Thanks in advance,

frankly, I haven't checked the licenses on the software. I'd suggest going 
to www.tivo.com/linux and download all the source for all the different 
versions there.

by the way, it looks like there is one wireless driver that they ship in 
some releases but don't provide the source for. but if you plan on going 
after them for that you better go after everyone who ships binary kernel 
modules.

David Lang

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

[Alexandre Oliva]

On Jun 21, 2007, david@lang.hm wrote:

> how exactly can they prevent a system that's been tampered with from
> accessing their network?

By denying access to their servers?  By not granting whatever is
needed to initiate network sessions?

And note, "it's been tampered with" is not necessarily enough of a
reason to cut someone off, it has to meet these requirements:

>   when the modification itself materially and adversely affects the
>   operation of the network or violates the rules and protocols for
>   communication across the network.

> (something even you say they have a right to do)

as long as this right is not used by the software distributor to
impose restrictions on the user's ability to adapt the software to
their own needs.  The GPLv3 paragraph above makes a fair concession in
this regard, don't you agree?

-- 
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 21, 2007, david@lang.hm wrote:

> On Thu, 21 Jun 2007, Alexandre Oliva wrote:
>> On Jun 20, 2007, david@lang.hm wrote:
>> 
>>> but the signature isn't part of the kernel, and the code that checks
>>> the signature is completely independant.
>> 
>> Well, then remove or otherwise mangle the signature in the disk of
>> your TiVo DVR and see at what point the boot-up process halts.

> I have actually done exactly that. what happens is that the bootloader
> detects a problem and switches to the other active partition ane
> reboots. ir neither of the boot partitions meet the requirements the
> cycle will continue forever.

Oh, too bad.  That must be a bug in the boot loader, right? :-)


BTW, since you got a TiVo...  I'm writing an article on Tivoization,
could you (or anyone else) please help me get information as to what
other GPLed programs or libraries TiVo includes in their devices?

Thanks in advance,

-- 
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]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 20, 2007, david@lang.hm wrote:
>
>> On Wed, 20 Jun 2007, Alexandre Oliva wrote:
>>> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
>>>
>>> On Jun 20, 2007, Andrew McKay <amckay@iders.ca> wrote:
>>>
>>>> However, I don't see how this would ever require a company like Tivo
>>>> or Mastercard to have their networks play nice with a unit that has
>>>> been modified by the end user, potentially opening up some serious
>>>> security holes.
>>>
>>> Which is why the GPLv3 doesn't make the requirement that you stated.
>
>> so if the BIOS checked the checksum of the boot image and if it found
>> it wasn't correct would disable the video input hardware but let you
>> boot the system otherwise it would be acceptable to you and the GPLv3?
>
> I don't think so, but IANAL.  What do you think?  Here's what I
> think to be the relevant passages.
>
>  [...] 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.
>
>  [...]
>
>  The requirement to provide Installation Information does not include
>  a requirement to continue to provide support service, warranty, or
>  updates for a work that has been modified or installed by the
>  recipient, or for the User Product in which it has been modified or
>  installed.  Network access may be denied when the modification
>  itself materially and adversely affects the operation of the network
>  or violates the rules and protocols for communication across the
>  network.

Ok, so if refusing to run software that's tampered with isn't acceptable, 
and disabling the hardware that would be needed to talk on the network 
isn't acceptable. how exactly can they prevent a system that's been 
tampered with from accessing their network? (something even you say they 
have a right to do)

asking a device that's running software that you haven't verified to give 
you a checksum of itself isn't going to work becouse the software can just 
lie to you.

you claim they have this right, but then claim to prohibit every possible 
method of them excercising that right.

pick one side or the other, you don't get both.

David Lang

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

[Alexandre Oliva]

On Jun 20, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> It allows everybody do make that choice that I consider to be really 
> important: the choice of how something _you_ designed gets used.

> And it does that exactly by *limiting* the license to only that one work. 
> Not trying to extend it past the work.

Actually, the two paragraphs above are contradictory, and the second
is not true in as much as it gives way to the former.

Consider an independent file contributed to Linux.  It is an
independent work.  But the moment it gets combined with Linux, the
only license you can use is GPLv2.

Consider a patch, or any modified version of Linux.  It's another
work.  But the license applies to it as well.

Similarly, the GPL affects patents that somehow cover the work: the
distributor can no longer enforce them against downstream users of the
software.

See?

> The GPLv3 can never do that.

It does it in just the same way that GPLv1 and GPLv2 do: "no further
restrictions".  That it has to make some of them explicit, such that
people understand they apply, or such that this provision can't be
trampled on by other laws that by default trample copyright law, is
just a legal implementation detail.

> It's missing the point that "morals" are about _personal_ choices. You 
> cannot force others to a certain moral standpoint. 

FWIW, I tend to consider morals more of a society issue than an
individual issue.  Morals encode what the society understands to be
the common good, and that's often something evolutionary, even with
biological roots.

Laws (as you say) try to reflect the morals of the society, but since
it's based on morals, it often lags behind.  Which is why some laws
become forgotten and no longer applied, even if still applicable in
theory.  And that's also why (as you alluded to in your message), when
laws actively diverge from morals, you observe a lot of civil
disobedience: think DRM, DMCA, non-benevolent dictatorships and other
abusive regimes.


I must say that it *is* lovely to watch you talk about morals in ways
that I agree so much with, even if I dissent in a some details.  This
has further increased my admiration for you.  Thank you.

-- 
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 20, 2007, "Jesper Juhl" <jesper.juhl@gmail.com> wrote:

> On 18/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> Your analysis stopped at the downside of prohibiting tivoization.  You
>> didn't analyze the potential upsides,

> Maybe that's because I don't really see any up sides.

You do:

> a few vendors that currently tivoize hardware may open up their
> hardware but I doubt that will be very many

You just don't think they'd prevail over the downsides.  *This* is an
opinion I can respect, even if it's as much of a guestimate as mine.
I'm sure both are highly influenced by personal opinions, wishful
thinking and fears.  This is all very human.

>> so you may indeed come to different conclusions, and they may very
>> well be wrong.

> Just because I come to a different conclusion than you doesn't
> nessesarily make it wrong.

Agreed.  I didn't say they were.  I said they could be.  Can you prove
they're right?  Do you even have any supporting evidence to back your
guestimates?  Heck, you may even have more than I do.

I openly admit mine is mostly theoretical.  I extrapolate the initial
success of GNU+Linux on the PC environment, due in a large part to the
ability for users to tinker with their computers, and expect it not to
be so significantly different for other kinds of computers.

For sure you'll get a far lower *percentage* of hackers in consumer
devices than on PCs, whose users used to be far more
technically-inclined and thus more propense to become hackers when
GNU+Linux started than these days.

But then I think of all of these computer users who helped make
GNU+Linux what it is today, and other hackers that hadn't discovered
this inclination before because they haven't had access to hackable
computers.  They could be tinkering with their DVRs, cell phones,
wireless routers et al, and bringing the same kind of exciting
community development to these kinds of computers.

I'm saddened that the major Linux developers are willing to trade all
of this (which I openly admit may be just a figment of my imagination,
or just a tip of an iceberg) for some professional contributions
(good) and some additional exposure that won't do justice to their
software (bad), because these users will miss a big part of the
picture by not being able to tinker with the software in the
environment where they use the software.

>> It's very human to look only at the potential downside of an action
>> and conclude it's a bad action.

> And you believe yourself to be immune to that - right?

Last I looked, I was still human.  So no.  I try to use logic to
reason out such behaviors when I realize they might be in action.
But, as the saying goes, logic is a tool we use to justify our
intutions.  Or, logical reasoning is a tool to make the wrong
decisions with a greater amount of confidence ;-)

>> You can create the device using GPLv3 software in it.

> Not as long as I want to prevent the user from tampering with it, no.

<broken record>mumble ROM mumble</broken record>

> But do you really expect a vendor to put a device on the market where
> they also lock themselves out of upgrading it and releasing new
> software for it?

Depends on how badly they want to use the GPLed software.

Don't you guys think Linux is so technically superior that some
vendors might prefer to stick with it (should it move to GPLv3, or
tivoization be found to be already forbidden by GPLv2 in a US court or
elsewhere) even if this means going to ROM or respecting users'
freedoms?

Or are Linux advantages so thin and fragile (if they exist at all)
that you're just hoping nobody realizes there are better choices out
there, and you're desperate for vendors not to realize this?

> For a few select individuals that may be true. But for the majority of
> the population it won't mean a thing.

Agreed.  You're thinking of percentages (fewer percent hackers among
consumers of user products than among PC users).  I'm thinking all
hackers in PCs could become hackers of such devices as well, and then
some.

>> This is the upside that you left out from your analysis, and from
>> every other analysis that set out to "prove" that anti-tivoization is
>> bad that I've seen so far.

> I'm sorry, but I don't think it holds water.

Fair enough.

-- 
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 20, 2007, david@lang.hm wrote:

> On Wed, 20 Jun 2007, Alexandre Oliva wrote:
>> We already know the vendor doesn't care about the user, so why should
>> we take this into account when analyzing the reasoning of the vendor?

> no, we don't know this. you attribute the reason for the lockdown to
> be anti-user. others view it as being pro-user becouse it lets the
> user get functionality that they wouldn't have access to otherwise.

Yeah, yeah, now please put your hands to the back such that I can
place your handcuffs, such that you can't use your hands to kill
someone.  Well, you might still be able to use your hands for this
purpose if you're really clever, or you can kill people by other
means.  But at least I'll be able to claim that I did my part.

:-/

-- 
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 20, 2007, david@lang.hm wrote:

> On Wed, 20 Jun 2007, Alexandre Oliva wrote:
>> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
>> 
>> On Jun 20, 2007, Andrew McKay <amckay@iders.ca> wrote:
>> 
>>> However, I don't see how this would ever require a company like Tivo
>>> or Mastercard to have their networks play nice with a unit that has
>>> been modified by the end user, potentially opening up some serious
>>> security holes.
>> 
>> Which is why the GPLv3 doesn't make the requirement that you stated.

> so if the BIOS checked the checksum of the boot image and if it found
> it wasn't correct would disable the video input hardware but let you
> boot the system otherwise it would be acceptable to you and the GPLv3?

I don't think so, but IANAL.  What do you think?  Here's what I
think to be the relevant passages.

  [...] 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.

  [...]

  The requirement to provide Installation Information does not include
  a requirement to continue to provide support service, warranty, or
  updates for a work that has been modified or installed by the
  recipient, or for the User Product in which it has been modified or
  installed.  Network access may be denied when the modification
  itself materially and adversely affects the operation of the network
  or violates the rules and protocols for communication across the
  network.

-- 
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 20, 2007, "Tomas Neme" <lacrymology@gmail.com> wrote:

>> > However, I don't see how this would ever require a company like Tivo
>> > or Mastercard to have their networks play nice with a unit that has
>> > been modified by the end user, potentially opening up some serious
>> > security holes.
>> 
>> Which is why the GPLv3 doesn't make the requirement that you stated.

> Why, if you let user-compiled kernels to run in a TiVo, it might be
> modified so the TiVo can be used to pirate-copy protected content,

And then the user who uses such features in ways not permitted by the
copyright holders are committing a crime.  They can be prosecuted by
the copyright holders and convicted of the crime.

That TiVo can somehow become liable for this just shows how broken the
legal system in the US is.  It's like making a knife manufacturer
liable for a killing using a knife they made, just because the knife
didn't have technical measures intended to prevent the knife from
being used to kill people.

-- 
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 20, 2007, david@lang.hm wrote:

> On Wed, 20 Jun 2007, Alexandre Oliva wrote:
>> On Jun 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:
>>>> It is the duty of the FSF to defend these freedoms.  It's its public
>>>> mission.  That's a publicly stated goal of the GPL, for anyone who
>>>> cares to understand it, or miss it completely and then complain about
>>>> changes in spirit.
>> 
>>> I wouldn't call it a duty.  It is the chosen mission perhaps, but nobody
>>> is making them do it.
>> 
>> Everyone who donates to it does so understanding what the mission is.
>> Detracting from that mission would be failing the public commitment.

> true, but selecting the GPL as the license for your project is not
> donating to the FSF.

Oh, that's what you meant.  Indeed, absolutely not.

The GPL is "just" a set of permissions you, as an author, grant to
anyone who comes across your program.

Whether you share FSF's goals or not, you can do that.

If you share FSF's goals of not only respecting users' freedoms, but
also defending them as much as deemed legally possible under copyright
law, you can also offer your code under any later version of the GPL,
such that it remains usable by the community who cares about this.

In theory, this shouldn't be a problem for anyone who chose the GPLv2,
since all of the permissions granted by GPLv3 are granted by GPLv2,
and this is how it should be.  The difference is that GPLv3 plugs some
holes that were found in GPLv2, in a similar way that GPLv2 plugged
holes found in GPLv1, and GPL "plugs holes" in LGPL, which "plugs
holes" in other even more permissive licenses.

Each GPL revision is expected to plug holes ("address new problems",
as in the legal terms of GPLv2) that might enable licensees to deny
other licensees the rights you meant to grant them.

This will necessarily make each revision incompatible with the
previous, for being stricter, thus imposing further restrictions, even
if only by removing exploitable ambiguities.  This should have been
clear since GPLv1, anyone who understands the goals of the GPL and
with enough foresight to understand the recommendation of permitting
relicensing under newer versions should be able to see this.

So, since new restrictions are always on licensees' ways to deny other
licensees the enjoyment of the permissions you meant to grant them, if
you mean to permit people to use your work in the ways permitted by
GPLv2, not permitting them to be used in GPLv3 software amounts to
pure selfishness: "if I won't get to use your code, you don't get to
use mine."  Tit-for-tat, for sure, but certainly not in the spirit of
sharing clearly established early in the preamble of every version of
the GPL.

Permitting such relicensing wouldn't deny anyone any freedom, and it
wouldn't create any obligations whatsoever for the licensor.  Whoever
wanted to use the work under the more liberal terms of the earlier
version of the GPL under which the work was licensed still could: this
license can't be unilaterally revoked, not by you, not by anyone else.

-- 
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]

On Thu, 21 Jun 2007, Alexandre Oliva wrote:

> On Jun 20, 2007, david@lang.hm wrote:
>
>> but the signature isn't part of the kernel, and the code that checks
>> the signature is completely independant.
>
> Well, then remove or otherwise mangle the signature in the disk of
> your TiVo DVR and see at what point the boot-up process halts.

I have actually done exactly that. what happens is that the bootloader 
detects a problem and switches to the other active partition ane reboots. 
ir neither of the boot partitions meet the requirements the cycle will 
continue forever.

David Lang

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

[Alexandre Oliva]

On Jun 20, 2007, "David Schwartz" <davids@webmaster.com> wrote:

> A key is a number. A signature is a number.

And a program is a number.

http://asdf.org/~fatphil/maths/illegal.html

Your point?

-- 
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

[Michael Poole]

david@lang.hm writes:

> On Wed, 20 Jun 2007, Michael Poole wrote:
>
>> david@lang.hm writes:
>>
>>> if the GPL can excercise control over compilations, then if Oracle
>>> were to ship a Oracle Linux live CD that contained the Oracle Database
>>> in the filesystem image, ready to run. then the GPL would be able to
>>> control the Oracle Database code.
>>
>> By copyright law, it could.  By its language, it does not.
>
> many people (including many lawyers will disagree that it could by
> copyright law

On what grounds would Oracle have a license to ship that part of
Linux?  Unless you are the sole copyright owner, you have no right to
copy a given piece of software _at all_ without a license.  The GPL is
a remarkably giving license in terms of how little it requires.

(This potentially wide scope is one of the major reasons that the GPL
mentions "mere aggregation" and that the Debian Free Software
Guidelines' include guideline #9.)

>>> if the GPL can't do this then it can't control the checksum either.
>>>
>>> again, it's not just the kernel that's part of the checksum on a tivo,
>>> the checksum is over the kernel + initial filesystem, much of which
>>> contains code not covered by the gPL)
>>
>> Again, did you miss where I pointed out that this makes it *worse* for
>> Tivo, because they are tying together -- and making inseparable -- a
>> combination that would otherwise be "mere aggregation"?
>
> and it makes most distro CD's illegal since they contain code under
> different incompatible licenses and they make a checksum across the
> entire CD image.

When distributions generate checksums (as opposed to signatures) over
images, they do provide all of the inputs.  When most distributions
provide signatures, the signatures do not function as statements or
instructions to computers -- they function as statements to users.
Tivo's digital signatures differ in both of these respects.

Michael Poole

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

[Alexandre Oliva]

On Jun 20, 2007, david@lang.hm wrote:

> but the signature isn't part of the kernel, and the code that checks
> the signature is completely independant.

Well, then remove or otherwise mangle the signature in the disk of
your TiVo DVR and see at what point the boot-up process halts.

-- 
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]

On Wed, 20 Jun 2007, Michael Poole wrote:

> david@lang.hm writes:
>
>> if the GPL can excercise control over compilations, then if Oracle
>> were to ship a Oracle Linux live CD that contained the Oracle Database
>> in the filesystem image, ready to run. then the GPL would be able to
>> control the Oracle Database code.
>
> By copyright law, it could.  By its language, it does not.

many people (including many lawyers will disagree that it could by 
copyright law

>> if the GPL can't do this then it can't control the checksum either.
>>
>> again, it's not just the kernel that's part of the checksum on a tivo,
>> the checksum is over the kernel + initial filesystem, much of which
>> contains code not covered by the gPL)
>
> Again, did you miss where I pointed out that this makes it *worse* for
> Tivo, because they are tying together -- and making inseparable -- a
> combination that would otherwise be "mere aggregation"?

and it makes most distro CD's illegal since they contain code under 
different incompatible licenses and they make a checksum across the entire 
CD image.

David Lang

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

[Alexandre Oliva]

On Jun 20, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> And anybody who thinks others don't have the "right to choice", and then 
> tries to talk about "freedoms" is a damn hypocritical moron.

Yeah, it is indeed possible to twist it such that it sounds bad.

The important point is that one's freedom ends where another's
starts.  Unlimited freedom would be freedom to trample over others'
freedoms too, and that's not right.  That's why freedom of choice has
to be used with care.  Even fundamental human rights sometimes clash
with each other.

We don't fight for the freedoms as goals in themselves.  We fight for
them because we understand they're essential for the common good.

-- 
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

[Michael Poole]

david@lang.hm writes:

> if the GPL can excercise control over compilations, then if Oracle
> were to ship a Oracle Linux live CD that contained the Oracle Database
> in the filesystem image, ready to run. then the GPL would be able to
> control the Oracle Database code.

By copyright law, it could.  By its language, it does not.

> if the GPL can't do this then it can't control the checksum either.
>
> again, it's not just the kernel that's part of the checksum on a tivo,
> the checksum is over the kernel + initial filesystem, much of which
> contains code not covered by the gPL)

Again, did you miss where I pointed out that this makes it *worse* for
Tivo, because they are tying together -- and making inseparable -- a
combination that would otherwise be "mere aggregation"?

Michael Poole

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> David Schwartz writes:
>
>>> However, compilations (even to the extent they are creative
>>> combinations) are not necessarily derivative works of their elements.
>>> For more details, see
>>> http://www.copyright.gov/circs/circ14.html#compilations
>>
>> Because compilation copyrights don't really affect the Tivo and GPLv2/GPLv3
>> issue, I tend to ignore them when discussing that subject. If you think I'm
>> wrong and there is some relationship between them, please let me know. I
>> admit I may not have given that possibility enough thought.
>
> I believe compilation copyrights do bear on GPL-licensed software, by
> virtue of the GPL's sentence "[...] rather, the intent is to exercise
> the right to control the distribution of derivative _or collective_
> works based on the Program." (emphasis added).
>
> There is a lot of grey and/or arguable area about what constitutes a
> GPL-encumbered collective work versus mere aggregation.  Although I
> disagree, I understand and respect that some believe that the kernel
> plus a digital signature over it is "mere aggregation".  I would like
> to focus the discussion on that question, though, rather than whether
> the GPL is worded to control the rights to compilations-in-general
> that include GPLed works.

if the GPL can excercise control over compilations, then if Oracle were to 
ship a Oracle Linux live CD that contained the Oracle Database in the 
filesystem image, ready to run. then the GPL would be able to control the 
Oracle Database code.

if the GPL can't do this then it can't control the checksum either.

again, it's not just the kernel that's part of the checksum on a tivo, the 
checksum is over the kernel + initial filesystem, much of which contains 
code not covered by the gPL)

David Lang

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

[Michael Poole]

David Schwartz writes:

>> There is a lot of grey and/or arguable area about what constitutes a
>> GPL-encumbered collective work versus mere aggregation.
>
> I think it's technically/legally clear what the standards are, but certainly
> arguable whether particular works meet that standard. If the choice of works
> to combine is sufficiently creative (above and beyond any choices dictated
> by functional considerations), it's a GPL-encumbered collective work.
>
> I don't think it's arguable that a signature shipped along with a binary is
> a collective work. In any event, if that were true, I think we should be
> able to agree that Linus would be required to release his kernel signing
> keys.

The distinction between GPL-covered works and "mere aggregation" is
not a function only of legal classifications.  If it were, the GPL
would be worded differently than it is -- and have different effects
than most people believe it does.

Michael Poole

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

[David Schwartz]

> I believe compilation copyrights do bear on GPL-licensed software, by
> virtue of the GPL's sentence "[...] rather, the intent is to exercise
> the right to control the distribution of derivative _or collective_
> works based on the Program." (emphasis added).

Ahh, good. So there's no problem with the GPL. I already thought it the most
sensible reading that it included collective works (and it's clear that it
can legally do so), but that makes it clear.

I didn't mean that compilation copyrights have no relevance to GPL issues in
general, just none to the issue of GPLv2 versus GPLv3 and Tivoization. Are
you going to argue that there's a compilation copyright justified by
combining a kernel binary with a signature for that binary? That seems
untenable to me.

> There is a lot of grey and/or arguable area about what constitutes a
> GPL-encumbered collective work versus mere aggregation.

I think it's technically/legally clear what the standards are, but certainly
arguable whether particular works meet that standard. If the choice of works
to combine is sufficiently creative (above and beyond any choices dictated
by functional considerations), it's a GPL-encumbered collective work.

I don't think it's arguable that a signature shipped along with a binary is
a collective work. In any event, if that were true, I think we should be
able to agree that Linus would be required to release his kernel signing
keys.

> Although I
> disagree, I understand and respect that some believe that the kernel
> plus a digital signature over it is "mere aggregation".

It clearly is. What else could it be? The digital signature is a separate
item, a pure number for which there is no copyright interest, that is simply
appended to the kernel. It does not contain significant protected elements
of the kernel. No creative process is used to generate it or attach it. The
decision to append is dictated by purely functional considerations (nobody
creatively picks which signature to bundle with which kernel).

> I would like
> to focus the discussion on that question, though, rather than whether
> the GPL is worded to control the rights to compilations-in-general
> that include GPLed works.

If the kernel plus a digital signature over it is not mere aggregation, then
Linus is violating the GPL by shipping kernel plus digital signatures but
not including the "source code" to produce the signatures. If the
ridiculousness of that is not sufficiently obvious, I'm not sure what else
to say. Where is the outcry that Linus is keeping his signing key secret,
failing to include the source code that he used to build the Linux kernel
distibution?

The past few times this has come up, every possible irrelevent side issue
was raised. For example, that the signature in the case of Linux is not
functional. These considerations do not have anything to do with whether the
combination is mere aggregation or not.

Ironically, this case is even clearer than linking. The two works are quite
literally stapled together. There is no intermingling at all.

DS

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

[Michael Poole]

David Schwartz writes:

>> However, compilations (even to the extent they are creative
>> combinations) are not necessarily derivative works of their elements.
>> For more details, see
>> http://www.copyright.gov/circs/circ14.html#compilations
>
> Because compilation copyrights don't really affect the Tivo and GPLv2/GPLv3
> issue, I tend to ignore them when discussing that subject. If you think I'm
> wrong and there is some relationship between them, please let me know. I
> admit I may not have given that possibility enough thought.

I believe compilation copyrights do bear on GPL-licensed software, by
virtue of the GPL's sentence "[...] rather, the intent is to exercise
the right to control the distribution of derivative _or collective_
works based on the Program." (emphasis added).

There is a lot of grey and/or arguable area about what constitutes a
GPL-encumbered collective work versus mere aggregation.  Although I
disagree, I understand and respect that some believe that the kernel
plus a digital signature over it is "mere aggregation".  I would like
to focus the discussion on that question, though, rather than whether
the GPL is worded to control the rights to compilations-in-general
that include GPLed works.

Michael Poole

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

[David Schwartz]

> By "creative combination" do you mean what US copyright law refers to
> as compilations (or their subset collective works)?

Not only. By "creative combination" I mean either a compilation or a
derivative work. I was a bit unclear about that because I wasn't really
addressing compilation rights at the time.

For example, if you adapt a FreeBSD driver to work on Linux, you may be
creatively combining aspects of the driver with code from the kernel. The
result is one that you have copyright interest in because it is a derivative
work but probably not a compilation copyright. The choice of one driver and
one OS, where the goal is to make the driver work on the OS, probably is not
sufficiently creative to justify a compilation copyright. However, this is
clearly not mere aggregation if significant changes are needed to make the
driver work with Linux.

> Compilations can be creative combinations while still being mere
> aggregation under the GPL.  For example, if applications are selected
> to run with a Linux kernel, and they are distributed together, the
> collection is a creative selection -- and this seems to be one of the
> cases evoked by the GPL's reference to "mere aggregation".  See also
> practically every Linux distribution on the planet.

You are quite correct. In this case, the GPL may not require you to license
the compilation copyright. I believe this is so even if all the works are
covered by the GPL. Arguably, that's a defect in the GPL because it means
there might be situations in which you might receive a CD that contains only
GPL'd software and not be able to redistribute it due to a compilation
copyright.

I honestly have no position on whether "mere aggregation" should include
aggregating works where there is sufficient creative input to justify a
compilation copyright on the result. I think either position can be argued.
I think the intent of the GPL was probably that mere aggregation not include
compilation rights because that leads to strange results. I don't know of
any evidence that compilation rights were considered when the GPL was
written. If so, the deliberate lack of mention might weigh in the balance.

> Compilations also can be creative combinations and *more* than mere
> aggregation: for example, Linux with respect to its subsystems, or any
> case where a larger work is derivative of one of its components.

Of course. If I write a Linux kernel module, it might be a derivative work
because it contains significant portions of the Linux kernel source code.
This is true before anyone compiles it or links it.

When I say linking cannot create a derivative work, I mean assuming the work
was not derivative in the first place. I am also further assuming there is
insufficient creativity in the choice of which works to link to justify a
compilation copyright.

> However, compilations (even to the extent they are creative
> combinations) are not necessarily derivative works of their elements.
> For more details, see
> http://www.copyright.gov/circs/circ14.html#compilations

Because compilation copyrights don't really affect the Tivo and GPLv2/GPLv3
issue, I tend to ignore them when discussing that subject. If you think I'm
wrong and there is some relationship between them, please let me know. I
admit I may not have given that possibility enough thought.

DS

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

[Michael Poole]

David Schwartz writes:

>> Most of this list has
>> already dismissed your rather unique -- I would even say frivolous --
>> idea of how far "mere aggregation" goes: I, for one, have better
>> things to do than explain why a C file is not a "mere aggregation" of
>> the functions it contains.)
>>
>> Michael Poole
>
> Of course it's not mere aggregation. The functions in a C file are
> creatively combined. How many times do I have to say that the opposite of
> "mere aggregation" is creative combination?
>
> It is not unique, it is part of the definition of a "derivative work".

By "creative combination" do you mean what US copyright law refers to
as compilations (or their subset collective works)?

Compilations can be creative combinations while still being mere
aggregation under the GPL.  For example, if applications are selected
to run with a Linux kernel, and they are distributed together, the
collection is a creative selection -- and this seems to be one of the
cases evoked by the GPL's reference to "mere aggregation".  See also
practically every Linux distribution on the planet.

Compilations also can be creative combinations and *more* than mere
aggregation: for example, Linux with respect to its subsystems, or any
case where a larger work is derivative of one of its components.

However, compilations (even to the extent they are creative
combinations) are not necessarily derivative works of their elements.
For more details, see
http://www.copyright.gov/circs/circ14.html#compilations

Michael Poole

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

[Dave Neuer]

On 6/20/07, David Schwartz <davids@webmaster.com> wrote:
>
> > This argument is the obvious nonsense. "Runs on TiVO" is a property of
> > the software that TiVO distributes -- such an important property that
> > it would be nonsensical for them to distribute it with their hardware.
> > But they do distribute it, and only the GPL allows them to.
>
> Why does the importance of the property matter to the validity of the
> argument?

>From a legal standpoint, perhaps you're right, it doesn't matter what
the function is. From a moral standpoint it should be obvious to you
that "runs on TiVO" is TiVO's sole motivation to distribute the
software at all, it is "the software" and arguing that they have an
equivalent obligation WRT it as to some incidental thing like Linus'
signing key is just preposterous.

> > > Tivo's choice is an authorization decision. It is similar to
> > > you not having
> > > root access to a Linux box. Sorry, you can't run a modified
> > > kernel on that
> > > machine, but you can still modify the kernel and run it on any hardware
> > > where authorization decisions don't stop you from doing so. The GPL was
> > > never about such authorization decisions.
>
> > Says judge Schwartz. Oops. That's right, you're not a judge in any
> > legal jurisdiction, nor an author of the GPL.
>
> Nice argument. I'm wrong because people can disagree with me.

No, in this case you are wrong because absent authority to decide the
meaning from a dispositive legal standpoint (the law says the license
means this) or knowledge of the intent of the author of the GPL (I the
author intended it to mean this), your statement that the GPL was
"never about" "such decisions" is meaningless, AFAICT.

>
> DS

Dave

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

[Tomas Neme]

On 6/20/07, Dave Neuer <mr.fred.smoothie@pobox.com> wrote:
> On 6/20/07, Tomas Neme <lacrymology@gmail.com> wrote:
> >
> > I'm about this far to Linus'izing my wording and calling you stupid,
> > hypocrite, or bullshitter
>
> Knock yourself out, it will no doubt lend much moral and logic weight
> to your rhetoric.

I might not have the best rhetoric, but I still hold my point about
the credit card. Ask yourself: are you going to complain about Firefox
(GPL'ed) not passing information unencrypted because it stops
potential users (crackers ARE users) from doing what they want to with
it? What's a security issue and what's not is a matter of legality and
it's each part's duty to enforce legality in every way they can (I'm
not saying that I agree, I'm an anarchist, but it's just how it goes).

The content providers do it by not allowing DVRs to work if they're
not secure, and DVRs are secure by doing whatever is legally possible
to avoid crackers from bypassing security measures. On the other hand
is legal for you to bypass those security measures as long as you
don't make illegal use of those bypasses.

The kernel TiVo distributes works on TiVo boxes, The kernel modified
by you, is no longer the kernel TiVo distributes, and therefore the
key that the original kernel had no longer applies to it. Try running
your TiVo kernel on a PC, I think you won't be able to without a lot
of modification.. and then again, once you do the proper modifying,
you will be able to use it on your multimedia computer, and use all of
the wonderful things you DIE to be able to modify the TiVo kernel
for.. If you modify your TiVo kernel and say make it so it doesn't
have an IDE controller module anymore, you won't be able to run it on
your TiVo either.. at least not in any useful way, and would you be
complaining?

And someone said this already, if the signature is created via a known
algorithm, and only the key isn't provided, saying that that's not
GPLv2 compliant is like saying that I can't publish investigation work
that was produced sharing via a secured network unless I also publish
the SSH key I used through investigation, or the original value of the
srand() if the investigation relied on random number generation,
because the exact same results won't be reproducible.

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Dave Neuer]

On 6/20/07, Tomas Neme <lacrymology@gmail.com> wrote:
>
> I'm about this far to Linus'izing my wording and calling you stupid,
> hypocrite, or bullshitter

Knock yourself out, it will no doubt lend much moral and logic weight
to your rhetoric.

Dave

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

[David Schwartz]

> Most of this list has
> already dismissed your rather unique -- I would even say frivolous --
> idea of how far "mere aggregation" goes: I, for one, have better
> things to do than explain why a C file is not a "mere aggregation" of
> the functions it contains.)
>
> Michael Poole

Of course it's not mere aggregation. The functions in a C file are
creatively combined. How many times do I have to say that the opposite of
"mere aggregation" is creative combination?

It is not unique, it is part of the definition of a "derivative work".

DS

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

[David Schwartz]

> This argument is the obvious nonsense. "Runs on TiVO" is a property of
> the software that TiVO distributes -- such an important property that
> it would be nonsensical for them to distribute it with their hardware.
> But they do distribute it, and only the GPL allows them to.

Why does the importance of the property matter to the validity of the
argument?

> Linus' key is not required to use the software Linus distributes under
> the GPL, by contrast.

Why does whether or not the key is required to use the software matter? It
may be impossible to use a Linux kernel on a particular piece of hardware
without the BIOS, that doesn't mean the BIOS source code is part of the
kernel source code even if the kernel is shipped for that hardware.

> > Tivo's choice is an authorization decision. It is similar to
> > you not having
> > root access to a Linux box. Sorry, you can't run a modified
> > kernel on that
> > machine, but you can still modify the kernel and run it on any hardware
> > where authorization decisions don't stop you from doing so. The GPL was
> > never about such authorization decisions.

> Says judge Schwartz. Oops. That's right, you're not a judge in any
> legal jurisdiction, nor an author of the GPL.

Nice argument. I'm wrong because people can disagree with me.

DS

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

[Michael Poole]

David Schwartz writes:

>> I do not say that the BIOS is doing anything (legally) wrong.  The
>> wrong act is distributing the binary kernel image without distributing
>> complete source code for it.
>
> Why are you not complaining that Linus does not distribute the keys he uses
> to sign kernel source distributions? If a digital signature is part of the
> distribution, why is the key used to produce that signature not part of the
> distribution?
>
> If you can cite some legal reason there is a difference, I would be quite
> impressed.
>
> In any event, the argument is obvious nonsense. The signature is merely
> aggregated with the kernel. Cooperation, dependent function, and convergent
> design can't break mere aggregation or you get ridiculous results. (For
> example, a device shipped with the Linux kernel and some applications would
> have to GPL all the applications.)

Do you make it a habit to pose ranty questions to people while neither
attributing their text nor cc'ing them?  Especially when you claim the
person's argument is "obvious nonsense", it seems quite rude.

(Since you have dismissed my argument as nonsense before hearing my
response, I will not bother answering your question.  Since you are
acting like a troll, I will dismiss you as one.  Most of this list has
already dismissed your rather unique -- I would even say frivolous --
idea of how far "mere aggregation" goes: I, for one, have better
things to do than explain why a C file is not a "mere aggregation" of
the functions it contains.)

Michael Poole

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

[Dave Neuer]

On 6/20/07, David Schwartz <davids@webmaster.com> wrote:
>
> > Tomas Neme writes:
>
> > > I have been following this discussion for the last week or so, and
> > > what I haven't been able to figure out is what the hell is the big
> > > deal with TiVO doing whatever they want to with their stupid design.
> > > They made a design, they build a machine, they sell it as is, and
> > > provide source code for GPL'ed software... what's your problem?
>
> > It's simple: they don't provide _complete_ source code.  They keep the
> > source code for the part of their Linux kernel images that provides
> > the functionality "runs on Tivo DVRs".  The GPL requires that
> > distributors of binary versions provide complete source code, not just
> > the parts of source code that are convenient.
> >
> > Michael Poole
>
> That leads to lots of obvious nonsense unless you fix it with all kinds of
> made up ad-hoc changes just to get the result you want. Why doesn't Linus
> have to release the keys he uses to sign the Linux kernel source
> distributions? That provides the functionality "can be proven to be
> authorized by Linus". What you call "runs on Tivo DVRs", I call "can be
> proven to be authorized by Tivo to run on Tivo DVRs".

This argument is the obvious nonsense. "Runs on TiVO" is a property of
the software that TiVO distributes -- such an important property that
it would be nonsensical for them to distribute it with their hardware.
But they do distribute it, and only the GPL allows them to.

Linus' key is not required to use the software Linus distributes under
the GPL, by contrast.

>
> Tivo's choice is an authorization decision. It is similar to you not having
> root access to a Linux box. Sorry, you can't run a modified kernel on that
> machine, but you can still modify the kernel and run it on any hardware
> where authorization decisions don't stop you from doing so. The GPL was
> never about such authorization decisions.

Says judge Schwartz. Oops. That's right, you're not a judge in any
legal jurisdiction, nor an author of the GPL.

Dave

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

[David Schwartz]

> I do not say that the BIOS is doing anything (legally) wrong.  The
> wrong act is distributing the binary kernel image without distributing
> complete source code for it.

Why are you not complaining that Linus does not distribute the keys he uses
to sign kernel source distributions? If a digital signature is part of the
distribution, why is the key used to produce that signature not part of the
distribution?

If you can cite some legal reason there is a difference, I would be quite
impressed.

In any event, the argument is obvious nonsense. The signature is merely
aggregated with the kernel. Cooperation, dependent function, and convergent
design can't break mere aggregation or you get ridiculous results. (For
example, a device shipped with the Linux kernel and some applications would
have to GPL all the applications.)

DS

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

[Andrew McKay]

david@lang.hm wrote:
> On Wed, 20 Jun 2007, Alexandre Oliva wrote:
> 
>> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
>>
>> On Jun 20, 2007, Andrew McKay <amckay@iders.ca> wrote:
>>
>>> However, I don't see how this would ever require a company like Tivo
>>> or Mastercard to have their networks play nice with a unit that has
>>> been modified by the end user, potentially opening up some serious
>>> security holes.
>>
>> Which is why the GPLv3 doesn't make the requirement that you stated.
> 
> so if the BIOS checked the checksum of the boot image and if it found it 
> wasn't correct would disable the video input hardware but let you boot 
> the system otherwise it would be acceptable to you and the GPLv3?
> 
> somehow I doubt it, but that's what it would take to prevent modified 
> software from interacting with their networks (remembering that these 
> networks are the cable and satellite networks in some cases)
> 
> it also seems that if this was the case it would be a trivial 
> work-around for the GPLv3 if it was acceptable.
> 

That is exactly where I was going with that.  A trivial work around in the Tivo 
case.  I guess GPLv4 will have to close up that hole?

Andrew McKay

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

[Andrew McKay]

Alexandre Oliva wrote:
> On Jun 20, 2007, Andrew McKay <amckay@iders.ca> wrote:
> 
>> However, I don't see how this would ever require a company like Tivo
>> or Mastercard to have their networks play nice with a unit that has
>> been modified by the end user, potentially opening up some serious
>> security holes.
> 
> Which is why the GPLv3 doesn't make the requirement that you stated.
> 

So if it's not a requirement of the GPLv3, then Tivo could deny content based 
signing the binary image of the Linux kernel and using that signature as 
authentication on their network (or their content providers network).  A 
modified Tivo box would not be able to preform it's original task of being a PVR 
at that point, at least with the content provider's signal.  Seems pretty 
pointless to me.  Seems like almost the same thing as not allowing an unsigned 
Linux kernel to boot on the system.  Though it would still be possible to get 
the Tivo box to play tetris or something like that.

Andrew McKay

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

[Dave Neuer]

On 6/20/07, Tomas Neme <lacrymology@gmail.com> wrote:
>
> Why, if you let user-compiled kernels to run in a TiVo, it might be
> modified so the TiVo can be used to pirate-copy protected content,

1) It may be far more likely that in the majority of cases it will be
modified with the intent to allow functionality which has no bearing
on copyrighted entertainment copyright or which is permitted under the
Fair Use doctrine. Neither you, nor I, nor anyone else on the list
knows whether that is the case.

2) There are far easier ways to pirate copyrighted entertainment
content (like buy the discs, professional duplicating hardware, and
just dup them) which I would wager is what actual pirates do 99% of
the time

> which is a serious security hole.

If you are a content company it's a security hole, if you are a TiVO
owner, it's a feature.

Dave

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

[david]

On Wed, 20 Jun 2007, Alexandre Oliva wrote:

> On Jun 20, 2007, "Jesper Juhl" <jesper.juhl@gmail.com> wrote:
>
>> On 19/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>>> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>>>> In the GPLv3 world, we have already discussed in this thread how you can
>>>> follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM
>>>> instead of using a flash device.
>>>
>>> Yes.  This is one option that doesn't bring any benefits to anyone.
>>> It maintains the status quo for users and the community, but it loses
>>> the ability for the vendor to upgrade, fix or otherwise control the
>>> users.  Bad for the vendor.
>
>> Also bad for the user
>
> We already know the vendor doesn't care about the user, so why should
> we take this into account when analyzing the reasoning of the vendor?

no, we don't know this. you attribute the reason for the lockdown to be 
anti-user. others view it as being pro-user becouse it lets the user get 
functionality that they wouldn't have access to otherwise.

David Lang

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

[david]

On Wed, 20 Jun 2007, Alexandre Oliva wrote:

> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
> 
> On Jun 20, 2007, Andrew McKay <amckay@iders.ca> wrote:
>
>> However, I don't see how this would ever require a company like Tivo
>> or Mastercard to have their networks play nice with a unit that has
>> been modified by the end user, potentially opening up some serious
>> security holes.
>
> Which is why the GPLv3 doesn't make the requirement that you stated.

so if the BIOS checked the checksum of the boot image and if it found it 
wasn't correct would disable the video input hardware but let you boot the 
system otherwise it would be acceptable to you and the GPLv3?

somehow I doubt it, but that's what it would take to prevent modified 
software from interacting with their networks (remembering that these 
networks are the cable and satellite networks in some cases)

it also seems that if this was the case it would be a trivial work-around 
for the GPLv3 if it was acceptable.

David Lang

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

[Tomas Neme]

> > However, I don't see how this would ever require a company like Tivo
> > or Mastercard to have their networks play nice with a unit that has
> > been modified by the end user, potentially opening up some serious
> > security holes.
>
> Which is why the GPLv3 doesn't make the requirement that you stated.

Why, if you let user-compiled kernels to run in a TiVo, it might be
modified so the TiVo can be used to pirate-copy protected content,
which is a serious security hole. TiVo would need to read, approve of,
and sign any modified kernels the users intend to use on their
hardware. If GPLv3 allows for this, it'd be doing exactly that

Tomás

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[david]

On Wed, 20 Jun 2007, Alexandre Oliva wrote:

> On Jun 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:
>>> It is the duty of the FSF to defend these freedoms.  It's its public
>>> mission.  That's a publicly stated goal of the GPL, for anyone who
>>> cares to understand it, or miss it completely and then complain about
>>> changes in spirit.
>
>> I wouldn't call it a duty.  It is the chosen mission perhaps, but nobody
>> is making them do it.
>
> Everyone who donates to it does so understanding what the mission is.
> Detracting from that mission would be failing the public commitment.

true, but selecting the GPL as the license for your project is not 
donating to the FSF.

David Lang

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

[Alexandre Oliva]

On Jun 20, 2007, "Jesper Juhl" <jesper.juhl@gmail.com> wrote:

> On 19/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>> > In the GPLv3 world, we have already discussed in this thread how you can
>> > follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM
>> > instead of using a flash device.
>> 
>> Yes.  This is one option that doesn't bring any benefits to anyone.
>> It maintains the status quo for users and the community, but it loses
>> the ability for the vendor to upgrade, fix or otherwise control the
>> users.  Bad for the vendor.

> Also bad for the user

We already know the vendor doesn't care about the user, so why should
we take this into account when analyzing the reasoning of the vendor?

The rest of your paragraph was covered in what I wrote, 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 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> On Tue, Jun 19, 2007 at 05:04:52AM -0300, Alexandre Oliva wrote:
>> Once again, now with clearer starting conditions (not intended to
>> match TiVo in any way, BTW; don't get into that distraction)
>> 
>> 
>> Vendor doesn't care about tivoizing, their business works the same
>> either way.

> Not true.  A PVR that can record pay per view and encrypted digital
> channels

You see the "not intended to match TiVo" above?

Do you see that it's pointless to dispute antecedents of a logical
inference rule, if you don't know what role it plays in the full
argument?

Consider that this could be a proof by contradiction to realize how
pointless your objection is, no matter how true the point you state
is.  It bears no relationship with the argument at hand, and you said
so yourself, by disputing the assumptions of the inference, rather
than its conclusions.

Assumptions that were not even used to arrive at the conclusions, BTW.

>> Can you point out any flaw in this reasoning, or can we admit it as
>> true?

> Certainly fails to be true.

Once you change the conditions to twist whatever else you want, then
you arrive at different conclusions.  What's the surprise here?

What you're doing is like, after getting an argument like this:

  assumptions:
    1+1 = 2
    2+1 = 3

  provable consequence:

    1+1+1 = 3

responding:

  no, no, that's wrong!  the right argument is:

  assumptions:

    3+4 = 7
    2+1 = 3

  provable consequence:

    2+1+4 = 7

  therefore the answer is 5, not 3!

You see how illogical this is?

It doesn't matter whether your argument is correct.  It just doesn't
dispute the proposition at hand.

-- 
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 20, 2007, Andrew McKay <amckay@iders.ca> wrote:

> However, I don't see how this would ever require a company like Tivo
> or Mastercard to have their networks play nice with a unit that has
> been modified by the end user, potentially opening up some serious
> security holes.

Which is why the GPLv3 doesn't make the requirement that you stated.

-- 
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 20, 2007, "H. Peter Anvin" <hpa@zytor.com> wrote:

> Alexandre Oliva wrote:
>> 
>>> b) the manufacturer is able to update the device _in_ _the_ _field_.
>> Sure, it would be more costly, but it's not like the
>> law (or the agreements in place) *mandate* tivoization.

> The sad part is that the FCC, especially, are pretty fond of doing
> exactly that.

<broken-record>
  It does not mandate the use of *copyleft* Free Software in non-ROM
  such a way that the user cannot modify it.
</broken-record>

> This comes more from a general cluelessness about technology

And the meaning of tivoization ;-)

Tivoization doesn't mean "user can't modify".  It's more than that.


But I agree with the feeling.  It's like mandating knife manufacturers
to design ways to stop people from hurting or killing others with
knives.  So much for self defense...

-- 
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 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> On Mon, Jun 18, 2007 at 06:12:57PM -0300, Alexandre Oliva wrote:
>> Aah, good question.  Here's what the draft says about this:
>> 
>> Mere interaction with a user through a computer network, with no
>> transfer of a copy, is not conveying.
>> 
>> The requirements as to "installation information" apply to conveying
>> the program along with a user product.

> So if I go use a computer running some GPL software, and I copy the
> contents of /bin to a CD and bring it home, does the owner of the
> machine now owe me a copy of the GPL sources?

According to one of the rationales of GPLv3, it is understood that
lending someone a computer for a short period of time does not amount
to conveying the software in it.  I assume this is backed by strong
legal reasoning I won't pretend to know or understand.  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

[Alexandre Oliva]

On Jun 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> On Sun, Jun 17, 2007 at 12:52:38AM -0300, Alexandre Oliva wrote:
>> Why should restrictions through patents be unacceptable, but
>> restrictions through hardware and software be acceptable.

>> Both are means to disrespect users' freedoms.

> A patent prevents you from using the software in any way at all,
> while a hardware restriction prevents you from using the software on
> that particular hardware, but not on lots of other hardware.  Very
> big difference.

So, one disrespects a lot, the other disrespects a little.  Is that
relevant, when the requirement is "no further restrictions"?

>> It is the duty of the FSF to defend these freedoms.  It's its public
>> mission.  That's a publicly stated goal of the GPL, for anyone who
>> cares to understand it, or miss it completely and then complain about
>> changes in spirit.

> I wouldn't call it a duty.  It is the chosen mission perhaps, but nobody
> is making them do it.

Everyone who donates to it does so understanding what the mission is.
Detracting from that mission would be failing the public commitment.

Sure, it may have been self-imposed in the beginning, but maybe not
even then.  At least in Brazil, foundations are started by an initial
donor, who determines its mission, and it has a legal obligation to
pursue that mission, and IIRC it cannot be changed except by a court
order, and even than within certain limits.

> So what would happen if some company was to make software for a tivo and
> released their binaries signed with some specific key, and they released
> information on how to check this was signed with their key, and then
> some other companies went and made tivo hardware and decided that they
> would only allow code signed by the first companies key to run on it,

I was pretty sure this had been covered in the section about technical
barriers to modification in the third draft's rationale, but I can't
find it right now.  http://gplv3.fsf.org/gpl3-dd3-rationale.pdf

Anyhow, the argument I read went like: if there's an agreement between
the parties to do this, then the copyright holder can probably enforce
the license regardless of the software and hardware distributor being
different parties, since the software is being distributed with
information whose purpose is to enable the hardware to deny the user
the freedom to run modified versions of the software.

However, if there's no such agreement, if the copyright holder has no
copyright claims over the hardware or works shipped in it, there's
nothing the copyright holder can do about it, and that's probably how
it should be, since a copyright license (!= contract) can't possibly
prohibit people from creating hardware limited in function, it can
only tell people that, in order for them to have permission to modify
or distribute the covered work, they must abide by certain conditions.
And if they don't want to abide by the conditions, and they don't
manage to obtain a license from the copyright holders that doesn't
impose conditions they can't accept, they just can't modify or
distribute 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

[Tim Post]

On Wed, 2007-06-20 at 16:25 -0400, Michael Poole wrote:
> Tomas Neme writes:
> 
> >>     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.
> >>     -- US Code, Title 17, Section 101
> >
> > so?
> 
> People keep arguing that the signature is somehow not part of the
> kernel or not subject to copyright law.  I suspect they do not realize
> how broad the legal definition of "computer program" is.


"A computer program is a set of statements or instructions used by a
computer."

I have no idea what logic was employed that made continuing beyond that
point desirable. 

However, the GPL was designed so you couldn't be forced to pay to use
the software released under it, my guess is if it weren't for such a
broad definition the GPL would look a bit different than it does. 

When laws give you ground that gets you free stuff, typically its unwise
to suggest that they take some back ;)

Best,
--Tim

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

[Jesper Juhl]

On 18/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 17, 2007, "Jesper Juhl" <jesper.juhl@gmail.com> wrote:
>
> > On 17/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> > [snip]
>
> >> Serious, what's so hard to understand about:
>
> >> no tivoization => more users able to tinker their formerly-tivoized
> >> computers => more users make useful modifications => more
> >> contributions in kind
>
> > I have to disagree.
>
> Your analysis stopped at the downside of prohibiting tivoization.  You
> didn't analyze the potential upsides,

Maybe that's because I don't really see any up sides.

As I see it, if we prevent tivoization, then the most likely outcome
will be that a very few number of vendors will switch to ROM based
solutions or similar (everyone lose, both vendor and user), a few
vendors that currently tivoize hardware may open up their hardware but
I doubt that will be very many, and the vast majority of vendors will
move to *BSD or proprietary software since they simply can't or won't
open up their hardware.

So no, I don't think there are any upsides. We'll lose a huge number
of developers, testers and users inside the business comunity and
we'll lose a lot of exposure (like "hey, did you know TiVO actually
runs Linux inside? Isn't that cool?)...  Gaining a few hobyists at the
expense of driving a huge number of businesses away from GPL'ed
software does not look like an upside to me.


>so you may indeed come to
> different conclusions, and they may very well be wrong.
>
Just because I come to a different conclusion than you doesn't
nessesarily make it wrong.

> It's very human to look only at the potential downside of an action
> and conclude it's a bad action.
>
And you believe yourself to be immune to that - right?


> > Let's say that for some reason I don't want the end users of my
> > device to tinker with the software inside my device.
>
> Ok, keep the *want* in mind.  This is very important.
>
No, it is not. When I wrote that I meant "don't want" as in "really
don't want to since it'll destroy our business" or "really really
don't want to since we would be breaking the law" etc.

> > Now I think you can agree to these things being positive:
>
> Yes, even if I'd phrase them slightly differently.
>
> > The only downside is that the end user purchasing the device can't
> > install modified versions of the software on it.
>
> And therefore you severely limit the number of end users who might
> turn into contributors because of self interest in hacking the device
> to suit their needs.
>
Most people don't care about hacking their devices, and of the few who
do only a subset have the skill and only a subset of those will
actually contribute anything back. This is a *small* set of people and
gaining that small set at the expense of losing the large number of
contributers from various companies doesn't make sense to me.


> > Now let's try it in a GPLv3 universe.  Since I can no longer create my
> > device without having to allow the end user to install modified
> > software on it
>
> False assumption.  You can create the device using GPLv3 software in
> it.

Not as long as I want to prevent the user from tampering with it, no.

>So your acccounting of necessary downsides is only one of the
> possibilities.  The other possibility would be to have the program in
> ROM, of course, which would come with a completely different set of
> downsides, but that would retain all of the "these things being
> positive" you mentioned above.
>
But do you really expect a vendor to put a device on the market where
they also lock themselves out of upgrading it and releasing new
software for it?  That's just rediculous.


> And, remember, since you merely don't *want* the end user of the
> device to tinker with the software, you have the option to do let them
> do that.
>
See above.

> And, if you do, they may find in themselves reasons and incentives to
> change the software in the device, and the improvements are likely to
> get back to the community and thus back to you.  Everybody wins.
>
For a few select individuals that may be true. But for the majority of
the population it won't mean a thing.

> This is the upside that you left out from your analysis, and from
> every other analysis that set out to "prove" that anti-tivoization is
> bad that I've seen so far.
>
I'm sorry, but I don't think it holds water.

> It appears that people are so concerned about whatever little they
> might lose from requiring respect for users' freedoms that they don't
> even consider what they might win, and that they *would* win if at
> least some of the vendors were to make an choice more favorable to
> their users and the community.

Contrary to you, I don't believe any significant number of companies
will do that. It's simply better for business to just use other
software in that case.

-- 
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

[Michael Poole]

Tomas Neme writes:

>>     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.
>>     -- US Code, Title 17, Section 101
>
> so?

People keep arguing that the signature is somehow not part of the
kernel or not subject to copyright law.  I suspect they do not realize
how broad the legal definition of "computer program" is.

> Not GPL related, but casino machine software that needs to be approved
> by the casino regulation office in Argentina need to provide source,
> compiling instructions AND binaries, and the binaries must pass a diff
> check. This is impossible without a hacked compiler since the
> timestamps WILL differ.
>
> Just an example that legality doesn't always comply with itself, and
> even less make sense.

This discussion is about copyright and the GPL, not legal quirks.  I
am aware of a large number of silly results reached by law.

> plus, and I repeat myself.. the program comes with no warranties whatsoever.
>
> and if your complains are purely moral, see it this way: if TiVo
> didn't sign their kernel, digital cable providers wouldn't give them
> their hash keys, and they wouldn't be able to show HD signals,
> rendering them useless, and making them go bankrupt... so they'd go
> BSD, because they ARE a company after all, and they are after The
> Moneys..
>
> no?

The GPL does not guarantee anyone a viable business model.  Following
it is not conditional on profitability.  It is only conditional on
exercising rights that are granted by the GPL.

Michael Poole

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

[Alexandre Oliva]

On Jun 20, 2007, lsorense@csclub.uwaterloo.ca (Lennart Sorensen) wrote:

> On Fri, Jun 15, 2007 at 04:26:34PM -0300, Alexandre Oliva wrote:
>> If the bug is in the non-GPLed BIOS, not in the GPLed code, too bad.
>> One more reason to dislike non-Free Software.

> Maybe the Tivo only loading signed kernels is a bug in their bios. :)

That might be so.  And in the US, a court might end up finding that
piece of code was taken at random, for no particular reason, from a
sample of garbage produced by code monkeys (in a very literal sense)
typing at random on computer keyboards over a large period of time ;-)

-- 
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 20, 2007, Manu Abraham <abraham.manu@gmail.com> wrote:

> Alan Cox wrote:
>>> Well, it is not Tivo alone -- look at http://aminocom.com/ for an
>>> example. If you want the kernel sources pay USD 50k and we will provide
>>> the kernel sources, was their attitude.
>> 
>> GPLv2 deals with that case, and they can (and should) be sued for it
>> [except that US copyright law is designed for large music companies not
>> people]

> Their argument was that the mentioned sources contain propreitary
> closed stuff from IBM/AMCC for the PPC 405/440 and or for the NXP
> (MIPS based) chips.

As you probably know, this is not a valid excuse to distribute the
software under conditions that disrespect its license.

It doesn't mean you can force them to give you the source code, it
only means the copyright holder can stop them from distributing the
software this way.

-- 
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

[Michael Poole]

david@lang.hm writes:

> On Wed, 20 Jun 2007, Michael Poole wrote:
>
>> Please retract that claim.  I have said no such thing, and have
>> avoided saying anything that I thought might be misconstrued in that
>> direction.
>>
>> To be absolutely clear: My complaints with Tivo as a hardware or BIOS
>> vendor are moral and pragmatic, not legal.  My complaint with Tivo as
>> a distributor of Linux is what hinges on legal issues.
>
> but if the GPL doesn't control the BIOS how in the world are you
> saying that the fact that the GPL covers the kernel makes what the
> BIOS does wrong (even if the kernel was covered by GPLv3)?

I do not say that the BIOS is doing anything (legally) wrong.  The
wrong act is distributing the binary kernel image without distributing
complete source code for it.

>>> that's a seperate body of code that is in no way derived from the
>>> linux kernel (even the anti-tampering functions would work equally
>>> well with other Operating systems and are in no way linux
>>> specific). it's no even loaded on the same media (the BIOS is in
>>> flash/rom on the botherboard, the OS is on the hard drive)
>>>
>>> and note that the software that is checked to make sure that it hasn't
>>> been changed includes much more then the kernel. it checks the kernel
>>> and the initrd.
>>
>> Not legally relevant.
>
> I disagree. it's very relevant if your argument is that becouse the
> checksum if a checksum of the kernel that the license for the kernel
> somehow controlls what can be done with it.

To the extent that it is relevant, it strengthens the argument against
Tivo: they are tying together many works of authorship, including some
GPLed works, in a way that makes them effectively inseparable.  This
is beyond "mere aggregation" on a distribution medium, and tends to
implicate *all* parts of the whole as GPL encumbered.

Michael Poole

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

[Tomas Neme]

> Just an example that legality doesn't always comply with itself, and
> even less make sense.

plus, and I repeat myself.. the program comes with no warranties whatsoever.

and if your complains are purely moral, see it this way: if TiVo
didn't sign their kernel, digital cable providers wouldn't give them
their hash keys, and they wouldn't be able to show HD signals,
rendering them useless, and making them go bankrupt... so they'd go
BSD, because they ARE a company after all, and they are after The
Moneys..

no?

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Dave Neuer]

On 6/20/07, david@lang.hm <david@lang.hm> wrote:
>
> but the signature isn't part of the kernel

But some would argue that it's part of the source with which the
binary is derived (only a court could meaningfully decide if they're
right).

> and the code that checks the
> signature is completely independant.

Irrelevant.

> and finally the PVR functions are not
> part of the kernel

Booting is.

Dave

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

[Tomas Neme]

>     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.
>     -- US Code, Title 17, Section 101

so?

Not GPL related, but casino machine software that needs to be approved
by the casino regulation office in Argentina need to provide source,
compiling instructions AND binaries, and the binaries must pass a diff
check. This is impossible without a hacked compiler since the
timestamps WILL differ.

Just an example that legality doesn't always comply with itself, and
even less make sense.

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> david@lang.hm writes:
>
>> On Wed, 20 Jun 2007, Michael Poole wrote:
>>
>>> david@lang.hm writes:
>>>
>>>> no, saying that the result must be acceptable to other software (in
>>>> this case the software running in the BIOS) is not part of the source
>>>> code.
>>>
>>> Why not?  The digital signature is a statement (which translates
>>> roughly to "Tivo approves this") to be used in a computer in order to
>>> bring about a certain result.  That result is making it boot on the
>>> PVR.  Source code simply means the original forms or inputs used to
>>> generate machine-readable statements.
>>
>> but the signature isn't part of the kernel, and the code that checks
>> the signature is completely independant. and finally the PVR functions
>> are not part of the kernel (and not under the GPL in any case)
>>
>> if your argument was true then Oracle releasing a database appliance
>> would require Oracle to give you the source to their database since
>> it's part of 'bringing about a certain result' namely operating as a
>> database server.
>
> From the kernel's COPYING file:
>
>   NOTE! This copyright does *not* cover user programs that use kernel
> services by normal system calls - this is merely considered normal use
> of the kernel, and does *not* fall under the heading of "derived work".
>
> See also the portion below.

the PVR software does exactly this, so the functionality as a PVR has 
nothing to do with the kernel by your own arguments.

>> if your argument was true then releasing a GPL package for windows
>> would require that the windows kernel source be released, after it
>> it's nessasary for 'brining about a certain result' namely letting
>> your code run.
>
> From section 3 of the GPL:
>
>  However, as a special exception, the source code distributed need
>  not include anything that is normally distributed (in either source
>  or binary form) with the major components (compiler, kernel, and so
>  on) of the operating system on which the executable runs, unless
>  that component itself accompanies the executable.

but then how can the kernel impose any restriction on the BIOS? (and 
remember, you are the one arguing that it somehow does)

David Lang

>> these are both nonsense results.
>
> .. which is why they are recognized to be different.
>
> Michael Poole
>

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> david@lang.hm writes:
>
>> On Wed, 20 Jun 2007, Michael Poole wrote:
>>
>>> david@lang.hm writes:
>>>
>>>> this is very much NOT true. if you take the source the provide you can
>>>> compile a kernel that will run on the tivo, the only thing you have to
>>>> do (on some models) is to change the bios to skip the step that checks
>>>> if the kernel has been tampered with.
>>>
>>> If we are opining whether Tivo provided complete source code for their
>>> Linux kernel images, the requirement to change non-GPLed software as a
>>> condition to exercise GPL-protected rights speaks for itself.
>>
>> no, the GPL protected rights don't say anything about the hardware the
>> system runs on.
>>
>> you are saying that the GPL now controls what the BIOS software is
>> allowed to do or not allowed to do.
>
> Please retract that claim.  I have said no such thing, and have
> avoided saying anything that I thought might be misconstrued in that
> direction.
>
> To be absolutely clear: My complaints with Tivo as a hardware or BIOS
> vendor are moral and pragmatic, not legal.  My complaint with Tivo as
> a distributor of Linux is what hinges on legal issues.

but if the GPL doesn't control the BIOS how in the world are you saying 
that the fact that the GPL covers the kernel makes what the BIOS does 
wrong (even if the kernel was covered by GPLv3)?

>> that's a seperate body of code that is in no way derived from the
>> linux kernel (even the anti-tampering functions would work equally
>> well with other Operating systems and are in no way linux
>> specific). it's no even loaded on the same media (the BIOS is in
>> flash/rom on the botherboard, the OS is on the hard drive)
>>
>> and note that the software that is checked to make sure that it hasn't
>> been changed includes much more then the kernel. it checks the kernel
>> and the initrd.
>
> Not legally relevant.

I disagree. it's very relevant if your argument is that becouse the 
checksum if a checksum of the kernel that the license for the kernel 
somehow controlls what can be done with it.

>>> Out of curiosity, what do you have to do on models besides those?  Are
>>> newer models more or less restrictive in what they run?  If newer
>>> models are more restrictive, I think that also speaks to whether Tivo
>>> thinks it is conveying complete source code.
>>
>> newer models do tend to be more restrictive, but they also tend to
>> connect to more propriatary networks (satellite or cable)
>
> What they connect to is also not relevant.  That imples that because a
> vendor has been issued or licensed patents, they are not obliged to
> follow the GPL -- that the vendor has other obligations that supercede
> the GPL's license claims.  GPL section 7 addresses that situation.

you are arguing that the fact that later models are more locked down is 
'proof' that tivo knows that it's doing the wrong thing. I'm pointing out 
an alternate reason, the people who control the networks that the newer 
models are connecting to are imposing additional restrictions on tivo that 
cause them to need to be locked down more.

David Lang

> Michael Poole
>

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

[David Schwartz]

> The kernel you build from the source code that Tivo distributes must
> be accepted by Tivo's hardware without making other modifications (to
> Tivo's hardware or bootloader).  If that is possible, I will retract
> what I said.  If it is not possible, they are omitting part of the
> program's source code:
>
>     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.
>     -- US Code, Title 17, Section 101

A key is a number. A signature is a number. They are neither statements nor
instructions. The argument that GPLv2 prohibits Tivoization is really and
truly absurd. It has neither a legal nor a moral leg to stand on.

DS

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

[Michael Poole]

david@lang.hm writes:

> On Wed, 20 Jun 2007, Michael Poole wrote:
>
>> david@lang.hm writes:
>>
>>> no, saying that the result must be acceptable to other software (in
>>> this case the software running in the BIOS) is not part of the source
>>> code.
>>
>> Why not?  The digital signature is a statement (which translates
>> roughly to "Tivo approves this") to be used in a computer in order to
>> bring about a certain result.  That result is making it boot on the
>> PVR.  Source code simply means the original forms or inputs used to
>> generate machine-readable statements.
>
> but the signature isn't part of the kernel, and the code that checks
> the signature is completely independant. and finally the PVR functions
> are not part of the kernel (and not under the GPL in any case)
>
> if your argument was true then Oracle releasing a database appliance
> would require Oracle to give you the source to their database since
> it's part of 'bringing about a certain result' namely operating as a
> database server.

>From the kernel's COPYING file:

   NOTE! This copyright does *not* cover user programs that use kernel
 services by normal system calls - this is merely considered normal use
 of the kernel, and does *not* fall under the heading of "derived work".

See also the portion below.

> if your argument was true then releasing a GPL package for windows
> would require that the windows kernel source be released, after it
> it's nessasary for 'brining about a certain result' namely letting
> your code run.

>From section 3 of the GPL:

  However, as a special exception, the source code distributed need
  not include anything that is normally distributed (in either source
  or binary form) with the major components (compiler, kernel, and so
  on) of the operating system on which the executable runs, unless
  that component itself accompanies the executable.

> these are both nonsense results.

.. which is why they are recognized to be different.

Michael Poole

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

[Michael Poole]

david@lang.hm writes:

> On Wed, 20 Jun 2007, Michael Poole wrote:
>
>> david@lang.hm writes:
>>
>>> this is very much NOT true. if you take the source the provide you can
>>> compile a kernel that will run on the tivo, the only thing you have to
>>> do (on some models) is to change the bios to skip the step that checks
>>> if the kernel has been tampered with.
>>
>> If we are opining whether Tivo provided complete source code for their
>> Linux kernel images, the requirement to change non-GPLed software as a
>> condition to exercise GPL-protected rights speaks for itself.
>
> no, the GPL protected rights don't say anything about the hardware the
> system runs on.
>
> you are saying that the GPL now controls what the BIOS software is
> allowed to do or not allowed to do.

Please retract that claim.  I have said no such thing, and have
avoided saying anything that I thought might be misconstrued in that
direction.

To be absolutely clear: My complaints with Tivo as a hardware or BIOS
vendor are moral and pragmatic, not legal.  My complaint with Tivo as
a distributor of Linux is what hinges on legal issues.

> that's a seperate body of code that is in no way derived from the
> linux kernel (even the anti-tampering functions would work equally
> well with other Operating systems and are in no way linux
> specific). it's no even loaded on the same media (the BIOS is in
> flash/rom on the botherboard, the OS is on the hard drive)
>
> and note that the software that is checked to make sure that it hasn't
> been changed includes much more then the kernel. it checks the kernel
> and the initrd.

Not legally relevant.

>> Out of curiosity, what do you have to do on models besides those?  Are
>> newer models more or less restrictive in what they run?  If newer
>> models are more restrictive, I think that also speaks to whether Tivo
>> thinks it is conveying complete source code.
>
> newer models do tend to be more restrictive, but they also tend to
> connect to more propriatary networks (satellite or cable)

What they connect to is also not relevant.  That imples that because a
vendor has been issued or licensed patents, they are not obliged to
follow the GPL -- that the vendor has other obligations that supercede
the GPL's license claims.  GPL section 7 addresses that situation.

Michael Poole

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

[David Schwartz]

> Tomas Neme writes:

> > I have been following this discussion for the last week or so, and
> > what I haven't been able to figure out is what the hell is the big
> > deal with TiVO doing whatever they want to with their stupid design.
> > They made a design, they build a machine, they sell it as is, and
> > provide source code for GPL'ed software... what's your problem?

> It's simple: they don't provide _complete_ source code.  They keep the
> source code for the part of their Linux kernel images that provides
> the functionality "runs on Tivo DVRs".  The GPL requires that
> distributors of binary versions provide complete source code, not just
> the parts of source code that are convenient.
>
> Michael Poole

That leads to lots of obvious nonsense unless you fix it with all kinds of
made up ad-hoc changes just to get the result you want. Why doesn't Linus
have to release the keys he uses to sign the Linux kernel source
distributions? That provides the functionality "can be proven to be
authorized by Linus". What you call "runs on Tivo DVRs", I call "can be
proven to be authorized by Tivo to run on Tivo DVRs".

The problem is that your description of the functionality as "runs on Tivo
DVRs" is an ad-hoc choice. You could describe that functionality any number
of other ways, and this is the only one that supports your argument. (Which
would be wrong anyway since functionality has nothing to do with whether
something is part of the source code or not. Copyright is not functional in
operation.)

Tivo's choice is an authorization decision. It is similar to you not having
root access to a Linux box. Sorry, you can't run a modified kernel on that
machine, but you can still modify the kernel and run it on any hardware
where authorization decisions don't stop you from doing so. The GPL was
never about such authorization decisions.

Suppose I make a machine that automatically accepts any kernel signed by
Linus and configures it, compiles it, and installs it. Does this change
Linus' signature into "works with my machine's kernel autoinstall"
functionality? The signature is proof Linus 'blessed' the release. Others
who trust Linus can use this functionally to make authorization decisions.
However, your releases are not blessed by Linus and there's no reason you
should be able to give them this 'blessed by Linus' functionality.

There are other legal reasons why this can't be right (most of them have
been stated at least three times in this thread), but I think the
commonsense argument is the most persuasive. Far from being part of the
source code, the signature is proof of source and authorization. These are
fact that don't apply to your modified release.

If I only want to run kernels signed by Linus, you have no right to trick me
into running a modified kernel. The Tivo only wants to run kernels signed by
Tivo.

Your argument is equivalent to saying that I have the right not just to run
modified Linux kernels on my own hardware but to compel others to run my
modifications even when authorization decisions say they won't run my
changes. (By bogusly calling authorization decisions 'functionality'.)

DS

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

[Lennart Sorensen]

On Tue, Jun 19, 2007 at 07:28:22PM +0400, Manu Abraham wrote:
> Well, it is not Tivo alone -- look at http://aminocom.com/ for an
> example. If you want the kernel sources pay USD 50k and we will provide
> the kernel sources, was their attitude.

Hmm, set top boxes are often rented from the cable company rather than
sold.  Stupid grey area for sure.  At least tivo does give you the
sources, without demanding more money.  Rather big difference.

> Well, it is not Tivo alone, a large chunk of the vendors do that. The
> vendors who actually do it the clean way are just few and can be counted
> very easily.

Well at least where I work we don't try to lock down the hardware, we do
contribute our changes and bug fixes to upstream when it makes sense
(and where our changes wouldn't make sense for upstream, they are still
clearly included with the sources we have.)  If a customer wants a copy
of the sources, they will get a nice DVD, although strangely none have
asked for one yet.

--
Len Sorensen

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> david@lang.hm writes:
>
>> no, saying that the result must be acceptable to other software (in
>> this case the software running in the BIOS) is not part of the source
>> code.
>
> Why not?  The digital signature is a statement (which translates
> roughly to "Tivo approves this") to be used in a computer in order to
> bring about a certain result.  That result is making it boot on the
> PVR.  Source code simply means the original forms or inputs used to
> generate machine-readable statements.

but the signature isn't part of the kernel, and the code that checks the 
signature is completely independant. and finally the PVR functions are not 
part of the kernel (and not under the GPL in any case)

if your argument was true then Oracle releasing a database appliance would 
require Oracle to give you the source to their database since it's part of 
'bringing about a certain result' namely operating as a database server.

if your argument was true then releasing a GPL package for windows would 
require that the windows kernel source be released, after it it's 
nessasary for 'brining about a certain result' namely letting your code 
run.

these are both nonsense results.

David Lang

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> david@lang.hm writes:
>
>> this is very much NOT true. if you take the source the provide you can
>> compile a kernel that will run on the tivo, the only thing you have to
>> do (on some models) is to change the bios to skip the step that checks
>> if the kernel has been tampered with.
>
> If we are opining whether Tivo provided complete source code for their
> Linux kernel images, the requirement to change non-GPLed software as a
> condition to exercise GPL-protected rights speaks for itself.

no, the GPL protected rights don't say anything about the hardware the 
system runs on.

you are saying that the GPL now controls what the BIOS software is allowed 
to do or not allowed to do.

that's a seperate body of code that is in no way derived from the linux 
kernel (even the anti-tampering functions would work equally well with 
other Operating systems and are in no way linux specific). it's no even 
loaded on the same media (the BIOS is in flash/rom on the botherboard, the 
OS is on the hard drive)

and note that the software that is checked to make sure that it hasn't 
been changed includes much more then the kernel. it checks the kernel and 
the initrd.

> Out of curiosity, what do you have to do on models besides those?  Are
> newer models more or less restrictive in what they run?  If newer
> models are more restrictive, I think that also speaks to whether Tivo
> thinks it is conveying complete source code.

newer models do tend to be more restrictive, but they also tend to connect 
to more propriatary networks (satellite or cable)

David Lang

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

[Michael Poole]

david@lang.hm writes:

> no, saying that the result must be acceptable to other software (in
> this case the software running in the BIOS) is not part of the source
> code.

Why not?  The digital signature is a statement (which translates
roughly to "Tivo approves this") to be used in a computer in order to
bring about a certain result.  That result is making it boot on the
PVR.  Source code simply means the original forms or inputs used to
generate machine-readable statements.

Michael Poole

>>    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.
>>    -- US Code, Title 17, Section 101

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

[david]

On Wed, 20 Jun 2007, Michael Poole wrote:

> Tomas Neme writes:
>
>>> It's simple: they don't provide _complete_ source code.  They keep the
>>> source code for the part of their Linux kernel images that provides
>>> the functionality "runs on Tivo DVRs".  The GPL requires that
>>
>> http://en.wikipedia.org/wiki/Tivoization does not agree that this is
>> the problem but rather "TiVo circumvented this goal by making their
>> products run programs only if the program's digital signature matches
>> those authorised by the manufacturer of the TiVo."
>>
>> I'm downloading the sources now.. if they compile, then you're lying
>> to me, right? Moreover, if I compile them as is, and I can run them on
>> a TiVo (let's say upgrading the machine's kernel) then you're even
>> more so..
>
> Whether the sources you download from Tivo compile says very little.
> It certainly does not mean I am wrong.  I could give you binary for
> version A and sources for version B -- and the sources would compile.
>
> The kernel you build from the source code that Tivo distributes must
> be accepted by Tivo's hardware without making other modifications (to
> Tivo's hardware or bootloader).  If that is possible, I will retract
> what I said.  If it is not possible, they are omitting part of the
> program's source code:

no, saying that the result must be acceptable to other software (in this 
case the software running in the BIOS) is not part of the source code.

David Lang

>    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.
>    -- US Code, Title 17, Section 101
>
> Michael Poole
> -
> 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/
>

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

[Michael Poole]

david@lang.hm writes:

> this is very much NOT true. if you take the source the provide you can
> compile a kernel that will run on the tivo, the only thing you have to
> do (on some models) is to change the bios to skip the step that checks
> if the kernel has been tampered with.

If we are opining whether Tivo provided complete source code for their
Linux kernel images, the requirement to change non-GPLed software as a
condition to exercise GPL-protected rights speaks for itself.

Out of curiosity, what do you have to do on models besides those?  Are
newer models more or less restrictive in what they run?  If newer
models are more restrictive, I think that also speaks to whether Tivo
thinks it is conveying complete source code.

Michael Poole

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

[Michael Poole]

Tomas Neme writes:

>> It's simple: they don't provide _complete_ source code.  They keep the
>> source code for the part of their Linux kernel images that provides
>> the functionality "runs on Tivo DVRs".  The GPL requires that
>
> http://en.wikipedia.org/wiki/Tivoization does not agree that this is
> the problem but rather "TiVo circumvented this goal by making their
> products run programs only if the program's digital signature matches
> those authorised by the manufacturer of the TiVo."
>
> I'm downloading the sources now.. if they compile, then you're lying
> to me, right? Moreover, if I compile them as is, and I can run them on
> a TiVo (let's say upgrading the machine's kernel) then you're even
> more so..

Whether the sources you download from Tivo compile says very little.
It certainly does not mean I am wrong.  I could give you binary for
version A and sources for version B -- and the sources would compile.

The kernel you build from the source code that Tivo distributes must
be accepted by Tivo's hardware without making other modifications (to
Tivo's hardware or bootloader).  If that is possible, I will retract
what I said.  If it is not possible, they are omitting part of the
program's source code:

    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.
    -- US Code, Title 17, Section 101

Michael Poole

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

[Dave Neuer]

On 6/20/07, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> On Wed, 20 Jun 2007, Dave Neuer wrote:
> > >
> > > And anybody who thinks others don't have the "right to choice", and then
> > > tries to talk about "freedoms" is a damn hypocritical moron.
> >
> > One might say the same thing about someone who claims not to have a
> > moral right to force certain choices on others in some circumstances
> > (e.g. when those others have used copyrighted work in a product and
> > ought to understand that for some not insignificant portion of the
> > copyright holders, the terms implicitly included preserving certain
> > "freedoms" for downstream recipients) while reserving a very similar
> > moral right with others (e.g. potential murderers, theives,
> > tresspassers, distributors of proprietary derived works).
>
> I don't disagree that "morals" are something very personal, and you can
> thus never really argue on morals *except*for*your*own*behaviour*.
>
> So I claim that for *me* the right choice is GPLv2 (or something similar).
> I think the GPLv3 is overreaching.
>
> There's a very fundamental, and very basic rule that is often a good
> guideline. It's "Do unto others".
>
> So the reason I *personally* like the GPLv2 is that it does unto others
> exactly what I wish they would do unto me.
>
> It allows everybody do make that choice that I consider to be really
> important: the choice of how something _you_ designed gets used.
>
> And it does that exactly by *limiting* the license to only that one work.
> Not trying to extend it past the work.
>
> See?
>
> The GPLv3 can never do that. Quite fundamentally, whenever you extend the
> "reach" of a license past just the derived work, you will *always* get
> into a situation where people who designed two different things get into a
> conflict when they meet. The GPLv2 simply avoids the conflict entirely,
> and has no problem at all with the "Do unto others as you would have them
> do unto you".
>
> In a very real sense, the GPLv3 asks people to do things that I personally
> would refuse to do. I put Linux on my kids computers, and I limit their
> ability to upgrade it. Do I have that legal right (I sure do, I'm their
> legal guardian), but the point is that this is not about "legality", this
> is about "morality". The GPLv3 doesn't match what I think is morally where
> I want to be. I think it *is* ok to control peoples hardware. I do it
> myself.
>
> So your arguments about "potential murderes", "thieves", "trespassers" and
> "distributors of proprietary derived works" is totally missing the point.
>
> It's missing the point that "morals" are about _personal_ choices. You
> cannot force others to a certain moral standpoint.
>
> Laws (like copyright law) and legal issues, on the other hand, are
> fundamentally *not* about "personal" things, they are about interactions
> that are *not* personal. So laws need to be fundamnetally different from
> morals. A law has to take into account that different people have
> different moral background, and a law has to be _pragmatic_.
>
> So trying to mix up a moral argument with a legal one is a fundamental
> mistake. They have two totally different and separate areas.
>
> The GPLv2 is a *legal* license. It's not a "moral license" or a "spiritual
> guide". Its raison-d'etre is that pragmatic area where different peoples
> different moral rules meet.
>
> In contrast, a persons *choice* to use the GPLv2 is his private choice.
> Totally different. My choice of the GPLv2 doesn't say anything about my
> choice of laws or legal issues.
>
> You don't have to agree with it - but exactly because it's his private
> choice, it's a place where the persons moral rules matter, in a way that
> they do *not* matter in legal issues.
>
> So killing, thieving, and distributing proprietary derived works are about
> *legal* choices. Are they also "immoral"? Who knows. Sometimes killing is
> moral. Sometimes thievery can me moral. Sometimes distributing derived
> works can be moral. Morality != legality. They are two totally different
> things.
>
> Only religious fanatics and totalitarian states equate "morality" with
> "legality". There's tons of examples of that from human history. The ruler
> is not just a king, he's a God, so disagreeing with him is immoral, but
> it's also illegal, and you can get your head cut off.
>
> In fact, a lot of our most well-known heroes are the ones that actually
> saw the difference between morals and laws.
>
> A German soldier who refused to follow orders was clearly the more "moral"
> one, wouldn't you say? Never mind law. Gandhi is famous for his peaceful
> civil disobedience - was that "immoral" or "illegal"?
>
> Or Robin Hood. A romantic tale, but one where the big fundamnetal part of
> the picture is the _difference_ between morality and legality.
>
> Think about it.

Oh, I have, professor.

>
> Yes, there is obviously overlap, in that a lot of laws are there to
> protect things that people also consider "moral". But the fact that there
> is correlation should *not* cause anybody to think that they are at all
> about the same thing.

No, they overlap in cases where the reason for the law is to enforce
some moral edict, and when they're not in sync it's often because some
person or group of people casuistically draw a moral distinction which
is absent in the law (likely because it conflicts with some other
group's moral values).

Your example of Robin Hood is instructive. Rich people quite likely
don't agree with you that taking from the rich to give to the poor is
moral, and they get more say (in a culturally-dependent fashion) when
it comes time to codify the morals into laws.

>
> > To call people who draw the line in a different place than you
> > hypocrites is BS.
>
> That was *not* what I did.

It is.

>
> I don't think it's hypocritical to prefer the GPLv3. That's a fine choice,
> it's just not *mine*.

I understand your oft-repeated (in this message alone) preference,
which revolves around your personal "bright line" boundary of moral
coercion at "I designed it."

>
> What I called hypocritical was to do so in the name of "freedom", while
> you're at the same time trying to argue that I don't have the "freedom" to
> make my own choice.
>
> See? THAT is hypocritical.

It is no more hypocritical than any other instance of restricting one
person's freedom to protect some freedom of some other person. The law
restricts your freedom to take things from my house in order to
preserve my freedom to own things. That you don't disagree with that
law indicates that you agree with the majority of people about where
to draw the line between one person's freedom and another in that
instance, not that you are a hypocrite or that people who steal are
hypocrites (they're hypocrites if they complain about people stealing
from them).

You have every right to prefer using your copyright to coerce people
to provide source but not keys used to make binaries from that source
run on hardware with which those binaries are distributed.

But people who prefer the GPLv3 in the name of freedom simply favor
making explicit what they felt was always implicit in earlier
versions, which is their wish that people who want to use what "they
designed" do so in ways which comport with freedoms which are more
important to them than the freedom of the hardware developer to design
their system unconstrained (like the freedom of the end user to make
their TiVO work some other way). They are no more hypocrites than you
are for apparently feeling that it's OK for Robin Hood to steal from
the rich to feed the poor.

>
>                         Linus
>

Dave

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

[Andrew McKay]

Alan Cox wrote:
>> Secondly GPLv3 will cause companies like TIVO, router companies, security 
>> companies to not adopt Linux as an operating system, because they can't secure 
>> their system.  Placing code in a ROM so they can't upgrade their own systems is 
> 
> You've made an important mistake. You said "their system". Now its "our
> code" and "whoever bought the units' hardware" so it isn't their anything.


Yes, the hardware belongs to the user, and the software belongs to the Linux 
community.   However I think I wasn't 100% clear, I also mean keeping companies 
networks and content secured.  Credit card companies insuring the software 
hasn't been modified to skim cards (not that it's the only way to skim a card), 
  or Tivo making sure that their content providers are protected.  Lets look at 
the credit card example.  Sure the user could modify the system and boot their 
own kernel, but it doesn't have to play nice with Mastercard's network anymore. 
  Or better yet, would actually report that a certain business's card reader had 
been tampered with.

> 
> You've made a second mistake I think by assuming that vendor held keys
> "improve" security and must be vendor held and secret for it to work. In
> fact vendor owned key systems that cannot be changed usually reduce
> security.
> 
> There are very very good reasons for having vendor owned secret keys.
> There are also very very good reasons for being able to rekey or disable
> the key on your box.
> 
> Ask people whose product vendor went bankrupt. With the ability to
> override/replace the keys they could have maintained their system
> securely instead they could make no updates and the boxes were left
> insecure.

I do see what you're saying here, and I can see how this is a problem.  However 
what is the solution?  Sure having the system open for users to replace software 
and tinker is great.  It's how I got into engineering.  I can also appreciate 
the ability for the end user to fix and continue to use a system long after a 
vendor goes out of business.  However, I don't see how this would ever require a 
company like Tivo or Mastercard to have their networks play nice with a unit 
that has been modified by the end user, potentially opening up some serious 
security holes.  From what I understand this would still violate GPLv3 because 
the system could no longer preform the task it was designed to do with modified 
code, but maybe I have misunderstood.

Andrew McKay

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

[david]

On Wed, 20 Jun 2007, H. Peter Anvin wrote:

> Michael Poole wrote:
>> Tomas Neme writes:
>>
>>> I have been following this discussion for the last week or so, and
>>> what I haven't been able to figure out is what the hell is the big
>>> deal with TiVO doing whatever they want to with their stupid design.
>>> They made a design, they build a machine, they sell it as is, and
>>> provide source code for GPL'ed software... what's your problem?
>>
>> It's simple: they don't provide _complete_ source code.  They keep the
>> source code for the part of their Linux kernel images that provides
>> the functionality "runs on Tivo DVRs".  The GPL requires that
>> distributors of binary versions provide complete source code, not just
>> the parts of source code that are convenient.
>>
>
> If true, that's a clear violation of the GPLv2, but it's very different
> from preventing the loading of modified images on their hardware (which
> appears to be permitted by v2.)
>
> 	-hpa

this is very much NOT true. if you take the source the provide you can 
compile a kernel that will run on the tivo, the only thing you have to do 
(on some models) is to change the bios to skip the step that checks if the 
kernel has been tampered with.

in fact, people have ported the nessary changes all the way up to the 2.6 
kernels (the tivos run a 2.1 derived kernel). you can buy hard drives to 
plug into your tivo with newer kernels (and yes, you do get source from 
those folks as well)

by the way, the first hacks that disabled these checks were done by people 
who were interested in getting their tivo to boot faster (they found an 
expensive bios routing that wasn't nessasary to boot), they weren't 
interested in loading different software on the box. when they released 
the hack other people realized that they had opened the door for other 
changes.

David Lang

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

[Tomas Neme]

> It's simple: they don't provide _complete_ source code.  They keep the
> source code for the part of their Linux kernel images that provides
> the functionality "runs on Tivo DVRs".  The GPL requires that

http://en.wikipedia.org/wiki/Tivoization does not agree that this is
the problem but rather "TiVo circumvented this goal by making their
products run programs only if the program's digital signature matches
those authorised by the manufacturer of the TiVo."

I'm downloading the sources now.. if they compile, then you're lying
to me, right? Moreover, if I compile them as is, and I can run them on
a TiVo (let's say upgrading the machine's kernel) then you're even
more so..

if this is all true, then I read GPLv2 and it tells me (clipping and ** mine):

0.  This License applies to any *program* or other work [...]. The
"Program", below, refers to any such program or work, and a "work
based on the Program" means either the Program or any derivative work
under copyright law: that is to say, *a work containing the Program or
a portion of it, either verbatim or with modifications* [...]

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 copy and
distribute such modifications [...]

3.  You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:

    a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections 1
and 2 above on a medium customarily used for software interchange;
[...]

The source code for a work means the preferred form of the work for
making modifications to it. 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. However, as a
special exception, the source code distributed
*need*not*include*anything that is *normally*distributed (in *either*
source or *binary* form) with the major components (compiler, kernel,
and so on) of the operating system on which the executable runs,
unless that component itself accompanies the executable.

11.  BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO
WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY
KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME
THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

so, it very clearly says that the program and the modifications come
with NO WARRANTY of working, either expressed or implied, including
the warranty of fitness for a particular purpose (for example being
able to run in your TiVo box). Other than that the licence says you
can do whatever you want as long as you give out the source code.

Let me put it this way: if TiVo gave microsoft the keys to run in
TiVo, and Microsoft made a Windows Kernel for TiVo and it run, they
would not be needing to give you any source codes because the box
itself isn't GPL'ed and even if it was, there's no need to GLP stuff
that work on a system based on The Program (say Adobe finally ports
Photoshop into linux) but is not a modification or aggregation.
Whatever checks the kernel's key is outside of the kernel itself, and
probably not a modification of any GPL'ed software

plus, what I marked in point 3. says that they don't need to give you
anything that is normally distributed with the system the program runs
on. They don't need to give you the source or the design of their
hardware, or any embebbed software

so.. you're free to modify the kernel, try to figure out what TiVo
does for authentication and modify your binaries so they look what
TiVo wants them to.. add some garbage bits in the end, or something.

One more thing: even if the program they use to make their binaries
TiVo-key-check compliant was GPL'ed, they don't need to give it to
you, because the kernel does not use them. *IT* uses the compiled
kernel as input data (and produce signed kernel binaries as output),
not the other way around

Tomás

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[H. Peter Anvin]

Michael Poole wrote:
> Tomas Neme writes:
> 
>> I have been following this discussion for the last week or so, and
>> what I haven't been able to figure out is what the hell is the big
>> deal with TiVO doing whatever they want to with their stupid design.
>> They made a design, they build a machine, they sell it as is, and
>> provide source code for GPL'ed software... what's your problem?
> 
> It's simple: they don't provide _complete_ source code.  They keep the
> source code for the part of their Linux kernel images that provides
> the functionality "runs on Tivo DVRs".  The GPL requires that
> distributors of binary versions provide complete source code, not just
> the parts of source code that are convenient.
> 

If true, that's a clear violation of the GPLv2, but it's very different
from preventing the loading of modified images on their hardware (which
appears to be permitted by v2.)

	-hpa

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

[Michael Poole]

Tomas Neme writes:

> I have been following this discussion for the last week or so, and
> what I haven't been able to figure out is what the hell is the big
> deal with TiVO doing whatever they want to with their stupid design.
> They made a design, they build a machine, they sell it as is, and
> provide source code for GPL'ed software... what's your problem?

It's simple: they don't provide _complete_ source code.  They keep the
source code for the part of their Linux kernel images that provides
the functionality "runs on Tivo DVRs".  The GPL requires that
distributors of binary versions provide complete source code, not just
the parts of source code that are convenient.

Michael Poole

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

[Tomas Neme]

I have been following this discussion for the last week or so, and
what I haven't been able to figure out is what the hell is the big
deal with TiVO doing whatever they want to with their stupid design.
They made a design, they build a machine, they sell it as is, and
provide source code for GPL'ed software... what's your problem?

In order to play backuped games in, say, a PS2 I need to modify it's
hardware. Gut it up, solder some cables to a chip that will bypass the
signature key reading in the CD/DVDs. It's legal to mod it (it's my
hardware, I paid for it) although it voids my warranty. It's also
legal to copy the games (where I live) if I own the original, but
we're no talking about DMR here. Not exactly

Now, I'm sure some alike modification can be made to "help" TiVO
bypassing the signature key check... and I'm sure it's legal. Again,
what is your problem?

Linus said it already, but TiVO has seen himself FORCED to protect its
software for the very same reason Microsoft had to uncanningly
overhead EVERYTHING in Vista: the Media (Fox, Universal, Etcetera)
companies forced them too. But this is just a comment.. the thing is
that I really can't see the problem

T

-- 
|_|0|_|
|_|_|0|
|0|0|0|

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

[Linus Torvalds]

On Wed, 20 Jun 2007, Dave Neuer wrote: 
> > 
> > And anybody who thinks others don't have the "right to choice", and then
> > tries to talk about "freedoms" is a damn hypocritical moron.
> 
> One might say the same thing about someone who claims not to have a
> moral right to force certain choices on others in some circumstances
> (e.g. when those others have used copyrighted work in a product and
> ought to understand that for some not insignificant portion of the
> copyright holders, the terms implicitly included preserving certain
> "freedoms" for downstream recipients) while reserving a very similar
> moral right with others (e.g. potential murderers, theives,
> tresspassers, distributors of proprietary derived works).

I don't disagree that "morals" are something very personal, and you can 
thus never really argue on morals *except*for*your*own*behaviour*.

So I claim that for *me* the right choice is GPLv2 (or something similar). 
I think the GPLv3 is overreaching.

There's a very fundamental, and very basic rule that is often a good 
guideline. It's "Do unto others".

So the reason I *personally* like the GPLv2 is that it does unto others 
exactly what I wish they would do unto me.

It allows everybody do make that choice that I consider to be really 
important: the choice of how something _you_ designed gets used.

And it does that exactly by *limiting* the license to only that one work. 
Not trying to extend it past the work.

See?

The GPLv3 can never do that. Quite fundamentally, whenever you extend the 
"reach" of a license past just the derived work, you will *always* get 
into a situation where people who designed two different things get into a 
conflict when they meet. The GPLv2 simply avoids the conflict entirely, 
and has no problem at all with the "Do unto others as you would have them 
do unto you".

In a very real sense, the GPLv3 asks people to do things that I personally 
would refuse to do. I put Linux on my kids computers, and I limit their 
ability to upgrade it. Do I have that legal right (I sure do, I'm their 
legal guardian), but the point is that this is not about "legality", this 
is about "morality". The GPLv3 doesn't match what I think is morally where 
I want to be. I think it *is* ok to control peoples hardware. I do it 
myself.

So your arguments about "potential murderes", "thieves", "trespassers" and 
"distributors of proprietary derived works" is totally missing the point.

It's missing the point that "morals" are about _personal_ choices. You 
cannot force others to a certain moral standpoint. 

Laws (like copyright law) and legal issues, on the other hand, are 
fundamentally *not* about "personal" things, they are about interactions 
that are *not* personal. So laws need to be fundamnetally different from 
morals. A law has to take into account that different people have 
different moral background, and a law has to be _pragmatic_.

So trying to mix up a moral argument with a legal one is a fundamental 
mistake. They have two totally different and separate areas.

The GPLv2 is a *legal* license. It's not a "moral license" or a "spiritual 
guide". Its raison-d'etre is that pragmatic area where different peoples 
different moral rules meet.

In contrast, a persons *choice* to use the GPLv2 is his private choice.  
Totally different. My choice of the GPLv2 doesn't say anything about my 
choice of laws or legal issues. 

You don't have to agree with it - but exactly because it's his private 
choice, it's a place where the persons moral rules matter, in a way that 
they do *not* matter in legal issues.

So killing, thieving, and distributing proprietary derived works are about 
*legal* choices. Are they also "immoral"? Who knows. Sometimes killing is 
moral. Sometimes thievery can me moral. Sometimes distributing derived 
works can be moral. Morality != legality. They are two totally different 
things.

Only religious fanatics and totalitarian states equate "morality" with 
"legality". There's tons of examples of that from human history. The ruler 
is not just a king, he's a God, so disagreeing with him is immoral, but 
it's also illegal, and you can get your head cut off.

In fact, a lot of our most well-known heroes are the ones that actually 
saw the difference between morals and laws.

A German soldier who refused to follow orders was clearly the more "moral" 
one, wouldn't you say? Never mind law. Gandhi is famous for his peaceful 
civil disobedience - was that "immoral" or "illegal"? 

Or Robin Hood. A romantic tale, but one where the big fundamnetal part of 
the picture is the _difference_ between morality and legality.

Think about it.

Yes, there is obviously overlap, in that a lot of laws are there to 
protect things that people also consider "moral". But the fact that there 
is correlation should *not* cause anybody to think that they are at all 
about the same thing. 

> To call people who draw the line in a different place than you 
> hypocrites is BS.

That was *not* what I did.

I don't think it's hypocritical to prefer the GPLv3. That's a fine choice, 
it's just not *mine*.

What I called hypocritical was to do so in the name of "freedom", while 
you're at the same time trying to argue that I don't have the "freedom" to 
make my own choice.

See? THAT is hypocritical.

			Linus

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

[Jesper Juhl]

On 19/06/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > In the GPLv3 world, we have already discussed in this thread how you can
> > follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM
> > instead of using a flash device.
>
> Yes.  This is one option that doesn't bring any benefits to anyone.
> It maintains the status quo for users and the community, but it loses
> the ability for the vendor to upgrade, fix or otherwise control the
> users.  Bad for the vendor.
>
Also bad for the user since now the vendor can no longer supply
updated firmware to fix bugs or otherwise improve the device. Also,
the vendow has the problem that devices that get returned for repair
or similar can't easily be updated software wise. For the vendor not
to be able to update the box creates a lot of problems for both the
vendor and the end user.


> As another option, the vendor can respect users' freedoms, and then
> everybody wins big.  That's the option that anti-tivoization provides
> economic incentive for vendors to take.  Sure, they may still prefer
> the alternative above, or stick with an older version (which has its
> costs), or move to different software (which also has its costs), but
> it's unreasonable to claim that I'm advocating for vendors to move to
> ROM.
>
I am fairly confident that if too much software switches to GPLv3
we'll see a lot of businesses move to BSD or proprietary software.
That means *we* lose bigtime.

-- 
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

[Dave Neuer]

On 6/20/07, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
> > To call people who draw the line in
> > a different place than you hypocrites is BS.
>
> Very poor example. In many parts of the world "Just quit" is "just starve
> to death".

> So please DON'T equate the two. Tivo is a minor control argument about a
> silly little TV recording box. Employee rights is a matter of life, death
> and slavery for many many people.

Well, I consider your eloquent counter-argument to have strengthened
my point that such "rights to moral coercion" arguments are inherently
subjective and highly context-dependent.

Dave

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

[Lennart Sorensen]

On Tue, Jun 19, 2007 at 05:04:52AM -0300, Alexandre Oliva wrote:
> Yes.  How does this relate with the piece of the argument I've
> proposed so far, or the whole argument I've posted before?
> 
> Answer: It doesn't.  At all.  You're just showing you didn't
> understand the argument.  Which shows why I have to explain it piece
> by piece.  Which suggests you shouldn't try to jump to conclusions.
> 
> 
> 
> Once again, now with clearer starting conditions (not intended to
> match TiVo in any way, BTW; don't get into that distraction)
> 
> 
> Vendor doesn't care about tivoizing, their business works the same
> either way.

Not true.  A PVR that can record pay per view and encrypted digital
channels and other such things (as mandated by MPAA or whoever is
unfortunately in charge of such stupid thigns), is much easier to sell
to customers, and as a result they have to prevent things that the
content providers want prevented (never mind that such prevention is
futile because there is no DRM that can't be broken, only made more
difficult to break).  So their business does not work anywhere near as
well if they can't sell devices that do what their customers want them
to do (which is record stuff from TV, not play tetris, even though
tetris might be fun too).  Now if they don't do what the content
providers want, they have to sell an inferior PVR, while a competitor
that does what the content providers demand (in order to provide what
customers demand), then they won't sell very much.  They could go with
another non GPL os and software of course, and spend way more moeny on
it (or potentially use BSD or something) and deal with the problem that
way.  Either way GPL software gets no contributions under the GPL v3 if
the company wants to be competitive in that business.

> Vendor's employees will contribute the same, one way or another, so
> their contributions are out of the equation.

No, because they are no longer using GPL software.

> Users get source code in either case, and they can modify it and share
> it.  They're in no way stopped from becoming part of the community.

The users don't get any code now, because the vendor has no requirement
to release it, or may even have no permission to realse it.  And end
users are for the most part incapable of contributing anyhow.

> Given these conditions:
> 
> In a tivoized device, users will be unable to scratch their itches.
> This doesn't stop them from contributing to the project, but they may
> lack self-interest motivation to contribute, because they won't be
> able to use their modifications in the device they own.

I bought some hardware and built a box to run mythtv.  I can only record
stuff from regular analog cable, because I can't buy any hardware that
supports digital encrypted cable.  If I bought a locked down PVR
instead, then I would be able to record that, but I wouldn't be able to
play around.  To me playing around was more important than recording
encrypted channels, but to most customers, being able to record
everything is what is important.

> In a non-tivoized device, users can scratch their itches.  They can
> contribute just as much as they would in a tivoized device, but since
> they can use the changes they make to make their own devices work
> better for them, this works as a motivator for them to make changes,
> and perhaps to contribute them.  Therefore, they will tend to
> contribute more.

Their device can't physically do some of the thigns they wanted it to do
though, since those features disappeared when the tivoization was
removed due to licensing issues with the content makers.

> Can you point out any flaw in this reasoning, or can we admit it as
> true?

Certainly fails to be true.

Who is to blame?  The content makers for requiring lock down of devices
that can access their content?  Consumers for demanding to be able to
use that content and hence buying devices that can use it?  The company
that made a device that did what consumers demanded by implementing it
the way required by the content makers?

--
Len Sorensen

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

[Alan Cox]

> constraints it's valid for a copyright holder to try to enforce w/ a
> license -- I think it's immoral for an employer to force an employee
> to toil at a meaningless, soul-crushing job for the vast majority of
> one's single, short existence if they could make it more enjoyable,
> but I'd hate to see someone try to enforce that with a license (I'm
> happy telling a person in that situation to "just quit" just as you
> tell people "just don't buy it"). To call people who draw the line in
> a different place than you hypocrites is BS.

Very poor example. In many parts of the world "Just quit" is "just starve
to death".

And guess what - brand owners do use the rights to try and stop that kind
of abuse with varying degrees of success. The 'fair trade' logo is
controlled this way. Many companies do their best not to license
production of goods to sweat shops and child slaves. It's considered
morally correct and companies get pounded by the public and lose business
when they fail to police these restrictions.

So please DON'T equate the two. Tivo is a minor control argument about a
silly little TV recording box. Employee rights is a matter of life, death
and slavery for many many people.

Alan

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

[Tim Post]

On Wed, 2007-06-20 at 11:27 -0500, Andrew McKay wrote:

> I'm not going to address whether GPLv3 changed the spirit of GPLv2, but saying 
> that licensing the Linux Kernel under GPLv3 will result in more contributions is 
> absolute BS.

Is there a system in place that sort of keeps track? Argument aside, I'm
really interested to see how adoption in devices encourages
contributions (and what kinds of contributions it attracts). Anyone
would be I guess.

How do you guys keep up with it? My head almost blows up just thinking
about it given the traffic of just _this_ list.

TIA :)
--Tim

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

[Dave Neuer]

On 6/20/07, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>
> But they do have the right to make their own choices, and try their own
> strategies. And people shouldn't complain about that. If somebody doesn't
> like the Tivo box, and the Tivo service requirements, just don't *buy* the
> damn thing, and don't sign up for the service.

>
> And anybody who thinks others don't have the "right to choice", and then
> tries to talk about "freedoms" is a damn hypocritical moron.

One might say the same thing about someone who claims not to have a
moral right to force certain choices on others in some circumstances
(e.g. when those others have used copyrighted work in a product and
ought to understand that for some not insignificant portion of the
copyright holders, the terms implicitly included preserving certain
"freedoms" for downstream recipients) while reserving a very similar
moral right with others (e.g. potential murderers, theives,
tresspassers, distributors of proprietary derived works).

As I pointed out in a previous message in the thread, that some people
want the copyright holders -- who have economic power over the
hardware vendors who use Linux in their products -- to force a
desireable outcome for those without that power (i.e. the small number
of people who buy TiVOs or Linksys routers who actually care about
source availability and who get told "don't buy it" as if that will
change anything) is understandable.

I think that there's a legitimate concern about what types of
constraints it's valid for a copyright holder to try to enforce w/ a
license -- I think it's immoral for an employer to force an employee
to toil at a meaningless, soul-crushing job for the vast majority of
one's single, short existence if they could make it more enjoyable,
but I'd hate to see someone try to enforce that with a license (I'm
happy telling a person in that situation to "just quit" just as you
tell people "just don't buy it"). To call people who draw the line in
a different place than you hypocrites is BS.

Your pragmatic arguments make much more sense than your moralistic ones, IMHO.

Dave

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

[Alan Cox]

> Secondly GPLv3 will cause companies like TIVO, router companies, security 
> companies to not adopt Linux as an operating system, because they can't secure 
> their system.  Placing code in a ROM so they can't upgrade their own systems is 

You've made an important mistake. You said "their system". Now its "our
code" and "whoever bought the units' hardware" so it isn't their anything.

You've made a second mistake I think by assuming that vendor held keys
"improve" security and must be vendor held and secret for it to work. In
fact vendor owned key systems that cannot be changed usually reduce
security.

There are very very good reasons for having vendor owned secret keys.
There are also very very good reasons for being able to rekey or disable
the key on your box.

Ask people whose product vendor went bankrupt. With the ability to
override/replace the keys they could have maintained their system
securely instead they could make no updates and the boxes were left
insecure.

RPM for example intentionally follows this approach. RPM will check keys
and the keys for different vendors/projects will not be released. However
you can add keys, remove keys or even tell rpm "forget the key checking".

A silly example that rather makes the point is the X-Box. Using a bug in
a game you can get into the X-box system and run Linux instead. As the
owner of an X-box you can't fix the bug in the game because you don't
know the signing key, but if you could change or add keys you could stop
the "problem" occurring.

The Xbox is perhaps an oddity in that the insecurity is widely preferred
by the owners but the situation applies in cases where the owner would
prefer the reverse were true.

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

[Linus Torvalds]

On Wed, 20 Jun 2007, Helge Hafting wrote:

> Linus Torvalds wrote:
> > 
> > You do realize that Tivo makes all their money on the service, don't you?
> > The actual hardware they basically give away at cost, exactly to get the
> > service contracts. Not exactly a very unusual strategy in the high-tech
> > world, is it?
> >   
> Not unusual - but so what.
> This strategy may backfire if users find a way
> to use the cheap box for something without Tivo's service contract.

Yeah, I'm not at all trying to say that Tivo has the "right to make 
money". 

They can fail for all I care. 

But they do have the right to make their own choices, and try their own 
strategies. And people shouldn't complain about that. If somebody doesn't 
like the Tivo box, and the Tivo service requirements, just don't *buy* the 
damn thing, and don't sign up for the service.

Whining about Tivo's choices is just stupid.

And anybody who thinks others don't have the "right to choice", and then 
tries to talk about "freedoms" is a damn hypocritical moron.

		Linus

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

[Andrew McKay]

> No, I'm not.  You can say tivoization is *good* however much you like.
> This doesn't dispute in any way my claim that no tivoization would be
> *better*, that you'd get contributions from the people that, because
> of tivoization, don't feel compelled to develop and contribute,
> because they can't use the fruits of their efforts in the device where
> they would be most useful for them.
> 

I know I'm jumping in to this conversation in the middle and a few days behind, 
but that's one of the biggest fallacies in the open source communities.  First 
of all more system developers contribute to the community than end users do. 
Secondly GPLv3 will cause companies like TIVO, router companies, security 
companies to not adopt Linux as an operating system, because they can't secure 
their system.  Placing code in a ROM so they can't upgrade their own systems is 
an absolute joke, and not a viable option.  In the end, GPLV3 will cause less 
contribution to the Linux Kernel community. Systems developers won't be choosing 
Linux and directly contributing to Kernel development while developing their 
product.  Fewer products on the market will use linux, so even end users, who 
really don't contribute a lot, won't be contributing either.

I'm not going to address whether GPLv3 changed the spirit of GPLv2, but saying 
that licensing the Linux Kernel under GPLv3 will result in more contributions is 
absolute BS.

Andrew McKay
Iders Inc.

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

[H. Peter Anvin]

Alexandre Oliva wrote:
> 
>> b) the manufacturer is able to update the device _in_ _the_ _field_.
> Sure, it would be more costly, but it's not like the
> law (or the agreements in place) *mandate* tivoization.
> 

The sad part is that the FCC, especially, are pretty fond of doing
exactly that.  This comes more from a general cluelessness about
technology (FCC is mostly stuffed with political shills which have more
to do with who stuffed money into the current President's campaign than
anything else) than malice, but "not modifiable by the end user" is a
common requirement in FCC regulations, which have the force of law.

	-hpa

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

[Lennart Sorensen]

On Mon, Jun 18, 2007 at 06:12:57PM -0300, Alexandre Oliva wrote:
> Aah, good question.  Here's what the draft says about this:
> 
>   Mere interaction with a user through a computer network, with no
>   transfer of a copy, is not conveying.
> 
> The requirements as to "installation information" apply to conveying
> the program along with a user product.

So if I go use a computer running some GPL software, and I copy the
contents of /bin to a CD and bring it home, does the owner of the
machine now owe me a copy of the GPL sources?

--
Len Sorensen

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

[Lennart Sorensen]

On Sun, Jun 17, 2007 at 12:52:38AM -0300, Alexandre Oliva wrote:
> What it does is impose conditions for whoever wants to distribute the
> software.  And GPLv3 makes it explicit that one such condition is to
> permit the user to install and run modified versions of the program in
> the hardware that ships with the program.  A condition that is
> arguably already encoded in the "no further restrictions to the rights
> granted" by the license" and to the requirement for complete
> corresponding source code to accompany the binary.
> 
> That you disagree with it doesn't make you right.
> 
> But that it is within the spirit of the GPL defined by its authors
> (which is all I'm trying to show here), it is.
> 
> > The GPL (at least through version 2) is about free access to source
> > code.
> 
> Some think so, but this was GPLv1.
> 
> v2 added stuff such as:
> 
>   if a patent license would not permit royalty-free redistribution of
>   the Program by all those who receive copies directly or indirectly
>   through you, then the only way you could satisfy both it and this
>   License would be to refrain entirely from distribution of the
>   Program
> 
> Do you realize that the patent is unrelated with the program, but
> nevertheless the copyright license establishes conditions about what
> kind of patent licenses you may accept in order for you to have
> permission to distribute the program.
> 
> Why should restrictions through patents be unacceptable, but
> restrictions through hardware and software be acceptable.
> 
> Both are means to disrespect users' freedoms.

A patent prevents you from using the software in any way at all, while a
hardware restriction prevents you from using the software on that
particular hardware, but not on lots of other hardware.  Very big
difference.

> It is the duty of the FSF to defend these freedoms.  It's its public
> mission.  That's a publicly stated goal of the GPL, for anyone who
> cares to understand it, or miss it completely and then complain about
> changes in spirit.

I wouldn't call it a duty.  It is the chosen mission perhaps, but nobody
is making them do it.

> That's true.  Per the license, it's only who distributes the hardware
> to you that shouldn't impose such restrictions.

So what would happen if some company was to make software for a tivo and
released their binaries signed with some specific key, and they released
information on how to check this was signed with their key, and then
some other companies went and made tivo hardware and decided that they
would only allow code signed by the first companies key to run on it,
because that company had software which was acceptable to the
DMCA/RIAA/MPAA/etc and allowed them to get access to the hardware they
wanted to use in their box.  The second company now sells hardware to
make money, and the first company sells tv guide updates service to
people who want to use their software releases fully.

What does the GPL do now?  The software company still releases the
sources to the GPL software, but their binary releases are signed with a
key they don't give you.  They didn't provide you with any hardware, you
have to buy that from the hardware company that makes a product that
happens to run that software because it has the right bits of hardware
to record tv programs and such.  The hardware company put restrictions
on what software the box will run, although techicly the software
company that has the signing key could make lots of compatible software
for that particular locked down hardware, including vxworks or windows
based code if they chose to do so, while the hardware company just makes
hardware and decided to only allow software with the signature to run.
They didn't distribute any software, the buyer has to go get that from
the software company if they want the box to be useful (probably not a
good business plan for the majority of customers, but still possible in
theory).

> That's right.  But one of the obligations is to impose no further
> restrictions on the exercise of the rights.  What is "imposing a
> restriction"?  Installing the software in ROM isn't regarded as such,
> it's just a technical decision.  Installing the software in modifiable
> non-volatile storage, but denying the user the ability to change it,
> is regarded as imposing a restriction.  (note the "denying")  It is a
> matter of intent.
> 
> It's not because you only install say 32MB of RAM on the machine that
> you're denying the user the ability to run OOo on the machine.  But if
> you ship the computer with plenty of memory, but somehow configure the
> hardware or the operating system so as to prevent the user from
> upgrading an OOo that shipped with it, while you can still install
> that upgrade, then you're actively placing limits on the user's
> freedom WRT to that software, and an anti-tivoization clause would
> then stop you from distributing the software under these conditions.
> 
> I've never disputed that this is how they perceive it.
> 
> I've never disputed that GPLv2 serves this goal.
> 
> I still think GPLv3 serves this same goal, and better than v2.
> 
> But this is not what my participation here is about.
> 
> My participation here is about showing that GPLv3, and anti-tivozation
> in particular, don't violate the spirit of the defending users'
> freedoms WRT the covered software, such that the Free Software remains
> Free.

Well many people in the community disagrees, and you can't change their
mind on that it would seem, just as they can't change yours (or the FSFs
for that matter).  I would not be surprised to see some code forks when
the GPLv3 is finally released.

--
Len Sorensen

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

[Lennart Sorensen]

On Fri, Jun 15, 2007 at 04:26:34PM -0300, Alexandre Oliva wrote:
> If the bug is in the non-GPLed BIOS, not in the GPLed code, too bad.
> One more reason to dislike non-Free Software.

Maybe the Tivo only loading signed kernels is a bug in their bios. :)

> The freedom the GPL defends is not the freedom to modify and debug the
> system, but rather the covered software.
> 
> Now, if you find evidence that the "bug" is actually intentionally put
> there to stop you from doing what you wanted with the software, then
> there's clearly a violation of the spirit of the license, and you
> might even have a case of copyright infringement, but IANAL.

There are many interesting bugs out there.  Who is to say what was
intensional?

--
Len Sorensen

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

[Dave Neuer]

On 6/20/07, SL Baur <steve@xemacs.org> wrote:
> On 6/19/07, Dave Neuer <mr.fred.smoothie@pobox.com> wrote:
>
> > Linux was a tool for UNIX sysadmins and admin wannabes to
> > practice their UNIX chops at home - or a conveniently inexpensive
> > platform on which to run Apache. Companies -- other than Linux
> > distributors -- didn't bet their business on it.
>
> Wrong conclusion.  Been there, done that, helped bet the company
> on networks based on Linux servers.

One of thousands, no doubt.

>
> > Apache's success greatly contributed to the corporate acceptance of Linux, IMHO.
>
> Wrong again. Apache was not allowed to distribute strong
> encryption for e-commerce servers over that time frame.

In the US only, and it was easily available for free if people didn't
care about violating patents or you could buy Stronghold from
Covalent. And plenty of folks didn't need encryption or did without it
even if they did.

>
> By the time of Linux 2.0.x,

1996, right? When Linux had just got SMP support, and Apache was
running on a majority of webservers, according to Netcraft?

> it could stay up for years at a
> time even if it was running on garbage.  There was no alternative
> even *close*.

You're preaching to the choir about Linux quality. I started using it
in 1997 (to run Apache, on a cast-off i486sx w/ 8 MB ram IIRC), and it
was years before I ever had it crash on me.

My point was about perceptions, not quality -- not that Linux wasn't
good, but that it was not widely considered to be "enterprise class."
For certain uses, Apache was, in a way that was fairly novel for Free
Software.

Anyway, it was also meant to be somewhat tongue-in-cheek (what's the
emoticon for a tongue in a cheek?) -- anyone who thinks they know the
_one_ reason Linux was successful is on crack (well, it might have
been Linus' accent).

Dave

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

[Helge Hafting]

Linus Torvalds wrote:
> On Fri, 15 Jun 2007, Alexandre Oliva wrote:
>   
>> case 2'': tivo provides source, end user tries to improve it, realizes
>> the hardware won't let him use the result of his efforts, and gives up
>>     
>
> So you're blaming Tivo for the fact that your end user was a lazy bum and 
> wanted to take advantage of somebody elses hard work without permission?
>
> Quite frankly, I know who the bad guy in that scenario is, and it ain't 
> Tivo. It's your lazy bum, that thought he would just take what Tivo did, 
> sign the contract, and then not follow it. And just because the box 
> _contained_ some piece of free software, that lazy bum suddenly has all 
> those rights? Never mind all the *other* effort that went into bringing 
> that box to market?
>
> You do realize that Tivo makes all their money on the service, don't you? 
> The actual hardware they basically give away at cost, exactly to get the 
> service contracts. Not exactly a very unusual strategy in the high-tech 
> world, is it?
>   
Not unusual - but so what.
This strategy may backfire if users find a way
to use the cheap box for something without Tivo's service contract.

That don't make me feel sorry for Tivo - they can then sell their
boxes with some profit - or go bankrupt for doing stupid business.

Similiar to how unrealistic cheap printers backed by ink sales
fail when third parties undercuts the ridiculously expensive ink.
Or give-away cellphones tied to an expensive provider, being
unlocked by third parties so a switch to a cheaper provider will work.


Don't get me wrong - I have nothing against Tivo - and nothing for
them either.  Them making a locked device is just a fun arms
race to see who can reprogram the bootloader key
or find some other clever way to load software. . .

> Guys, in fighting for "your rights", you should look a bit at *other* 
> peoples rights too. Including the rights of hw manufacturers, and the 
> service providers. Because this is all an eco-system, where in order to 
> actually succeed, you need to make _everybody_ succeed.
>   
Nothing against a hw manufacturers right - but of course they
have no particular right to succeed with give-away hw and
expensive service.  When that strategy fail due to people operating
the hw without buing service then the cure is to charge properly
for the hw, not to ask everybody to please stop hacking.

Assuming that nobody can change the box after the sale is
their risk, and there is nothing unfair about failure here.

Helge Hafting

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

[Alan Cox]

On Wed, 20 Jun 2007 00:11:24 -0400
"Dave Neuer" <mr.fred.smoothie@pobox.com> wrote:

> On 6/19/07, david@lang.hm <david@lang.hm> wrote:
> >
> > it was the ability of the linux kernel to adapt to vastly different
> > hardware (including embeded hardware) that made Linux what it is today.
> 
> Which is why NetBSD is currently poised to take over the world...


To be honest I wish some of the people continually stealing our work
would just go off and use BSD instead since the BSD people welcome that
kind of binary only one way usage.

The license is/was a problem however because of the advertising clause
parts. It was one reason BSD was not used on a large companies choice of
routers - the 'credit required' bit at the time gave the lawyers
heartburn.

Alan

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

[Al Boldi]

david@lang.hm wrote:
> On Wed, 20 Jun 2007, Al Boldi wrote:
> > Scott Preece wrote:
> >> On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
> >>> Nicolas Mailhot wrote:
> >>>> Tivo didn't make the Linux success. More Tivos can definitely undo
> >>>> it.
> >>>
> >>> I don't think so.
> >>>
> >>> First, it's not Linux that made success, but rather GNU that uses
> >>> Linux as its kernel.  And, believe it or not, when people say Linux,
> >>> they really mean GNU.  People could care less what kernel they were
> >>> running, as long as the system is up and runs the procs that offer
> >>> their services.
> >>
> >> ---
> >>
> >> Actually, for use in devices (like TiVos or cell phones), it is very
> >> definitely the kernel that is of interest. Many such devices use
> >> little or no GNU software (some manufacturers have consciously avoided
> >> it because of the possibility of shifts like the GPLv3 changes).
> >
> > Sure, but was it Linux in embedded devices that made Linux what it is
> > today, or was it GNU/Linux?
>
> if it was the GNU that made linux what it is today and the linux kernel
> mearly an oppurtunist then the GNU/Hurd, GNU/Solaris, GNU/BSD and for that
> matter GNU/Microsoft distributions should be steadily and quickly gaining
> marketshare

They may never reach critical mass, as GNU/Linux already crossed that border.  
Their only hope may possibly be a drop-in kernel replacement to leverage the 
vast availability of GNU/Linux distributions.

So what's this got to do with the GPLv2/v3 debate?  It's the seemingly fair 
GPLv2 that was possibly instrumental in growing this large GNU/Linux 
community, and changing this fairness by hampering parties to enter this 
community may well be counterproductive.


Thanks!

--
Al

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

[Manu Abraham]

Alan Cox wrote:
>> Well, it is not Tivo alone -- look at http://aminocom.com/ for an
>> example. If you want the kernel sources pay USD 50k and we will provide
>> the kernel sources, was their attitude.
> 
> GPLv2 deals with that case, and they can (and should) be sued for it
> [except that US copyright law is designed for large music companies not
> people]


Their argument was that the mentioned sources contain propreitary closed
stuff from IBM/AMCC for the PPC 405/440 and or for the NXP (MIPS based)
chips. Even if the GPLv2 deals with it, well haven't reached anywhere
with it, inspite of talks with them. So most of the users just probably
stopped talking sense with them, just like me.

Have some of those Amino STB's, the software on it being buggy,
including myself many others wanted to fix those bugs, but then people
had to pay for their annual support to get the fixes. People who were
able to fix also were denied the same since there is no source
available. But if you wanted the sources, then you pay for the sources.
If you don't pay for their sources, then pay for their
Bronze/Silver/Gold Support schemes, where people pay through their nose.

For a specific case with which i wanted to attach a USB based device to
the box, they stated: we can port in a driver that which exists in the
vanilla kernel, to their device but just that they need to be paid for
that to be done, eventhough if someone else was willing to do that job,
but that wasn't possible because of no sources.

In either way, if you buy their devices, it is just that, you keep
paying us, if you want your device your work as expected.

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

[SL Baur]

On 6/19/07, Dave Neuer <mr.fred.smoothie@pobox.com> wrote:

> It was Apache. Apache showed corporate users and small businesses
> desperate to cash in on the Interweb c. 1995-1998 ...

Right time period ...

> Linux was a tool for UNIX sysadmins and admin wannabes to
> practice their UNIX chops at home - or a conveniently inexpensive
> platform on which to run Apache. Companies -- other than Linux
> distributors -- didn't bet their business on it.

Wrong conclusion.  Been there, done that, helped bet the company
on networks based on Linux servers.

> Apache's success greatly contributed to the corporate acceptance of Linux, IMHO.

Wrong again. Apache was not allowed to distribute strong
encryption for e-commerce servers over that time frame.  The
solution we bought was O/S agnostic.

And to quote your next message, you have given all the reasons
why NetBSD has already taken over the world.

By the time of Linux 2.0.x, it could stay up for years at a
time even if it was running on garbage.  There was no alternative
even *close*.

-sb

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

[SL Baur]

> Sure, but was it Linux in embedded devices that made Linux what it is today,
> or was it GNU/Linux?

No, it was the fact that Linux has always been able to run on garbage.

My introduction to Linux was in 1995 when I was given a network
of computers made out of back-laboratory garbage and US$0
software budget and told to make it work.  None of the BSDs
could cut it, but Linux could.

User space Unix tool rewrites all of which I could have gotten from
*BSD had absolutely nothing to do with it.  I doubt that I am
typical.

-sb

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

[Alexandre Oliva]

On Jun 20, 2007, "David Schwartz" <davids@webmaster.com> wrote:

> The freedom to "run the program, for any purpose" is just as much
> violated by Microsoft when they make the Xbox.

That's correct.

> You can't run the Linux kernel on that either, for the exact same
> reasons you can't run a modified Linux kernel on the Tivo.

Technically speaking, yes.

Legally speaking, these are completely different situations.

First of all, Microsoft is not in any way involved in how you received
the copy of the kernel Linux you'd like to run on the Xbox.  If it is,
you might have a claim that they couldn't impose this restriction on
you.  IANAL.

Second, you're not a user of Linux on the Xbox in the first place, so
you don't have an inherent right to run the program there that the
hardware manufacturer is denying you.  It's not much different than
wanting to run the program on on computers you have remote access to,
on which you have no permission to install Linux, or on computers you
have but with incompatible hardware architectures.

The case of incompatible hardware architectures is actually a bit
different, since nobody is really stopping you from running the
program there, it would just take you porting work to get to run it.

While in both the remote-access case and the Xbox case someone *is*
indeed stopping you from installing and running the software on those
computers, this does not render the software non-free, because you're
not a user of the software there in the first place, and stating that,
in order for software to be free, you'd have to have the ability to
install it on any existing computer in the world would be nonsensical.

However, in the case of TiVo, you have received the software from TiVo
with the express purpose of running it on the hardware they sold you,
and you can't install or run modified versions precisely because they
don't want to let you.  They're denying you a freedom that you are
entitled to have, which renders that binary copy of Linux installed on
it non-Free, and at the same time they're imposing further
restrictions on your freedoms that the GPL is designed to respect and
defend, which violates the spirit, if not the letter, of the GPL.

So you see, these are two very different cases, the key difference
being the obligation that TiVo accepted towards the copyright holders
of Linux, when it distributed the software to you, of respecting your
freedoms, by not imposing restrictions on your exercise of the
freedoms in addition the restrictions that the license itself imposes.

> (And GPL rights were always applicable equally to all hardware in
> the universe, not specially to some hardware and not others.)

Correct.  Whoever distributed you the software entitled you to enjoy
the freedoms wherever you manage to run the 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

[David Schwartz]

> Much as I hate to extend the life of this execrable thread, since I
> think Alexandre makes Sisyphus look like a hard-nosed pragmatist, it
> seems pretty clear that TiVO impinges "[my] freedom to run the
> program, for any purpose" if "any purpose" includes "make my TiVO do
> what I want," and likewise to "adapt it to [my] needs" -- freedoms 0
> and part of 1. It is just disingenuous to argue otherwise.
>
> Dave

The freedom to "run the program, for any purpose" is just as much violated
by Microsoft when they make the Xbox. You can't run the Linux kernel on that
either, for the exact same reasons you can't run a modified Linux kernel on
the Tivo. It is manifestly obvious that the "freedom to run the program for
any purpose" must be limited to those pieces of hardware where you are the
one who decides what hardware runs on that software. (And GPL rights were
always applicable equally to all hardware in the universe, not specially to
some hardware and not others.)

DS

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

[Dave Neuer]

On 6/19/07, david@lang.hm <david@lang.hm> wrote:
>
> it was the ability of the linux kernel to adapt to vastly different
> hardware (including embeded hardware) that made Linux what it is today.

Which is why NetBSD is currently poised to take over the world...

Dave

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

[Dave Neuer]

On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
> Scott Preece wrote:
> > On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
> > > Nicolas Mailhot wrote:
> > > > Tivo didn't make the Linux success. More Tivos can definitely undo it.
> > >
> > > I don't think so.
> > >
> > > First, it's not Linux that made success, but rather GNU that uses Linux
> > > as its kernel.  And, believe it or not, when people say Linux, they
> > > really mean GNU.  People could care less what kernel they were running,
> > > as long as the system is up and runs the procs that offer their
> > > services.
> >
> > ---
> >
> > Actually, for use in devices (like TiVos or cell phones), it is very
> > definitely the kernel that is of interest. Many such devices use
> > little or no GNU software (some manufacturers have consciously avoided
> > it because of the possibility of shifts like the GPLv3 changes).
>
> Sure, but was it Linux in embedded devices that made Linux what it is today,
> or was it GNU/Linux?

It was Apache. Apache showed corporate users and small businesses
desperate to cash in on the Interweb c. 1995-1998 that they could do
it w/out paying some proprietary vendor and get better performance,
security and support to boot (I reported a bug in Apache JServ in 1998
and a fix was released by the time I came back from lunch 1/2 hour
later). Linux was a tool for UNIX sysadmins and admin wannabes to
practice their UNIX chops at home - or a conveniently inexpensive
platform on which to run Apache. Companies -- other than Linux
distributors -- didn't bet their business on it.

Apache's success greatly contributed to the corporate acceptance of Linux, IMHO.

Dave

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

[david]

On Wed, 20 Jun 2007, Al Boldi wrote:

> Scott Preece wrote:
>> On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
>>> Nicolas Mailhot wrote:
>>>> Tivo didn't make the Linux success. More Tivos can definitely undo it.
>>>
>>> I don't think so.
>>>
>>> First, it's not Linux that made success, but rather GNU that uses Linux
>>> as its kernel.  And, believe it or not, when people say Linux, they
>>> really mean GNU.  People could care less what kernel they were running,
>>> as long as the system is up and runs the procs that offer their
>>> services.
>>
>> ---
>>
>> Actually, for use in devices (like TiVos or cell phones), it is very
>> definitely the kernel that is of interest. Many such devices use
>> little or no GNU software (some manufacturers have consciously avoided
>> it because of the possibility of shifts like the GPLv3 changes).
>
> Sure, but was it Linux in embedded devices that made Linux what it is today,
> or was it GNU/Linux?

if it was the GNU that made linux what it is today and the linux kernel 
mearly an oppurtunist then the GNU/Hurd, GNU/Solaris, GNU/BSD and for that 
matter GNU/Microsoft distributions should be steadily and quickly gaining 
marketshare

oh, sorry, people barely know that they exist (including people convinced 
that the linux kernel is junk compared to Solaris) and they are little 
more than proof-that-it-can-be-done showpieces.

it was the ability of the linux kernel to adapt to vastly different 
hardware (including embeded hardware) that made Linux what it is today.

and I do mean 'Linux" not 'GNU/Linux' in the statement above.

David Lang

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

[david]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> On Jun 19, 2007, david@lang.hm wrote:
>
>> if you also make the assumption that the company won't use propriatary
>> software instead then I think you would get agreement.
>
> Ah, good point.  When I posed the one of the two cases of the inicial
> scenario as "no tivoization", I meant Free Software without
> constraints.
>
>> but the disagrement is over this exact assumption. you assume that
>> these companies will use non-tivoized products if you make it hard
>> to use the software covered by the GPL, most other people are saying
>> that they disagree and the result would be fewer companies useing
>> software covered by GPL instead.
>
> I understand that.  And what I'm saying is that, even if fewer such
> companies use GPLed software, you may still be better off, out of
> additional contributions you'll get from customers of companies that
> switch from tivoization to unconstrained Free Software, because of the
> additional costs of the alternatives.
>
> And no, I can't prove it, but it's good that at least the argument is
> no longer completely disregarded while something else is disputed.
>
>
> Now that you guys at least understand what the argument is, you can
> figure out the solution by yourselves.

good, if you are no longer going to claim that your opinion on this 
unprovable point is the Truth (not the capitol) hopefully you can accept 
that a lot of very smart people are convinced that you are wrong on this 
point and not just 'confused'

David Lang

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

[Dave Neuer]

On 6/19/07, Jan Harkes <jaharkes@cs.cmu.edu> wrote:
>
> You keep referring to the four freedoms so I googled for them and found
>
>     http://www.gnu.org/philosophy/free-sw.html
>
> So which of the freedoms did Tivo take away?
>
>     * The freedom to run the program, for any purpose (freedom 0).
>
>     * The freedom to study how the program works, and adapt it to
>       your needs (freedom 1). Access to the source code is a
>       precondition for this.
>
>     * The freedom to redistribute copies so you can help your neighbor
>       (freedom 2).
>
>     * The freedom to improve the program, and release your improvements
>       to the public, so that the whole community benefits (freedom 3).
>       Access to the source code is a precondition for this.
>
> It doesn't seem to me they took away freedoms 1, 2 or 3. They released
> the source to any free software components and we can study, modify,
> redistribute, improve and release our improvements for the benefit of
> the whole community.
>
> btw. freedom 3 seems to be just repeating what we already got from
> freedoms 1 and 2.
>
> So the only one we could differ in opinion about is freedom 0. I would
> say that they in no way are limiting my use of the Linux kernel (which
> is the part I mostly care about) I can run the program for any purpose I
> see fit. What if I want to run mythtv on my PC at home? Tivo has no
> control whether or not I can do so even when my kernel contains any of
> their modification or improvements, so I claim that I in fact retained
> freedom 0.

Much as I hate to extend the life of this execrable thread, since I
think Alexandre makes Sisyphus look like a hard-nosed pragmatist, it
seems pretty clear that TiVO impinges "[my] freedom to run the
program, for any purpose" if "any purpose" includes "make my TiVO do
what I want," and likewise to "adapt it to [my] needs" -- freedoms 0
and part of 1. It is just disingenuous to argue otherwise.

Dave

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

[Al Boldi]

Scott Preece wrote:
> On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
> > Nicolas Mailhot wrote:
> > > Tivo didn't make the Linux success. More Tivos can definitely undo it.
> >
> > I don't think so.
> >
> > First, it's not Linux that made success, but rather GNU that uses Linux
> > as its kernel.  And, believe it or not, when people say Linux, they
> > really mean GNU.  People could care less what kernel they were running,
> > as long as the system is up and runs the procs that offer their
> > services.
>
> ---
>
> Actually, for use in devices (like TiVos or cell phones), it is very
> definitely the kernel that is of interest. Many such devices use
> little or no GNU software (some manufacturers have consciously avoided
> it because of the possibility of shifts like the GPLv3 changes).

Sure, but was it Linux in embedded devices that made Linux what it is today, 
or was it GNU/Linux?


Thanks!

--
Al

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

[Scott Preece]

On 6/19/07, Al Boldi <a1426z@gawab.com> wrote:
> Nicolas Mailhot wrote:
> >
> > Tivo didn't make the Linux success. More Tivos can definitely undo it.
> >
>
> I don't think so.
>
> First, it's not Linux that made success, but rather GNU that uses Linux as
> its kernel.  And, believe it or not, when people say Linux, they really mean
> GNU.  People could care less what kernel they were running, as long as the
> system is up and runs the procs that offer their services.
---

Actually, for use in devices (like TiVos or cell phones), it is very
definitely the kernel that is of interest. Many such devices use
little or no GNU software (some manufacturers have consciously avoided
it because of the possibility of shifts like the GPLv3 changes).

scott

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

[Alexandre Oliva]

On Jun 19, 2007, "Josh Williams" <yurimxpxman.lkml@gmail.com> wrote:

> On 6/18/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> Free Software is not about freedom of choice.  That's an OSI slogan
>> for "if you like, you can shoot your own foot, regardless of whether
>> the shrapnel hurts people around you".
>> http://www.fsfla.org/?q=en/node/139#1

> When it comes right down to it, the FSF is all about freedom, just
> as long as it benefits them.

How is "ensuring users have their freedoms respected" beneficial to
the FSF, specifically?  I can see how that's good for users (and FSF
users software), how that's good for thriving Free Software
development communities (in some of which the FSF participates),
but it looks like you're aiming at more than that.  Can you be more
specific about the benefits the FSF collects out of this

>> Free Software is about respect for the four freedoms.

> Yes, there *are* four freedoms, and they are very good. However,
> what happens when we have *only* these four freedoms? The FSF would
> do anything within their power to impose their "religious" beliefs
> upon the entire world, regardless of what's really good for us.

Freedom is good for you, such that you can choose.

> If we don't have the freedom of choice,

You do.  That's a consequence of the four freedoms.

Without them, you end up without choice.

It's not a core freedom for the FSF precisely because freedom of
choice can be used to do common good as much as it can to hurt your
neighbor.  The four freedoms are all freedoms with which you can help
yourself or your neighbor, never hurt anyone.  You can choose not to
help yourself or your neighbor, and then the world is no better off,
but if you choose to help yourself or your neighbor, there's progress.

Now, freedom of choice can be used against the common good.  Choosing
to shoot your own feet, on grounds that freedom of choice says you
can, could still hurt your neighbors with the shrapnel.  That's why
freedom of choice is regarded as a secondary freedom.  Which is not to
say that it's undesirable, just that it's not fundamentally conducive
of the common good.  It is conducive of common good, as it turns out,
in as much as it is a consequence of the other four freedoms.

> Each person must be free to make *his*own* decisions of what's best
> for him,

As long as this doesn't hurt society.

> If the FSF insists on defining freedom for us

It doesn't.  It only defines what Free Software means.

Do you by any chance have a problem that OSI defines what "Open Source
Software" means, even though it doesn't mean software whose sources
are open?

Why would you have a problem that the FSF defines what Free Software
means, then?

If there were other definitions of Free Software out there, it might
make sense to qualify them, but there aren't, and coming up with
one now that meant something different would be quite confusing.

> Yes, I understand and appreciate the problems with proprietary software, but
> that's not the _only_ issue that needs dealt with in this world.

Sure.  I don't think anyone could disagree with this.

It just so happens that the Free Software Foundation is a foundation
with a registered mission to work on this particular cause.  Whether
it's a good cause or not, whether it's the most important cause of all
or not, can sure be debated among those who'd like to debate it (I
don't :-)

But if RMS, myself and so many others decided to devote our lives to
this cause, and a lot of people agree it's good for society, many
agree it's a good cause, even if not everyone likes our methods, who's
to tell us we ought to work on some other cause?  (FWIW, even if
someone asked me, told me or insisted that I do it, I wouldn't go
"ooh, they're *forcing* me to work on some other cause!", that would
be silly :-)

>> It was OSI that tried to create a definition that matched exactly
>> the meaning of the Free Software definition under "more objective
>> criteria".  We already know they failed, since the Reciprocal
>> Public License is accepted as an OSS license, but it's a non-Free
>> Software license.  There may be other examples.

> How could an open source license not being a free software license be
> considered a failure?

Err, see above.  It didn't achieve the stated goal.

It's indeed difficult to define freedom.

> Pardon me for saying so, but I don't think the OSS developers really
> give a crap whether their software is considered to be "free" as
> long as their goals are met.

http://www.opensource.org/node/130

>> That said, since a number of people already understand the GPLv2
>> prohibits tivoization, your argument means that either the comment in
>> the OSD is wrong, and GPLv2 already fails to match the OSD, or that
>> GPLv3 complies with it in just the same way.

> *what*?? That's the first I've heard of this. Nonsense.

Ask Alan Cox, for example, he got legal advice about this.  Maybe
Harald Welte shares the same opinion, certainly backed by good legal
advice.

That said, maybe it couldn't be upheld in a US court.

This was covered in great detail early in this thread.

> If that were true, we wouldn't be having this debate, am I correct?

It feels nonsensical indeed, doesn't 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

[Jan Harkes]

On Tue, Jun 19, 2007 at 06:20:24PM -0300, Alexandre Oliva wrote:
> On Jun 19, 2007, Jan Harkes <jaharkes@cs.cmu.edu> wrote:
> > and which will most likely make GPLv3 software unusable for various
> > applications ranging from medical equipment to financial transaction
> > systems (and probably others)
> 
> Not unusable, except perhaps for the one example about credit card
> terminals presented so far.
> 
> > is there to just make it a _bit_ more inconvenient for vendors to
> > implement a tivo-like scheme?
> 
> I'm not sure they find it to be "just a bit".
> 
> Point is to keep Free Software Free freedoms, and ROM doesn't make it
> non-Free, so this provision is a means to ensure the compliance with
> the wishes of users who want their software to not be used in ways
> that make it non-Free.

You keep referring to the four freedoms so I googled for them and found

    http://www.gnu.org/philosophy/free-sw.html

So which of the freedoms did Tivo take away?

    * The freedom to run the program, for any purpose (freedom 0).

    * The freedom to study how the program works, and adapt it to
      your needs (freedom 1). Access to the source code is a
      precondition for this.

    * The freedom to redistribute copies so you can help your neighbor
      (freedom 2).

    * The freedom to improve the program, and release your improvements
      to the public, so that the whole community benefits (freedom 3).
      Access to the source code is a precondition for this.

It doesn't seem to me they took away freedoms 1, 2 or 3. They released
the source to any free software components and we can study, modify,
redistribute, improve and release our improvements for the benefit of
the whole community.

btw. freedom 3 seems to be just repeating what we already got from
freedoms 1 and 2.

So the only one we could differ in opinion about is freedom 0. I would
say that they in no way are limiting my use of the Linux kernel (which
is the part I mostly care about) I can run the program for any purpose I
see fit. What if I want to run mythtv on my PC at home? Tivo has no
control whether or not I can do so even when my kernel contains any of
their modification or improvements, so I claim that I in fact retained
freedom 0.

Your position (and I hope I will get this right), is that they should
allow you to run the program for any purpose _on the hardware it is
installed on_. Interpreted that way, the whole modification part doesn't
even come into play, your freedom is taken away if you cannot even run
your own software on top of the already installed kernel.

This 'freedom 0' the way (I hope) you are interpreting it, could clearly
never have been protected by the GPLv2 since it explicitly does not
cover "Activities other than copying, distribution and modification"
(GPLv2, term 0)

However the GPLv3 does not seem to address this point either, the whole
discussion in section 5 about user products and installation information
completely misses the fact that none of the 'four freedoms' (which I
assume formed the foundation for the license) is about allowing a user
to install the program on some particular piece of hardware and that is
exactly where I think all this anti-tivoization language is going wrong.
It is clearly having a hard time pinning down the exact requirements for
something that was not well defined in the first place.

The real issue with Tivo isn't that they signed their kernel or their
initrd, but that they do not allow you to use that kernel for any
purpose, for instance run mythtv or a webserver on their unmodified
kernel. Maybe the GPLv3 shouldn't try to talk about license keys or
installation information and then heap on some exceptions for
non-consumer devices or rom-based implementations and then some further
legaleze patches to close the really obvious loopholes. Maybe it if it
actually addressed the fact that you want to be able to use the
distributed software for any purpose you see fit by allowing you to run
your own applications on the kernel they distributed.

I still believe they do allow me to use the program for any purpose as
they can not limit my use of the Linux kernel whether or not my copy
contains any of their contributions, so excuse my while I'll go enjoy
some more of my freedom 0.

Jan

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

[Alexandre Oliva]

On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> Company X has requirement for restriction Y
>   => License on product Z disallows restriction Y
>   => Product Z loses Company X and the exposure use in their product gives
>   => License on product Z is bad for the product

> Understandable now?

Well, considering that I've made this claim myself as part of my
complete argument in a number of times I've presented it, yes, this is
understandable and correct.  This is indeed one of the cases.

That said, very few companies have a scrict *requirement* for keeping
the ability to modify the software on the customer's computer while
denying this ability to the customer.

So this case you're discussing is the least common case.  It could
nearly be dismissed, rather than being the dominating topic in the
discussion, as it's been so far.

> What I was stating is that are legal (and other reasons) why a
> company might have to lock down their software in a process similar
> to "Tivoization".

Ok.  Most of these can be addressed (with inconvenience) with ROM.
Others are business reasons, and for these, the increased cost and
inconvenience of the alternatives may shift them to an unlock
situation.

>> Therefore, this claim is false.

> Only when you define a term as specifically as you have done 
> for "Tivoization".

It's not my definition.  This was from Wikipedia.

> I should, perhaps, have used a different term - it would 
> then have been patently true.

Depends on what the different term was.

-- 
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 19, 2007, david@lang.hm wrote:

> if you also make the assumption that the company won't use propriatary
> software instead then I think you would get agreement.

Ah, good point.  When I posed the one of the two cases of the inicial
scenario as "no tivoization", I meant Free Software without
constraints.

> but the disagrement is over this exact assumption. you assume that
> these companies will use non-tivoized products if you make it hard
> to use the software covered by the GPL, most other people are saying
> that they disagree and the result would be fewer companies useing
> software covered by GPL instead.

I understand that.  And what I'm saying is that, even if fewer such
companies use GPLed software, you may still be better off, out of
additional contributions you'll get from customers of companies that
switch from tivoization to unconstrained Free Software, because of the
additional costs of the alternatives.

And no, I can't prove it, but it's good that at least the argument is
no longer completely disregarded while something else is disputed.


Now that you guys at least understand what the argument is, you can
figure out the solution by yourselves.

-- 
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

[Josh Williams]

On 6/18/07, Alexandre Oliva <aoliva@redhat.com> wrote:

> ... derived from the Debian Free Software Guidelines, engineered to
> reflect the Free Software definition ...

Yes, that's true, but it was modified in several key points. OSS and FS
developers have a very similar approach to developing software, but our goals
are different. So why does it come to a shock that the open source definition
greatly resembles the four rights of free software?

> Err...  Excuse me?  Whole point for whom?

> Free Software is not about freedom of choice.  That's an OSI slogan
> for "if you like, you can shoot your own foot, regardless of whether
> the shrapnel hurts people around you".
> http://www.fsfla.org/?q=en/node/139#1

And in Rms's opinion, a woman's freedom to make decisions about her  own body
are hers to choose, regardless of whether this means murdering an innocent
child. (Do I see a conflict here?)

When it comes right down to it, the FSF is all about freedom, just as long as
it benefits them. They couldn't care less whether our society is productive.
For them, the software is not top priority, as they have been very vocal
about, and as Linus mentioned earlier in this thread.

> Free Software is about respect for the four freedoms.

Yes, there *are* four freedoms, and they are very good. However, what happens
when we have *only* these four freedoms? The FSF would do anything within
their power to impose their "religious" beliefs upon the entire world,
regardless of what's really good for us.

If we don't have the freedom of choice, then are we really free? Each person
must be free to make *his*own* decisions of what's best for him, even if it
means one of his appliances is "tivoized". That's really *his* decision,
isn't it?

If the FSF insists on defining freedom for us (and very narrowly, at that),
then perhaps we'd be better off without them. After all, isn't true freedom
the right to choose whatever freedoms you want to practise?

Yes, I understand and appreciate the problems with proprietary software, but
that's not the _only_ issue that needs dealt with in this world. It's getting
to the point where the FSF has become just as bad as Microsoft, yet in a
different way.

> I don't think the FSF is at all concerned whether GPLv3 complies with
> the OSD.  They couldn't care less.

<sarcasm>
  Are you serious? WOW! =O
</sarcasm>

>                                            It was OSI that tried to create a
> definition that matched exactly the meaning of the Free Software
> definition under "more objective criteria".  We already know they
> failed, since the Reciprocal Public License is accepted as an OSS
> license, but it's a non-Free Software license.  There may be other
> examples.

How could an open source license not being a free software license be
considered a failure? Pardon me for saying so, but I don't think the OSS
developers really give a crap whether their software is considered to
be "free" as long as their goals are met.

> That said, since a number of people already understand the GPLv2
> prohibits tivoization, your argument means that either the comment in
> the OSD is wrong, and GPLv2 already fails to match the OSD, or that
> GPLv3 complies with it in just the same way.

*what*?? That's the first I've heard of this. Nonsense. If that were true, we
wouldn't be having this debate, am I correct?

-- 
It's common knowledge that most intruders come in through Windows.

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

[david]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
> 
> On Jun 19, 2007, david@lang.hm wrote:
>
>> if a company doesn't care about tivoizing then they won't do it, it
>> takes time and money to tivoize some product and it will cause
>> headaches for the company.
>
>> their reasons for wanting to tivoize a product may be faulty, but they
>> think that the reasons are valid or they wouldn't go to the effort.
>
> Absolutely right.  And we'll get to that.  Please just be patient.
>
> In fact, how much the company cares about tivoizing is completely
> irrelevant to that point.  I shouldn't even have included it, but I
> did because I thought it would be useful as a boundary condition.
>
> So just disregard that.
>
> Is there agreement that, comparing tivoized and non-tivoized hardware,
> we get'd more contributions if the hardware is not tivoized, because
> users can scratch their own itches, than we would for tivoized
> hardware?

if you also make the assumption that the company won't use propriatary 
software instead then I think you would get agreement. but the 
disagrement is over this exact assumption. you assume that these companies 
will use non-tivoized products if you make it hard to use the software 
covered by the GPL, most other people are saying that they disagree and 
the result would be fewer companies useing software covered by GPL 
instead.

David Lang

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

[Daniel Hazelton]

On Tuesday 19 June 2007 19:49:24 david@lang.hm wrote:
> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
> > On Jun 19, 2007, david@lang.hm wrote:
> >> remember, not all tivo models are locked down,
> >
> > Only the earliest that you can't find for sale any more, right?
> >
> >> as a result of watching the hacker groups I can safely say that the
> >> lockdown has not blocked many users. it has slowed modification of the
> >> hacks to new types of hardware, but not for very long.
> >
> > Well, then...  What's the point *for* tivoization, again?  To slow
> > down the contributions?  And that's good because...?
>
> no, the point is that while tivoization is not nessasarily the best thing
> it's far better then the company useing propriatary code.
>
> delayed contributions are better then no contributions.
>
> David Lang

This logic has been proven already. Apple used KHTML and KJS as the backend 
systems for Safari. While Safari was in development they held onto all their 
changes and modifications. When they released Safari, they contributed it all 
back, making both things better. Fact: KHTML and KJS are also the core of 
Apples "WebKit" and "WebCore" technologies - both KHTML and KJS are Open 
Source projects, making up a part of KDE.

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 19, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> Right to control what software runs on the hardware is no different.
>> For any hardware on which I can run the software, I'm a user there,
>> and I'm entitled to the rights granted by the license.

> Exactly. However, that right does not include the right to run the software
> on any particular piece of hardware.

I think that's correct, in general.  But when I receive the software
along with a particular piece of hardware, on which I'm to expected to
run it, I'm also receiving all other freedoms that ought to be
respected by whoever distributed the software to me, which means the
distributor can't impose restrictions on my running modified versions
of the program there (or anywhere else I can run the software, for
that matter).

> It includes the right to run the software on *ANY* hardware, so long
> as one is authorized to choose what software runs on that hardware.

That's correct, too, with the provision that the distributor cannot
impose restrictions on running modified versions of the program.  Not
on any other hardware I can control, not on the hardware with which I
received the software along with the freedoms to control it in as far
as running the GPLed software goes.

> You keep conflating things like access to the source code and the legal
> right to modify that source code with the authorization right to install
> that modified kernel on some particular piece of hardware. The GPL was never
> about how such authorization decisions are made.

That's correct.  All it says is that you can't impose further
restrictions.  This means you can't use anything (source code
deprivation, patents, copyright, authorization, nothing) to disrespect
users' freedoms.  Is this so hard to accept?  There's nothing special
about these authorization rights you're talking about that distinguish
them from copyrights or patent rights or any other rights.  The GPL
says: don't use them against users' freedoms regarding the GPLed
software.  No excuses.

> If I make you a user of my laptop, do I have to let you install a modified
> kernel?

Does interacting with your laptop over the network make me a user of
your kernel?

> What if I even let you access the kernel source, so I have
> distributed it to you.

You have distributed the source to me, I can modify it.

But did you distribute the kernel binary installed on your computer to
me?  That's the one I'm (possibly) using per the above.  Whether you
also gave me sources is irrelevant, unless you gave them to me as part
of your obligation of distributing the corresponding sources, in case
we conclude you distributed the kernel binary installed on your
computer to me when you granted me remote access to it.

And then, GPLv3 makes it clear that, in the case of remote access,
you're not conveying the binary to me, therefore the conditions for
conveying do not apply.

> The GPL was never, ever about such authorization decisions. They are
> completely alien to both the wording and the spirit of the GPL.

You can repeat that as much as you want, this won't change the fact
that the GPL has never permitted you to use whatever rights you have
to impose restrictions on users' freedoms as to GPLed software once
you've (implicitly) accepted the conditions 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

[david]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> On Jun 19, 2007, david@lang.hm wrote:
>
>> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
>>> You're losing all that.
>
>> based on the knowledge shown by these users you aren't loosing much.
>
> Remember, the sample is biased, the hackers who'd like to hack it are
> less likely to buy it, and some might be only making private changes
> to avoid legal hassles.
>
> And then, it doesn't have to be all that much.  Since you're talking
> specifically about TiVo customers, how much has TiVo effectly
> contributed in terms of code?  (It was pointed out before, I know, but
> it's important to keep this in perspective, lest people lose sight of
> what's at stake)

they contributed all changes they made to GPL code, exactly as required.

in addition they have provided the full build environments, not just 
listing the versions of the compilers, etc.

>> remember, not all tivo models are locked down,
>
> Only the earliest that you can't find for sale any more, right?
>
>> as a result of watching the hacker groups I can safely say that the
>> lockdown has not blocked many users. it has slowed modification of the
>> hacks to new types of hardware, but not for very long.
>
> Well, then...  What's the point *for* tivoization, again?  To slow
> down the contributions?  And that's good because...?

no, the point is that while tivoization is not nessasarily the best thing 
it's far better then the company useing propriatary code.

delayed contributions are better then no contributions.

David Lang

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

[Alexandre Oliva]

On Jun 19, 2007, david@lang.hm wrote:

> if a company doesn't care about tivoizing then they won't do it, it
> takes time and money to tivoize some product and it will cause
> headaches for the company.

> their reasons for wanting to tivoize a product may be faulty, but they
> think that the reasons are valid or they wouldn't go to the effort.

Absolutely right.  And we'll get to that.  Please just be patient.

In fact, how much the company cares about tivoizing is completely
irrelevant to that point.  I shouldn't even have included it, but I
did because I thought it would be useful as a boundary condition.

So just disregard that.

Is there agreement that, comparing tivoized and non-tivoized hardware,
we get'd more contributions if the hardware is not tivoized, because
users can scratch their own itches, than we would for tivoized
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

[Alexandre Oliva]

On Jun 19, 2007, david@lang.hm wrote:

> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
>> You're losing all that.

> based on the knowledge shown by these users you aren't loosing much.

Remember, the sample is biased, the hackers who'd like to hack it are
less likely to buy it, and some might be only making private changes
to avoid legal hassles.

And then, it doesn't have to be all that much.  Since you're talking
specifically about TiVo customers, how much has TiVo effectly
contributed in terms of code?  (It was pointed out before, I know, but
it's important to keep this in perspective, lest people lose sight of
what's at stake)

> remember, not all tivo models are locked down,

Only the earliest that you can't find for sale any more, right?

> as a result of watching the hacker groups I can safely say that the
> lockdown has not blocked many users. it has slowed modification of the
> hacks to new types of hardware, but not for very long.

Well, then...  What's the point *for* tivoization, again?  To slow
down the contributions?  And that's good because...?

> in the case of Tivo, there are many public boards that talk about
> hacking tivos, and tivo employees login and contribute to them.

Nice!


Thanks for the info, this was very enlightening.

-- 
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 Tuesday 19 June 2007 13:06:17 Alexandre Oliva wrote:
> On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Tuesday 19 June 2007 04:04:52 Alexandre Oliva wrote:
> >> On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> >> > On Tuesday 19 June 2007 02:44:32 Alexandre Oliva wrote:
> >> >> GPLv3 forbids tivoization, therefore developer has requirement for
> >> >> tivoization in the license, therefore GPLv3 forbidding tivoization
> >> >> is bad.
> >> >
> >> > However, my argument is straight logic, nothing "circular" about it. 
> >> > :) Replacing "X" in my logic path above with "tivoization" and
> >> > "license" with "GPLv3", as you've done, does produce a valid chain of
> >> > logic.
> >>
> >> Yes.  Isn't it funny though that tivoization became necessary as a
> >> consequence of GPLv3 forbidding it?
> >
> > -ELOGIC
>
> I see.  Try 'modprobe logic', it worked for me years ago ;-) :-D

Try using real logic rather than the logic of your religion.

> > It didn't become necessary as a result of the GPLv3 forbidding it.
>
> Which is why I said it was funny, because your inference chain stated
> *exactly* (with an implied "for the developers") that it did.
>
> Do you understand what an inference chain is?  A => B, as in A implies
> B, which can also be read as A therefore B if A is known to hold.

Okay, since you want it in a specific language rather than as a set of bullet 
points (which is what I used the => for)

Company X has requirement for restriction Y
  => License on product Z disallows restriction Y
  => Product Z loses Company X and the exposure use in their product gives
  => License on product Z is bad for the product

Understandable now?

> > there could be any number of reasons why "tivoization" is needed by
> > the manufacturer.
>
> This claim is false.
>
> Tivoization is when hardware manufacturer takes copyleft software and
> blocks updates by the user of the hardware.

By that definition you are correct.

> No single law so far has shown an example that even resembled to
> mandate copyleft software, and no contract could possibly establish a
> condition like this.

No argument here. What I was stating is that are legal (and other reasons) why 
a company might have to lock down their software in a process similar 
to "Tivoization".

> Therefore, this claim is false.

Only when you define a term as specifically as you have done 
for "Tivoization". I should, perhaps, have used a different term - it would 
then have been patently true. Though, at that point, you would likely have 
argued that it wasn't "tivoization"

> > This whole bit was to point out that you were inferring circular
> > logic where none existed.
>
> There *is* circular logic is in place.
>
> The initial premise of this fallacy is that anti-tivoization is bad
> for the project.
>
> This is used to conclude that licenses with such provisions should be
> rejected.
>
> This is then used to conclude that there are fewer developers who
> would develop under such licenses.
>
> Which is then used to conclude that anti-tivozation is bad for the
> project.

Your view of the logic is, in this case, flawed. It's more along the lines of 
what I've detailed above. Now, please, go away. You aren't doing 
your "religion" any good. In fact, you are damaging it - repeatedly.

DRH

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

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

[Al Boldi]

Nicolas Mailhot wrote:
>
> Tivo didn't make the Linux success. More Tivos can definitely undo it.
>

I don't think so.

First, it's not Linux that made success, but rather GNU that uses Linux as 
its kernel.  And, believe it or not, when people say Linux, they really mean 
GNU.  People could care less what kernel they were running, as long as the 
system is up and runs the procs that offer their services.

It was probably a strategic mistake for GNU to lock into Linux, instead of 
developing a system that would allow GNU to plug-and-play the Kernel.  A 
mistake that could easily be rectified, given enough desire.

Second, GPLv2 was/is probably instrumental to the success of GNU/Linux.  
That's because GPLv2 seems to be fair, which allows the largest possible 
community to form.  Changing this fairness, driven by paranoia arguments, to 
hamper commercial entities to enter this community may possibly reduce the 
growth of this community.

Interestingly, it's this same paranoia that drives kernel developer to 
renounce a "Stable Kernel API" for fear of exploitation by certain 
commercial entities.  Which makes you kind of wonder, why they wouldn't move 
to GPLv3.

Or does paranoia and paradox go hand in hand?


Thanks!

--
Al

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

[david]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> On Jun 19, 2007, david@lang.hm wrote:
>
>> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
>>>
>>> Once again, now with clearer starting conditions (not intended to
>>> match TiVo in any way, BTW; don't get into that distraction)
>>>
>>>
>>> Vendor doesn't care about tivoizing, their business works the same
>>> either way.
>>>
>>> Vendor's employees will contribute the same, one way or another, so
>>> their contributions are out of the equation.
>
>> no, this is what you are missing.
>
> No, this is just one of many possible situations.
>
> You're just thinking of a different scenario, in which tivoization is
> important for the tivoizer.  We'll get to that.
>
> Please be patient.  Don't get the illusion that you're disputing or
> weakening the argument with these statements.  Disputing premises of a
> logical inference is pointless as far as disputing a larger argument
> that this particular lemma would be a part of.

if a company doesn't care about tivoizing then they won't do it, it takes 
time and money to tivoize some product and it will cause headaches for the 
company.

their reasons for wanting to tivoize a product may be faulty, but they 
think that the reasons are valid or they wouldn't go to the effort.

David Lang

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

[Alexandre Oliva]

On Jun 19, 2007, "Dave Neuer" <mr.fred.smoothie@pobox.com> wrote:

> On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> But it takes only a small fraction of the tivoizers to decide to take
>> out the locks, when faced with the costs mentioned above, for us to
>> gain contributions from even a small fraction of their user base
>> (which would then grow in hacker density as a result of
>> non-tivoization) for us to end up better off.

> Even if you're correct, that only takes into account the manufacturers
> who are using Linux _now_ who might be pressured to allowed modified
> versions to run. What about the lost opportunity cost of all of the
> future manufacturers who decide to use ProprietaryOS + locks instead
> of Linux? We don't get any of their code.

True.  This is not left out of my complete argument, though.

-- 
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 19, 2007, david@lang.hm wrote:

> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
>> 
>> Once again, now with clearer starting conditions (not intended to
>> match TiVo in any way, BTW; don't get into that distraction)
>> 
>> 
>> Vendor doesn't care about tivoizing, their business works the same
>> either way.
>> 
>> Vendor's employees will contribute the same, one way or another, so
>> their contributions are out of the equation.

> no, this is what you are missing.

No, this is just one of many possible situations.

You're just thinking of a different scenario, in which tivoization is
important for the tivoizer.  We'll get to that.

Please be patient.  Don't get the illusion that you're disputing or
weakening the argument with these statements.  Disputing premises of a
logical inference is pointless as far as disputing a larger argument
that this particular lemma would be a part of.

-- 
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]

> > You keep smuggling in the same assumption without ever
> > defending it. There
> > is a user. There is a person who gets to decide what software runs on a
> > particular piece of hardware. You keep assuming they must be the same
> > person.

> No, I'm just saying that whoever gets to decide cannot restrict the
> user's freedoms as to the software the user received.

I agree. However, the freedom to run modified software on hardware for which
one is not the person who gets to decide what software runs on that hardware
is not one of those freedoms.

> Consider this:
>
> I get GPLed software.
>
> I make improvements to it.
>
> I give it to you, but I leave out the sources of my changes.
>
> You ask me for sources, because without them you can't enjoy the
> freedom to adapt the software.
>
> I say "No, they're mine.  I have the right to keep them and release
> them however I like.  Copyright law says so.!
>
> You talk to the copyright holder, and he revokes my license and gets a
> court order such that I can't distribute the software any more.

Right.

> You see?  It's not because I had a right that I can use it to impose
> restrictions on your freedoms, after I distribute the software to you.

Right.

> Right to control what software runs on the hardware is no different.
> For any hardware on which I can run the software, I'm a user there,
> and I'm entitled to the rights granted by the license.

Exactly. However, that right does not include the right to run the software
on any particular piece of hardware. It includes the right to get the source
code, redistribute it, modify it, and so on. It includes the right to run
the software on *ANY* hardware, so long as one is authorized to choose what
software runs on that hardware.

> It's really this simple.  Don't complicate the issue by trying to make
> hardware special.  It's just an illusion to try to convince yourself
> that you can deprive users of freedoms provided by the GPL.

The right to run GPL'd software on some particular piece of hardware was
never a GPL right. That is an *authorization* decision, similar to who has
'root' access on a Linux box and can modify and install the kernel.

You keep conflating things like access to the source code and the legal
right to modify that source code with the authorization right to install
that modified kernel on some particular piece of hardware. The GPL was never
about how such authorization decisions are made.

If I make you a user of my laptop, do I have to let you install a modified
kernel? What if I even let you access the kernel source, so I have
distributed it to you.

The GPL was never, ever about such authorization decisions. They are
completely alien to both the wording and the spirit of the GPL.

DS

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

[Alexandre Oliva]

On Jun 19, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> Right.  All GPL can say is that you cannot impose further restrictions
>> on how the user adapts the software, and since the user runs the
>> software on that computer, that means you must not restrict the user's
>> ability to upgrade or otherwise replace that software there, when you
>> gave the user the software along with the computer.

> You keep smuggling in the same assumption without ever defending it. There
> is a user. There is a person who gets to decide what software runs on a
> particular piece of hardware. You keep assuming they must be the same
> person.

No, I'm just saying that whoever gets to decide cannot restrict the
user's freedoms as to the software the user received.


Consider this:

I get GPLed software.

I make improvements to it.

I give it to you, but I leave out the sources of my changes.

You ask me for sources, because without them you can't enjoy the
freedom to adapt the software.

I say "No, they're mine.  I have the right to keep them and release
them however I like.  Copyright law says so.!

You talk to the copyright holder, and he revokes my license and gets a
court order such that I can't distribute the software any more.


You see?  It's not because I had a right that I can use it to impose
restrictions on your freedoms, after I distribute the software to you.

Right to control what software runs on the hardware is no different.
For any hardware on which I can run the software, I'm a user there,
and I'm entitled to the rights granted by the license.


It's really this simple.  Don't complicate the issue by trying to make
hardware special.  It's just an illusion to try to convince yourself
that you can deprive users of freedoms provided by the GPL.

-- 
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 19, 2007, david@lang.hm wrote:

> On Tue, 19 Jun 2007, Alexandre Oliva wrote:
>> On Jun 19, 2007, Daniel Drake <dsd@gentoo.org> wrote:
>> 
>>> I realise that the latest GPLv3 draft would not pose restrictions
>>> here, as such devices would not be classified as consumer
>>> products.
>> 
>> And even if they were, there's always ROM.
>> 
>> I don't know whether hardware seals that state "once you break this
>> seal, law prohibits the use of this device with human patients".

> once you break the seal the device is no longer certified. an
> uncertified device cannot be used.

Yup.  That's the law.

At which point it's not the hardware vendor imposing the restriction,
so this use is perfectly acceptable.


One could presumably implement similar seals in software.  Nothing
wrong with a signature used to indicate that the device has been
tampered with.  Even a led somewhere that reflects this status.

None of this prevents the user from enjoying the freedoms he's
entitled to, according to the laws of the place where he lives.

-- 
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]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> On Jun 19, 2007, david@lang.hm wrote:
>
>> based on my experiance looking at the software released for tivos, I
>> think you are over-estimating these numbers. if there are more then a
>> dozen people producing things that are good enough to be useful and
>> releasing their results as opensource software I would be surprised.
>
> Yup.
>
> And how many more would there be should it not be tivoized?  More
> hackers would buy the devices, a number of them with the explicit
> intent and interest in modifying the software in it.
>
> You're losing all that.

based on the knowledge shown by these users you aren't loosing much.

remember, not all tivo models are locked down, and most of those that are 
locked down can be unlocked pretty easily. it's only a couple models that 
have required a soldering iron.

the xbox hacks show that people will produce ways to bypass hardware 
restrictions that are easy for less techinical people to use (and people 
less technical then this aren't gong to be contributing code anyway)

as a result of watching the hacker groups I can safely say that the 
lockdown has not blocked many users. it has slowed modification of the 
hacks to new types of hardware, but not for very long.

>> and for all that the FSF is claiming that tivos can't being modified
>> it's really not that hard to change.
>
> But is it legal?

that depends on what lawyer you ask, and what they think of the 
applicability of the DMCA.

> How many would contribute changes to a list where there are TiVo
> people watching, which might expose these contributors to liabilities?

in the case of Tivo, there are many public boards that talk about hacking 
tivos, and tivo employees login and contribute to them. just about all of 
these boards ban specific topics. for example:

getting service without paying for it is banned on every board I 
personally look at

transferring video off of the tivo is banned on some boards, but not on 
others.

David Lang

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

[Dave Neuer]

On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> On Jun 19, 2007, Anders Larsen <al@alarsen.net> wrote:
>
> > Only, your statement above seems to run counter to your previous claims
> > that the "anti-tivoisation" provisions of GPLv3 would bring _more_
> > developers to copyleft software.
>
> > So which one is it?
>
> We might lose your contributions, that's true, I've never ever denied
> that.  And this will even have a cost for you, especially if you go
> proprietary rather than some other more liberal Free Software license,
> or stick with a GPLv2 Linux and hope it's never ruled as prohibiting
> tivoization, or move to Linux on ROM.
>
> But it takes only a small fraction of the tivoizers to decide to take
> out the locks, when faced with the costs mentioned above, for us to
> gain contributions from even a small fraction of their user base
> (which would then grow in hacker density as a result of
> non-tivoization) for us to end up better off.

Even if you're correct, that only takes into account the manufacturers
who are using Linux _now_ who might be pressured to allowed modified
versions to run. What about the lost opportunity cost of all of the
future manufacturers who decide to use ProprietaryOS + locks instead
of Linux? We don't get any of their code.

But all of this is moot anyway. You are not going to win the argument
on practical grounds anyway since Linus, Greg, Ingo and several other
developers with collectively many lines of code in the kernel have
stated in one form or another that they don't agree with the ethical
goals of GPLv3. I don't understand who you think you're going to
convince w/ continued debate?

Dave

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

[Alexandre Oliva]

On Jun 19, 2007, Jan Harkes <jaharkes@cs.cmu.edu> wrote:

> On Tue, Jun 19, 2007 at 02:40:59AM -0300, Alexandre Oliva wrote:
>> > The actual software is mailed to you on a credit card sized
>> > ROM when you activate service.
> ...
>> The GPLv3 won't remove every way in which people who want/need to stop
>> the user from making changes to the software could accomplishing this
>> (ROM).  It will just make this a bit more inconvenient, such that
>> vendors that have the option respect users' freedoms, and those that
>> find it too inconvenient respect the wishes of users who don't want
>> their software turned non-free.

> Or are you saying that all that anti-tivoization language that adds
> complex requirements which change depending on the market some device
> happens to be sold in

You allude to the definition of User Product, not geographies, I
suppose.

> and which will most likely make GPLv3 software unusable for various
> applications ranging from medical equipment to financial transaction
> systems (and probably others)

Not unusable, except perhaps for the one example about credit card
terminals presented so far.

> is there to just make it a _bit_ more inconvenient for vendors to
> implement a tivo-like scheme?

I'm not sure they find it to be "just a bit".

Point is to keep Free Software Free freedoms, and ROM doesn't make it
non-Free, so this provision is a means to ensure the compliance with
the wishes of users who want their software to not be used in ways
that make it non-Free.

As it so happens, this also places economic pressure on vendors who
tivoize, such that they either face more costly solutions, or enable
users to tinker with the software as well.  And if neither ROM nor
permission are an option for the vendor, well, too bad, the author
gets to decide how his software is to be used, right?

> So what exactly is the point of all this then?

Keeping Free Software Free.

(and, as a consequence that many of you may welcome, keeping
open-source software open source)

-- 
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]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

>
> Once again, now with clearer starting conditions (not intended to
> match TiVo in any way, BTW; don't get into that distraction)
>
>
> Vendor doesn't care about tivoizing, their business works the same
> either way.
>
> Vendor's employees will contribute the same, one way or another, so
> their contributions are out of the equation.

no, this is what you are missing.

with tivoizing, the vendors employees are working on linux and 
contributing

without tivoizing and with GPLv3 license involved the vendors employees 
are working on a propriatary OS and are contributing nothing back

> Users get source code in either case, and they can modify it and share
> it.  They're in no way stopped from becoming part of the community.

not if the GPLv3 achieves it's objectives of forcing the vendor to stop 
useing opensource software.

>
> Given these conditions:
>
> In a tivoized device, users will be unable to scratch their itches.
> This doesn't stop them from contributing to the project, but they may
> lack self-interest motivation to contribute, because they won't be
> able to use their modifications in the device they own.
>
> In a non-tivoized device, users can scratch their itches.  They can
> contribute just as much as they would in a tivoized device, but since
> they can use the changes they make to make their own devices work
> better for them, this works as a motivator for them to make changes,
> and perhaps to contribute them.  Therefore, they will tend to
> contribute more.
>
>
> Can you point out any flaw in this reasoning, or can we admit it as
> true?

in a tivoized device, users have access to the source code and can 
understand how things work, incoporate improvements into other projects, 
etc.

In addition, users who bypass the lockdown restrictions can modify the 
software on that device.

in a non-tivoized device users have a black-box and have to reverse 
engineer everything becouse the vendor releases no source at all.

David Lang

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

[Alexandre Oliva]

On Jun 19, 2007, Anders Larsen <al@alarsen.net> wrote:

> Only, your statement above seems to run counter to your previous claims
> that the "anti-tivoisation" provisions of GPLv3 would bring _more_
> developers to copyleft software.

> So which one is it?

We might lose your contributions, that's true, I've never ever denied
that.  And this will even have a cost for you, especially if you go
proprietary rather than some other more liberal Free Software license,
or stick with a GPLv2 Linux and hope it's never ruled as prohibiting
tivoization, or move to Linux on ROM.

But it takes only a small fraction of the tivoizers to decide to take
out the locks, when faced with the costs mentioned above, for us to
gain contributions from even a small fraction of their user base
(which would then grow in hacker density as a result of
non-tivoization) for us to end up better off.

Or so I believe ;-)

-- 
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 19, 2007, david@lang.hm wrote:

> based on my experiance looking at the software released for tivos, I
> think you are over-estimating these numbers. if there are more then a
> dozen people producing things that are good enough to be useful and
> releasing their results as opensource software I would be surprised.

Yup.

And how many more would there be should it not be tivoized?  More
hackers would buy the devices, a number of them with the explicit
intent and interest in modifying the software in it.

You're losing all that.

> and for all that the FSF is claiming that tivos can't being modified
> it's really not that hard to change.

But is it legal?

How many would contribute changes to a list where there are TiVo
people watching, which might expose these contributors to liabilities?

-- 
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]

On Tue, 19 Jun 2007, Alexandre Oliva wrote:

> On Jun 19, 2007, Daniel Drake <dsd@gentoo.org> wrote:
>
>> I realise that the latest GPLv3 draft would not pose restrictions
>> here, as such devices would not be classified as consumer
>> products.
>
> And even if they were, there's always ROM.
>
> I don't know whether hardware seals that state "once you break this
> seal, law prohibits the use of this device with human patients".

once you break the seal the device is no longer certified. an uncertified 
device cannot be used.

this is very common (in some areas it's widely ignored, in others it 
isn't)

this is just like the 'you void the warranty if you disrupt this sticker' 
stickers that you see on just about any hardware you buy today. some 
vendors are stickers for this, others don't really care.

David Lang

> Then the restriction is not being imposed by the manufacturer, only by
> law, and this does make lot of a difference as far as software freedom
> is concerned.
>
> But then, law might not find this to be enough.  Software patents are
> not the only stupid law that harms Free Software :-(
>
>

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

[Pekka Enberg]

Hi Alexandre,

At some point in time, I wrote:
> > Furthermore, Linus has repeatedly explained to you why he (and bunch
> > of other _kernel_ hackers -- this is a thread on LKML, remember)
> > thinks it's stupid for a software license to restrict hardware
> > design choices.

On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> Most of these explanations carried the assumption that this would
> lower the amount of contributions he'd get.

Then may I suggest you go back and re-read those threads because
that's not at all what I am reading.

On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> Now, if people want to keep on fooling themselves, I guess they can,
> as long as they don't hurt others in the process.  If they try to fool
> others, I feel it is my moral duty to intervene.

Sure. I just wanted to point out that your "moral duty" really comes
off as "trolling" at this end. So could we please move this discussion
to maybe "moral-duties@lists.fsf.org" or some other more appropriate
list and stop flooding LKML?

                                      Pekka

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

[David Schwartz]

> Right.  All GPL can say is that you cannot impose further restrictions
> on how the user adapts the software, and since the user runs the
> software on that computer, that means you must not restrict the user's
> ability to upgrade or otherwise replace that software there, when you
> gave the user the software along with the computer.

You keep smuggling in the same assumption without ever defending it. There
is a user. There is a person who gets to decide what software runs on a
particular piece of hardware. You keep assuming they must be the same
person. There are *MANY* legitimate reasons why the user of a piece of
hardware should not be the same person who controls what software runs on
that hardware.

> >> As long as you didn't hand me the hardware along with the software,
> >> for me to become a user of the software on that hardware, I agree.

> > This is, again, an argument that is totally alien to the GPL.

> No, it's not.  It is intended to ensure that free software remains
> free for all its users.

Exactly, on all hardware. Not "especially free" on some one particular
piece.

> When you receive the software, you become a
> user.  That's when you receive the rights, and that's what creates the
> obligation on the distributor to not impose restrictions on the
> freedoms, no matter by how means such restrictions could be legally or
> technically accomplished.

Exactly. And they place no restrictions on your ability to modify or use
that software on any hardware you like, provided of course you are the
person who gets to decide what software runs on that hardware.

> > The idea that you have 'special' rights to the software on some
> > hardware but not others is simply insane.

> I agree, to some extent.  It's not so much about the rights, but about
> the restrictions the vendor can impose on hardware.

The GPL is about what restrictions a particule piece of hardware, that
contains no GPL'd software, can impose?

> It's just that, for this particular hardware, as you say, the
> manufacturer has (or had) special rights.  This means it can decide
> what software runs, whom it gives the hardware to, etc.

Exactly. For any given piece of hardware, there must be some person or
entity that decides what software runs on it.

> However, by distributing software under the GPL, the vendor accepts
> the condition to not use any means whatsoever to impose restrictions
> on the recipient's exercise of the rights granted by the license by
> means of the distribution of the software.

Agreed.

> There's no reason to make the hardware special, or the right of
> authorization special, as a possible excuse to impose restrictions on
> the user.  It amounts to just that: an attempt to excuse oneself from
> the condition of not imposing restrictions on the enjoyment of the
> freedoms.

There is always the restriction that if you aren't the person who gets to
choose what software runs on a particular piece of hardware, then you can't
run modified software on that hardware.

There has to be someone who makes that decision for any given piece of
hardware. The idea that this person *MUST* be the user is totally alien to
the GPL. It's got nothing whatsoever to do with *ANY* of the freedoms the
GPL was protecting. All of those freedoms very critically apply to *ALL*
hardware in the entire universe.

> > The GPL is about being able to use the software on *ANY* hardware
> > for which you have the right to decide what software runs.

> Yes.  And, per the "pass on all rights you have" spirit in the
> preamble, that translates into "no further restrictions" in the legal
> terms, the user *must* receive this right from the distributor of the
> software.

That right is a right to that particular piece of hardware, it is not a
right to the GPL'd software. Your argument suggests that if I let you use my
laptop, I must let you modify the Linux kernel on it. That's just craziness.
The GPL was never about who was authorized to install modified software on
particular pieces of hardware.

> > You can state what the GPLv3 does as many times as you want, but special
> > rights to particular pieces of hardware is *TOTALLY* alien to
> > the spirit of the GPL.

> I agree.  That's the bug in GPLv2 that the anti-tivoization provision
> is trying to fix.

You can see it as a bug, and if you think Tivoization of free software is
bad, then that view makes sense. However, if you see the GPL as being about
getting the software, being free to modify the software, being able to
install that software on *ANY* hardware (whether or not that hardware
shipped with open-souruce software) and if you see the GPL as avoiding any
restrictions on authorization decisions, then the GPLv3 is not "fixing"
something but radically doing something else entirely.

>From my point of view, the biggest problem with the GPLv3 is not the change
in spirit but the change in scope. The GPLv3 attempts to control the Tivo
hardware, firmware, and keys, none of which contain any GPL'd software at
all. The idea that because you use GPL'd software, restrictions are imposed
on non-GPL'd works bothers me tremendously.

DS

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

[Nicolas Mailhot]

[-- Attachment #1: Type: text/plain, Size: 962 bytes --]

Le mardi 19 juin 2007 à 21:56 +0200, Diego Calleja a écrit :

> Please, stop pretending you are hardware manufacturers. You are not.

Please, stop pretending the end user has no say in the GPL. The GPL (v2
or v3) is written with the end user not the hardware manufacturer in
mind.

Also I can tell you the enterprises who make the living of Red Hat,
Novell and IBM (to name some major kernel contributors) care very much
about their part of the GPL deal. That is they can dump a supplier
(hardware or software) at any moment because he has no lock on their
system. Should the kernel devs ally themselves with entities like Tivo
who put the vendor lock-in back in free/libre systems, said systems
attractivity will decrease sharply (and it only takes a few managers to
notice their Linux systems are just as locked as the usual proprietary
ones)

Tivo didn't make the Linux success. More Tivos can definitely undo it.

-- 
Nicolas Mailhot

[-- Attachment #2: Ceci est une partie de message numériquement signée --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

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

[Jan Harkes]

On Tue, Jun 19, 2007 at 02:40:59AM -0300, Alexandre Oliva wrote:
> > The actual software is mailed to you on a credit card sized
> > ROM when you activate service.
...
> The GPLv3 won't remove every way in which people who want/need to stop
> the user from making changes to the software could accomplishing this
> (ROM).  It will just make this a bit more inconvenient, such that
> vendors that have the option respect users' freedoms, and those that
> find it too inconvenient respect the wishes of users who don't want
> their software turned non-free.

I am trying to read that last sentence and it just doesn't seem to make
any sense.

Or are you saying that all that anti-tivoization language that adds
complex requirements which change depending on the market some device
happens to be sold in and which will most likely make GPLv3 software
unusable for various applications ranging from medical equipment to
financial transaction systems (and probably others) is there to just
make it a _bit_ more inconvenient for vendors to implement a tivo-like
scheme?

So what exactly is the point of all this then?

Jan

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

[Anders Larsen]

On 2007-06-19 20:23:00, Alexandre Oliva wrote:
> If you take the Wikipedia definition of Tivoization, you'll see it's
> about copyleft software only, and no law mandates the use of copyleft
> software.  There's no end to bad laws, but a law that mandated the use
> of copyleft (=> free) software and at the same time prohibited
> modifications by the user would be a very contradictory one.

You're absolutely right...

Nobody forces us to use Linux in the credit-card terminals I'm currently
working on; of course we could have selected a proprietary solution (and
we would be forced to, were the Linux kernel and/or certain crucial
libraries or utilities GPLv3 only). 

Only, your statement above seems to run counter to your previous claims
that the "anti-tivoisation" provisions of GPLv3 would bring _more_
developers to copyleft software.

So which one is it?

Cheers
 Anders

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

[Alexandre Oliva]

On Jun 19, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> > Why is the fact that only the root user can load a kernel module not a
>> > further restriction?

>> Because the user (under whose control the computer is, be it person or
>> company) set up the root password herself?

> Well, duh. TiVo, under whose control the software running on my Tivo is, set
> up the signing key themself. *Someone* has to decide what software runs, the
> GPL cannot rationally decide who that is because it's an
> application-specific authorization decision.

Right.  All GPL can say is that you cannot impose further restrictions
on how the user adapts the software, and since the user runs the
software on that computer, that means you must not restrict the user's
ability to upgrade or otherwise replace that software there, when you
gave the user the software along with the computer.


>> > However, "you can't load your modified sofware on *MY* hardware" is
>> > not a further restriction.

>> As long as you didn't hand me the hardware along with the software,
>> for me to become a user of the software on that hardware, I agree.

> This is, again, an argument that is totally alien to the GPL.

No, it's not.  It is intended to ensure that free software remains
free for all its users.  When you receive the software, you become a
user.  That's when you receive the rights, and that's what creates the
obligation on the distributor to not impose restrictions on the
freedoms, no matter by how means such restrictions could be legally or
technically accomplished.

> The idea that you have 'special' rights to the software on some
> hardware but not others is simply insane.

I agree, to some extent.  It's not so much about the rights, but about
the restrictions the vendor can impose on hardware.

It's just that, for this particular hardware, as you say, the
manufacturer has (or had) special rights.  This means it can decide
what software runs, whom it gives the hardware to, etc.

However, by distributing software under the GPL, the vendor accepts
the condition to not use any means whatsoever to impose restrictions
on the recipient's exercise of the rights granted by the license by
means of the distribution of the software.

There's no reason to make the hardware special, or the right of
authorization special, as a possible excuse to impose restrictions on
the user.  It amounts to just that: an attempt to excuse oneself from
the condition of not imposing restrictions on the enjoyment of the
freedoms.

> The GPL is about being able to use the software on *ANY* hardware
> for which you have the right to decide what software runs.

Yes.  And, per the "pass on all rights you have" spirit in the
preamble, that translates into "no further restrictions" in the legal
terms, the user *must* receive this right from the distributor of the
software.


Oh, but what if the distributor doens't have this right in the first
place?  Well, let's see...

Either the distributor received the hardware with the software inside
it, which means it should have received this right along with the
software from whoever gave it the software, so it has this right, or
it installed the software itself, which means it does have this
right.  In both cases.

>> No, because the user is not becoming a user of the software on their
>> own computers.  Only in the computer that was shipped along with the
>> software.

> Yes, they are becoming a user. They might very well be using those
> computers.

This means they already were users there of those computers, just not
necessarily of that software on those computers.

And then, if they choose to copy and run the software on other
computers where they are entitled to install software, they're free to
do so, the vendor of that other piece of hardware must not impose
restrictions on that either.

> You can state what the GPLv3 does as many times as you want, but special
> rights to particular pieces of hardware is *TOTALLY* alien to the spirit of
> the GPL.

I agree.  That's the bug in GPLv2 that the anti-tivoization provision
is trying to fix.

> The GPL was always about equal rights to use the software in any
> hardware.

Exactly.  Thank you.  It finally sank in, it seems.

-- 
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 Tue, 19 Jun 2007 20:21:53 +0200, Nicolas Mailhot <nicolas.mailhot@laposte.net> escribió:

> traffic regulations only consider cars? I think not. Yet the same
> argument is the core of most GPL v3 objections we've seen in this
> thread.

No, the core argument of the GPLv3 objections is that you can NOT tell the
hardware manufacturers how to build hardware. You only  can tell software
users what hardware (tivoized) it's forbidden for them. Wether or not the
hardware manufacturers are going to care enought about your anti-tivo
software to remove their tivo protections is a completely different question
unrelated to the software license. Which is why the GPLv3 anti-tivoization
measures are stupid and pointless. 

Please, stop pretending you are hardware manufacturers. You are not.

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

[david]

On Tue, 19 Jun 2007, Daniel Hazelton wrote:

>> Dispute this:
>>
>> non-tivoized hardware => users can scratch their itches => more
>> contributions from these users
>>
>> tivoized hardware => users can't scratch their itches => fewer
>> contributions from these users
>
> Linus doesn't have to. Statistically the number of people that will even think
> of modifying the code running on a "tivoized" device is minute - at most 5%
> of the users of such a device. Of those people the ones with the skill to
> actually do the work is an even smaller number - figure 2.5 to 3% of them. Of
> those  with the skill, probably about 10% of them are actually *good* enough
> at it for their changes to be useful. Of that number, figure that only 25%,
> at most, will contribute the changes back.
>
> Apply that to a sample case:
> "tivoized" device total users: 1,000,000
> people that think about modifying: 50,000 (5%)
> people with skill: 1500 (3%)
> people who are good enough for the changes to be useful: 150 (10%)
> those who will contribute them back: 38 (25%)

based on my experiance looking at the software released for tivos, I think 
you are over-estimating these numbers. if there are more then a dozen 
people producing things that are good enough to be useful and releasing 
their results as opensource software I would be surprised.

and for all that the FSF is claiming that tivos can't being modified it's 
really not that hard to change.

David Lang

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

[Alexandre Oliva]

On Jun 19, 2007, Anders Larsen <al@alarsen.net> wrote:

> On 2007-06-18 21:50:12, Alexandre Oliva wrote:
>> Given the ROM exception in GPLv3, I guess you could seal and
>> anti-tamper it as much as you want, and leave the ROM at such a place
>> in which it's easily replaceable but with signature checking and all
>> such that the user doesn't install ROM that is not authorized by you.

> The manufacturer must be able to _remotely_ update the device
> firmware, so as I see it (IANAL), Tivoisation _is_ a requirement.

If you take the Wikipedia definition of Tivoization, you'll see it's
about copyleft software only, and no law mandates the use of copyleft
software.  There's no end to bad laws, but a law that mandated the use
of copyleft (=> free) software and at the same time prohibited
modifications by the user would be a very contradictory one.

-- 
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

[Nicolas Mailhot]

[-- Attachment #1: Type: text/plain, Size: 1224 bytes --]

Le mardi 19 juin 2007 à 10:50 -0700, David Schwartz a écrit :
> > > The GPL was never about allowing you to load modified software
> > > onto hardware
> > > where the legitimate creators/owners of that hardware say, "no,
> > > you may not
> > > modify the software running on this hardware".
> 
> > Good try but you had to add creators there so the sentence actually
> > supported your opinion. It's still an obvious alien insert.
> 
> It's simply shorter than saying "owners of the right or ability to decide
> what software runs on that hardware".

Right is not the same thing as ability. You have a technical ability
which has been converted in a "right" which in turn is used as argument
to reject GPLv3.

But did the original conversion happened with the approval of everyone
having rights to the result? I think not.

All the "GPLv2 didn't think of DRM therefore DRM is GPLv2-protected"
arguments make me sick. If tomorrow Ford starts mass+producing flying
saucers will they be exempt from traffic regulations because current
traffic regulations only consider cars? I think not. Yet the same
argument is the core of most GPL v3 objections we've seen in this
thread.

-- 
Nicolas Mailhot

[-- Attachment #2: Ceci est une partie de message numériquement signée --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

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

[Alexandre Oliva]

On Jun 19, 2007, Johannes Stezenbach <js@linuxtv.org> wrote:

> (Where "free loaders" is a term introduced by Alexandre, not by me.)

It's actually from game theory.  Or something sufficiently mangled by
translation back and forth between English and Portuguese.  I think
the original is actually free riders.  My bad.

> The GPLv2 is a sufficient tool to defend free software
> against those that don't even grasp tit-for-tat.

It's not.  It doesn't even demand tit-for-tat.  This is a
*consequence* of the player understanding the spirit.  You can't
convince a person who doesn't believe in these ideas by pointing at
the license and saying "see, look, you're going to get contributions
back", because there are no such guarantees in the license.  And
that's how it should be.

> But it has to be their decision, IMHO it's wrong to force them.

Agreed.  Nobody is forcing anyone to use GPLed software.  Nobody is
forcing anyone to accept existing software under the GPLv3.

What every GPL stands to do is to defend the freedoms of users,
respecting the wishes of the authors expressed through the GPL that
the free software mains free.  Of course, authors who use the GPL for
other purposes may differ, but whether or not their claims that the
GPL advances their stances bear any resemblance with reality is
irrelevant.  This doesn't change what the GPL stands to do in any way,
it only changes what they expect the GPL to accomplish.  And whether
they're right or wrong in their expectations is besides the point.

> The GPLv3 tries to be a tool to defend against those that
> don't subscribe to the full Free Software Definition.

Not quite.  There appears to be an occurrence of a very common mistake
in your message.  Please forgive the long digression to try to dispell
it.

Many people think that the GPL is what the Free Software Definition is
all about, that Free Software somehow implies GPL, or some other
misunderstandings (to be read without a condescending tone ;-)

Free Software, and in particular the Free Software Definition, talks
about *respecting* users' freedoms.  There are many Free Software
licenses that accomplish this.

Some are very liberal in this sense.  They let you do whatever you
want with the code.  Even use it to disrespect users' freedoms.

Others are not so liberal.  They go *beyond* the Free Software
Definition, i.e., beyond merely respecting users' freedoms.  They
require agreement from recipients to not disrespect some users'
freedoms with the software, in some specific ways.  In addition to
*respecting* some freedoms, they *defend* some freedoms.

Others take the stance of defending all the four users' freedoms,
requiring agreement to not disrespect any users' freedoms with the
software, and to pass on this requirement, such that that software is
never used to disrespect users' freedoms.  These are called copyleft
licenses.

GPL is just one among all copyleft licenses, that are just some among
all licenses that defend some users' freedoms, that are just some
among all licenses that abide by the Free Software definition.  It
just so happens that it's the most widely-used Free Software license.


Think of it this way (if you understand open source but don't know
much of its history): the Open Source Definition is a rewritten
version of the Free Software Definition, intended to retain the same
meaning, but with a different focus.  That's why, as a general rule,
Free Software licenses are Open Source licenses, and Open Source
licenses are Free Software licenses.  AFAIK there is only one known
exception, and that's the open-source license Reciprocal Public
License, which goes to show that the Open Source definition is not
equivalent to the Free Software definition.

The differences are held to be in the motivations behind each of the
movements.  While Free Software takes the respect for the freedoms as
a moral issue (it's the right thing to do), Open Source takes it as a
pragmatic issue (it's better for everyone).  As luck would have it,
the pragmatic benefits are a consequence of the respect for users'
freedoms.

And more, pragmatists who see value in ensuring that the software
remains open source (which any OSI board member will insist that is
far more than merely keeping the source code open) are perceiving a
consequence of ensuring that the software remains Free, of defending
the freedoms of all users of the software.


Sure, there are other pragmatists that don't even care about the
software remaining open source, but only about the source code
remaining available.  We still have a lot in common with them, and we
can happily work together in projects under a number of licenses in
which our goals overlap.

It's not like everyone needs to move to GPLv3.  It's that there would
be moral and practical benefits for everyone if everyone did.  But if
those who don't want to don't, nothing is really lost.  It's just that
both Free Software and Open Source Software advocates who care not
only about abiding by their definitions, but also about defending and
advancing their goals (i.e., not willing to see their software being
used against their goals, and wanting to see more software like that),
don't get an advance in these defenses without the relicensing.

That's unfortunate, but it's not the end of the world.  Nobody can,
should or will force any copyright holder to adopt the GPLv3.  Any
claims to the contrary are emotional reactions to peer pressure, which
very clearly exists, since there are indeed numerous people who want
to advance their goals and would like to use the GPLv3 as a tool.

-- 
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 was never about allowing you to load modified software
> > onto hardware
> > where the legitimate creators/owners of that hardware say, "no,
> > you may not
> > modify the software running on this hardware".

> Good try but you had to add creators there so the sentence actually
> supported your opinion. It's still an obvious alien insert.

It's simply shorter than saying "owners of the right or ability to decide
what software runs on that hardware".

The right to control the hardware vests originally with its owner/creator.
Tivo does not in fact transfer that right to the purchasers of their
hardware. (You can agree or disagree with this, but I don't think you can
coherently deny that they don't.)

The point is that GPL rights always applied equally to all hardware. Special
rights to run software on particular pieces of hardware are completely alien
to the original spirit of the GPL.

Someone has to decide what software runs on what hardware, and who makes
that decision can differ from the owner of the hardware for any number of
legitimate reasons. (And, IMO, for illegitimate reasons in the case of Tivo,
but that's not the issue here.)

DS

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

[David Schwartz]

> On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

> > Why is the fact that only the root user can load a kernel module not a
> > further restriction?

> Because the user (under whose control the computer is, be it person or
> company) set up the root password herself?

Well, duh. TiVo, under whose control the software running on my Tivo is, set
up the signing key themself. *Someone* has to decide what software runs, the
GPL cannot rationally decide who that is because it's an
application-specific authorization decision.

> >> > The GPL was never, until GPLv3, about who gets to make
> >> > authorization decisions.
>
> >> I can agree with that.  As long as the authorization decisions are not
> >> used as means to deprive users' of the freedoms that must not be
> >> restricted, they can be whatever the distributor fancies.

> > Right, which is the freedom to modify the software. The freedom
> > to get the
> > source code. The freedom to use the source code however you want, absent
> > legitimate authorization decisions to the contrary.

> What makes them lawful, given the "no further restrictions"?

That the person who decides what software runs on that hardware can remove
them if they please. *Someone* has to decide what software runs on a
particular piece of hardware, right?

> > However, "you can't load your modified sofware on *MY* hardware" is
> > not a further restriction.

> As long as you didn't hand me the hardware along with the software,
> for me to become a user of the software on that hardware, I agree.

This is, again, an argument that is totally alien to the GPL. The idea that
you have 'special' rights to the software on some hardware but not others is
simply insane. It is totally out of left field with respect to the GPL. The
GPL is about being able to use the software on *ANY* hardware for which you
have the right to decide what software runs.

> However, since they distribute GPLed software along with the hardware,
> such that I'd be a user of the software on that hardware, they should
> not impose further restrictions on my freedoms that the GPL stands to
> defend WRT the GPLed software.  So, they must not use their
> authorization right to deny me, the user of the software, in the
> hardware that they meant me to use the software, the freedom to adapt
> the software for my own needs, and run it for any purpose.

You can argue this, but it's not a GPL argument. It's a reasonable argument,
but it has nothing whatsoever to do with GPL rights. GPL rights are about
being able to use the software on *any* hardware you want, not special
rights to use the software on some one particular piece of hardware. GPL
rights are rights against obstacles to getting the source code and legally
modifying it and distributing it, not rights against authorization obstacles
placed by people who own hardware.

> > That would mean it doesn't permit the distribute to state "BTW,
> > you can't
> > install, modify or run this software on *OUR* computers that run our
> > corporate network".

> No, because the user is not becoming a user of the software on their
> own computers.  Only in the computer that was shipped along with the
> software.

Yes, they are becoming a user. They might very well be using those
computers. It's absolutely absurd to argue that the right to choose what
software runs on a piece of hardware must go to the user of that hardware.
In any event, it's totally alien to the GPL which is not at all about who
gets to decide what software runs on what hardware.

> >> The "no further restrictions" applies equally to all computers.
> >> It's not just because you have some control over some particular
> >> hardware that you deliver along with the software that you're
> >> entitled to use that to limit the user's freedoms.

> > I agree. However, that doesn't mean that people who own or control
> > particular pieces of hardware can't put authorization barriers that
> > prevent you from running whatever software you want on thos pieces
> > of hardware.

> That's correct, as long as they didn't give me that hardware with
> GPLed software in it.  The moment they do, I become recipient and user
> of GPLed software in that computer, and they should relinquish their
> power to impose restrictions on my exercise of the freedoms WRT that
> software.  And there's no reason whatsoever to exclude restrictions
> such as those implemented by means of authorization.

You can state what the GPLv3 does as many times as you want, but special
rights to particular pieces of hardware is *TOTALLY* alien to the spirit of
the GPL. The GPL was always about equal rights to use the software in any
hardware.

DS

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

[Alexandre Oliva]

On Jun 19, 2007, "Pekka Enberg" <penberg@cs.helsinki.fi> wrote:

> Hi Alexandre,
> On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
>> Dispute this:
>> 
>> non-tivoized hardware => users can scratch their itches => more
>> contributions from these users
>> 
>> tivoized hardware => users can't scratch their itches => fewer
>> contributions from these users

> Maybe, but in what numbers?

We'll get to that.

> Furthermore, Linus has repeatedly explained to you why he (and bunch
> of other _kernel_ hackers -- this is a thread on LKML, remember)
> thinks it's stupid for a software license to restrict hardware
> design choices.

Most of these explanations carried the assumption that this would
lower the amount of contributions he'd get.  And this is precisely the
circular logic in their arguments, and this is why this argument is
intended to show the initial assumption is false.  But hey, maybe I'm
wrong, and it won't show that at all, and I'll be proved wrong, and
then there's going to be one more very vocal defensor of your ideals.


Now, if people want to keep on fooling themselves, I guess they can,
as long as they don't hurt others in the process.  If they try to fool
others, I feel it is my moral duty to intervene.

-- 
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 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Tuesday 19 June 2007 04:04:52 Alexandre Oliva wrote:
>> So your claim is that a user's possibility to scratch her own itches
>> makes no difference whatsoever as to their amount of contributions she
>> is likely to make?

> Exactly.

Hmm, interesting...  Are you doing this just to make it more
interesting, or did you never take Open Source 101? ;-)  Or economy,
game theory, ecology, politics, for that matter?

Do you realize how many contradictions arise from this claim?

A few examples for you.  Let me know if you don't understand why these
contradict this claim, I'll be happy to explain it to you.

How many newbies have patches accepted that they didn't test?

At which point in becoming a Linux developer does one leave behind
this altruistic attitude and becomes moved by self interest only?

How many people you know got to know Linux in their TiVos, took the
Linux sources that TiVo distributes and built it for their own PC?

>> Yes.  And your estimates are way too low too, FWIW.  Any reason why
>> you changed your mind as to the 10% before?

> That 10% was, IIRC, a reference to the potential number of "Hackers" that 
> would own a TiVO. On thinking about it I realized that the number of hackers 
> owning a TiVO would be vanishingly small because of "tivoization".

Aah, ok.

So, you lowered the estimate for the case of a tivoized device, but
then claimed it was the same for an otherwise-identical non-tivoized
device.

Seriously, try 'modprobe logic', you'll like it ;-)

> Wrong. Nobody here needs a "piece by piece" explanation.

Then why do you keep making claims that are not related with either
the part of the argument I'm posing or the full argument I've already
presented?

> So, in the belief that you were intelligent enough to understand
> that, I was providing proof that refutes your argument entirely.

The only proof you provided was that you didn't understand the
argument.

> With a situation as complex as what exists you can't split the
> argument into two and claim that, since "Argument A" is true in the
> "split" argument that it is true when the argument isn't split. This
> holds true for almost all real-world situations.

Which proves you don't understand how logical proofs work.

Let me explain it to you.

First, you need to establish initial premises.  Whether or not they
resemble any similarity with anything else you're thinking of is
irrelevant.  They might even be known to be false, in proof by
contradiction.

Then, you apply logical inference rules to the initial premises, and
establish consequences of the initial premises.

One of these consequences may be what you are trying to prove.  Or,
you may come to a contradiction, and prove that initial premises are
self-contradictory.  Or you may come to no useful conclusion
whatsoever.

> Now, I am not enjoying the discussion anymore.

Understandable.

> I've asked once before - remove me from the CC list.

I'll try to remember to 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

[Alexandre Oliva]

On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Tuesday 19 June 2007 04:04:52 Alexandre Oliva wrote:
>> On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
>> > On Tuesday 19 June 2007 02:44:32 Alexandre Oliva wrote:
>> >> GPLv3 forbids tivoization, therefore developer has requirement for
>> >> tivoization in the license, therefore GPLv3 forbidding tivoization
>> >> is bad.
>> >
>> > However, my argument is straight logic, nothing "circular" about it.  :)
>> > Replacing "X" in my logic path above with "tivoization" and "license"
>> > with "GPLv3", as you've done, does produce a valid chain of logic.
>> 
>> Yes.  Isn't it funny though that tivoization became necessary as a
>> consequence of GPLv3 forbidding it?

> -ELOGIC

I see.  Try 'modprobe logic', it worked for me years ago ;-) :-D

> It didn't become necessary as a result of the GPLv3 forbidding it.

Which is why I said it was funny, because your inference chain stated
*exactly* (with an implied "for the developers") that it did.

Do you understand what an inference chain is?  A => B, as in A implies
B, which can also be read as A therefore B if A is known to hold.

> there could be any number of reasons why "tivoization" is needed by
> the manufacturer.

This claim is false.

Tivoization is when hardware manufacturer takes copyleft software and
blocks updates by the user of the hardware.

No single law so far has shown an example that even resembled to
mandate copyleft software, and no contract could possibly establish a
condition like this.

Therefore, this claim is false.

> This whole bit was to point out that you were inferring circular
> logic where none existed.

There *is* circular logic is in place.

The initial premise of this fallacy is that anti-tivoization is bad
for the project.

This is used to conclude that licenses with such provisions should be
rejected.

This is then used to conclude that there are fewer developers who
would develop under such licenses.

Which is then used to conclude that anti-tivozation is bad for the
project.

-- 
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

[Alan Cox]

> Well, it is not Tivo alone -- look at http://aminocom.com/ for an
> example. If you want the kernel sources pay USD 50k and we will provide
> the kernel sources, was their attitude.

GPLv2 deals with that case, and they can (and should) be sued for it
[except that US copyright law is designed for large music companies not
people]

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

[Nicolas Mailhot]

[-- Attachment #1: Type: text/plain, Size: 367 bytes --]


> The GPL was never about allowing you to load modified software onto hardware
> where the legitimate creators/owners of that hardware say, "no, you may not
> modify the software running on this hardware".

Good try but you had to add creators there so the sentence actually
supported your opinion. It's still an obvious alien insert.

-- 
Nicolas Mailhot

[-- Attachment #2: Ceci est une partie de message numériquement signée --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

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

[Manu Abraham]

Lennart Sorensen wrote:

> 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

Well, it is not Tivo alone -- look at http://aminocom.com/ for an
example. If you want the kernel sources pay USD 50k and we will provide
the kernel sources, was their attitude.

> 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.

Well, it is not Tivo alone, a large chunk of the vendors do that. The
vendors who actually do it the clean way are just few and can be counted
very easily.

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

[Pekka Enberg]

Hi Alexandre,

On 6/19/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> Dispute this:
>
> non-tivoized hardware => users can scratch their itches => more
> contributions from these users
>
> tivoized hardware => users can't scratch their itches => fewer
> contributions from these users

Maybe, but in what numbers? It's not like every user out there is able
to scratch their itches in the first place. Besides, people who want
to hack their hardware either work around the restrictions (like Linux
on xbox) or buy hardware that doesn't have any.

In any case, I hope we call quits on this thread. People have already
explained to you that there are certain devices (ATMs, medical
equipment) where you absolutely want "tivoization". Furthermore, Linus
has repeatedly explained to you why he (and bunch of other _kernel_
hackers -- this is a thread on LKML, remember) thinks it's stupid for
a software license to restrict hardware design choices. If you don't
see the point, fine, but lets stop wasting each others time, ok?

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

[Manu Abraham]

Johannes Stezenbach wrote:
> On Tue, Jun 19, 2007, Manu Abraham wrote:
>> Johannes Stezenbach wrote:
>>
>>> I argue that if you keep the free loaders out, you miss
>>> the chance to communicate with and educate them.
>>> Communication across borders doesn't work well, and you create
>>> a border between the morally "good" and the "bad".
>>>
>>> Of course you can't expect that every free loader will
>>> learn and accept the free software philosopy, some just
>>> won't. But to me that's acceptable, and the GPLv2, or indeed
>>> Linus' tit-for-tat interpretation of the GPLv2, is IMHO
>>> sufficient to protect my interests.
>> Err .. when you say protection on one hand and on the other you state
>> it's hard to keep free loaders away, 
> 
> I didn't say that.
> 
> IMHO it isn't even useful to try to keep free loaders away,
> it's better to try and integrate them gradually. That's part
> of the game.
> (Where "free loaders" is a term introduced by Alexandre, not by me.)
> 
> The GPLv2 is a sufficient tool to defend free software
> against those that don't even grasp tit-for-tat. But if
> they do, you can talk to them *as peers* and try to convince
> them that there's more to free software than just tit-for-tat.

What i _feel_ is that "some" (vendors) think that even if they utilize
existing resources what is open and keep what they have done, completely
closed (ie, they use existing infrastructure as a building block, but
they keep the stuff completely closed, in many cases the argument with
regards to IP is not even valid, as looking at symbol tables etc we find
some badly copied code from other parts etc, sewn together. The logical
thought what i have is that they do this to avoid a competition in the
short run) -- and they feel that they are doing things in a quite legal way.

> But it has to be their decision, IMHO it's wrong to force them.

Trying to force _anything_ on _anyone_ doesn't achieve anything, other
than just anger.

We at the worst could of course argue that users should avoid going for
that specific product, but i don't know how much that would work. I say
thus, because legally it would not be possible to challenge such vendors
globally as rules and regulations are different with different
governments and or countries. In such a case a tit-for-tat doesn't work
at all.

If all the people were to agree on common aspects, there wouldn't be any
wars at all ?

> The GPLv3 tries to be a tool to defend against those that
> don't subscribe to the full Free Software Definition.

If v3 defends the definition better, that would be a better use case,
don't you think so ? (But i don't see how v3 also will defend the
definition across international territories, in such a case what i
outlined, since have been bitten by this many times, even talking to the
black sheep which never helped)

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

[Johannes Stezenbach]

On Tue, Jun 19, 2007, Manu Abraham wrote:
> Johannes Stezenbach wrote:
> 
> > I argue that if you keep the free loaders out, you miss
> > the chance to communicate with and educate them.
> > Communication across borders doesn't work well, and you create
> > a border between the morally "good" and the "bad".
> > 
> > Of course you can't expect that every free loader will
> > learn and accept the free software philosopy, some just
> > won't. But to me that's acceptable, and the GPLv2, or indeed
> > Linus' tit-for-tat interpretation of the GPLv2, is IMHO
> > sufficient to protect my interests.
> 
> Err .. when you say protection on one hand and on the other you state
> it's hard to keep free loaders away, 

I didn't say that.

IMHO it isn't even useful to try to keep free loaders away,
it's better to try and integrate them gradually. That's part
of the game.
(Where "free loaders" is a term introduced by Alexandre, not by me.)

The GPLv2 is a sufficient tool to defend free software
against those that don't even grasp tit-for-tat. But if
they do, you can talk to them *as peers* and try to convince
them that there's more to free software than just tit-for-tat.
But it has to be their decision, IMHO it's wrong to force them.

The GPLv3 tries to be a tool to defend against those that
don't subscribe to the full Free Software Definition.


Johannes

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

[Manu Abraham]

Johannes Stezenbach wrote:

> I argue that if you keep the free loaders out, you miss
> the chance to communicate with and educate them.
> Communication across borders doesn't work well, and you create
> a border between the morally "good" and the "bad".
> 
> Of course you can't expect that every free loader will
> learn and accept the free software philosopy, some just
> won't. But to me that's acceptable, and the GPLv2, or indeed
> Linus' tit-for-tat interpretation of the GPLv2, is IMHO
> sufficient to protect my interests.

Err .. when you say protection on one hand and on the other you state
it's hard to keep free loaders away, then don't you think that those 2
are two completely different things ?

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

[Johannes Stezenbach]

On Mon, Jun 18, 2007, Alexandre Oliva wrote:
> On Jun 18, 2007, Johannes Stezenbach <js@linuxtv.org> wrote:
> 
> > Hm, you only talk about people who already use free software,
> > but I tried to make you aware of the importance of
> > _promoting_ free software, i.e. winning new people and
> > companies for the free software idea.
> 
> Aah, I see.  Indeed, I'd missed that aspect.  Sorry about that.
> 
> My take on it is that bringing free loaders in doesn't help us much,
> and bringing them in in a way that they don't learn the essential
> aspects of the community will hurt the community in the long run.
> 
> So they must become aware that respecting others' freedoms is not only
> the right thing to do, from a moral and ethical standpoint, but also
> that this is precisely what enables our community to thrive, and to
> enable everyone to get the best out of the software we cooperate to
> develop.

The keywords here are "learn" and "become aware": It's a process
which takes time, and which requires ongoing communication.

I argue that if you keep the free loaders out, you miss
the chance to communicate with and educate them.
Communication across borders doesn't work well, and you create
a border between the morally "good" and the "bad".

Of course you can't expect that every free loader will
learn and accept the free software philosopy, some just
won't. But to me that's acceptable, and the GPLv2, or indeed
Linus' tit-for-tat interpretation of the GPLv2, is IMHO
sufficient to protect my interests.

> Of course we might get some additional contributions here and there,
> but then more and more users would still be stuck, unable or limited
> in the ways and incentives they have to participate in our community.
> Permitting this is very short-sighted.  It might bring us apparent
> advantages in the short run, but the more such disrespects there are,
> the more there will be, and the fewer users will be able to become
> developers.  In the end, this may kill the whole process, in a tragedy
> of the commons.  In the article linked below, I argue this very point,
> comparing how the demand for respecting users' freedoms is what keeps
> the free-loaders away and makes the GPL the most cost-effective
> license for software development, compared with permissive licenses
> and non-Free licenses.  The very same arguments apply to a comparison
> between a license that permits tivoization and one that doesn't,
> because the latter is more likely to have more contributors to share
> the load, and both equally reduce the likelihood of unmergeable forks.
> http://www.lsd.ic.unicamp.br/~oliva/papers/free-software/BMind.pdf

I'm not arguing about the GPL, especially not against the GPLv2.

What I'm concerned about is that the language you use trying to
promote the GPLv3 is IMHO anti-promotion of free software.

I believe executives don't read licenses. What they'll read is
the random article about "GPLv3 to outlaw tivoization",
"FSF wants to keep free loaders away" etc.

What is more likely, that they'll go to www.gnu.org to read
and absorb the GNU philosopy in order to become an accepted
member of the free software community, or that they'll decide to
stick with their proprietary RTOS then?


Johannes

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

[Anders Larsen]

On 2007-06-18 21:50:12, Alexandre Oliva wrote:
> Given the ROM exception in GPLv3, I guess you could seal and
> anti-tamper it as much as you want, and leave the ROM at such a place
> in which it's easily replaceable but with signature checking and all
> such that the user doesn't install ROM that is not authorized by you.

Sorry, I didn't state the regulations requirement clearly enough:

The manufacturer must be able to _remotely_ update the device
firmware, so as I see it (IANAL), Tivoisation _is_ a requirement.

Cheers
 Anders

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

[Daniel Hazelton]

On Tuesday 19 June 2007 04:04:52 Alexandre Oliva wrote:
> On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Tuesday 19 June 2007 02:44:32 Alexandre Oliva wrote:
> >> GPLv3 forbids tivoization, therefore developer has requirement for
> >> tivoization in the license, therefore GPLv3 forbidding tivoization
> >> is bad.
> >
> > However, my argument is straight logic, nothing "circular" about it.  :)
> > Replacing "X" in my logic path above with "tivoization" and "license"
> > with "GPLv3", as you've done, does produce a valid chain of logic.
>
> Yes.  Isn't it funny though that tivoization became necessary as a
> consequence of GPLv3 forbidding it?

-ELOGIC

It didn't become necessary as a result of the GPLv3 forbidding it. As I 
pointed out in text that was cut to keep the post short, there could be any 
number of reasons why "tivoization" is needed by the manufacturer. Other 
people have also pointed that out. This whole bit was to point out that you 
were inferring circular logic where none existed. 

<snip>
> >> Wait a minute, these figures you made up are for the tivoized hardware
> >> (no changes allowed to the GPLed software in it), or for the
> >> non-tivoized hardware (changes allowed to the GPLed software in it)?
> >
> > Actually, any generic "TiVO"-like hardware - whether it is tivoized or
> > not.
>
> So your claim is that a user's possibility to scratch her own itches
> makes no difference whatsoever as to their amount of contributions she
> is likely to make?

Exactly.

> Am I the only one who thinks this is utter nonsense?
>
> >> > those who will contribute them back: 38 (25%)
> >>
> >> Regardless of what you meant, this is 38 developers *on top* of
> >> however many the company pays to work on that, unless you're jumping
> >> the gun and spoiling the multi-part argument.
> >
> > 38ppm is a fairly small amount, regardless.
>
> Yes.  And your estimates are way too low too, FWIW.  Any reason why
> you changed your mind as to the 10% before?

That 10% was, IIRC, a reference to the potential number of "Hackers" that 
would own a TiVO. On thinking about it I realized that the number of hackers 
owning a TiVO would be vanishingly small because of "tivoization". So in this 
new set of numbers I dropped it entirely.

... crap I am tempted to respond to nastily has been cut ...

> > I think I'd rather see a guaranteed increase of developers - even if
> > it is only 10 - rather than hoping that the potential pool of 38
> > actually follows through. Wouldn't you?
>
> Yes.  How does this relate with the piece of the argument I've
> proposed so far, or the whole argument I've posted before?
>
> Answer: It doesn't.  At all.  You're just showing you didn't
> understand the argument.  Which shows why I have to explain it piece
> by piece.  Which suggests you shouldn't try to jump to conclusions.

Wrong. Nobody here needs a "piece by piece" explanation. So, in the belief 
that you were intelligent enough to understand that, I was providing proof 
that refutes your argument entirely.

With a situation as complex as what exists you can't split the argument into 
two and claim that, since "Argument A" is true in the "split" argument that 
it is true when the argument isn't split. This holds true for almost all 
real-world situations.

Now, I am not enjoying the discussion anymore. I've asked once before - remove 
me from the CC list.

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 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Tuesday 19 June 2007 02:44:32 Alexandre Oliva wrote:
>> GPLv3 forbids tivoization, therefore developer has requirement for
>> tivoization in the license, therefore GPLv3 forbidding tivoization
>> is bad.

> However, my argument is straight logic, nothing "circular" about it.  :)
> Replacing "X" in my logic path above with "tivoization" and "license" 
> with "GPLv3", as you've done, does produce a valid chain of logic.  

Yes.  Isn't it funny though that tivoization became necessary as a
consequence of GPLv3 forbidding it?

> FWIW the Linux Kernel shouldn't be as homogeneous a population as it
> is.

Nah.  Communities tend to form around similar values.  Linus started
the community.

>> Wait a minute, these figures you made up are for the tivoized hardware
>> (no changes allowed to the GPLed software in it), or for the
>> non-tivoized hardware (changes allowed to the GPLed software in it)?

> Actually, any generic "TiVO"-like hardware - whether it is tivoized or not. 

So your claim is that a user's possibility to scratch her own itches
makes no difference whatsoever as to their amount of contributions she
is likely to make?

Am I the only one who thinks this is utter nonsense?

>> > those who will contribute them back: 38 (25%)

>> Regardless of what you meant, this is 38 developers *on top* of
>> however many the company pays to work on that, unless you're jumping
>> the gun and spoiling the multi-part argument.

> 38ppm is a fairly small amount, regardless.

Yes.  And your estimates are way too low too, FWIW.  Any reason why
you changed your mind as to the 10% before?

>> > What you are arguing is that people should abandon

>> I'm not arguing any such thing.  Where's any such argument above?

>> At this point, I'm only comparing a tivoized device with a
>> non-tivoized device.  Nothing but it.

> You've been making the argument the entire time you've been arguing that 
> the "anti-tivoization" language in the GPLv3 is necessary.

And then I decided that, since the argument wasn't getting through, I
had to break it into pieces.

The piece I've presented so far has no abandonment whatsoever.  It's a
comparison between two different situations, to evaluate which of them
brings more contributions from users, regardless of the contributions
from the vendor, that are assumed to be the same, since there's no
material difference as far as the vendor is concerned (as in, vendor
has no reason to tivoize)

So your arguments bear zero relationship with the piece I've
proposed.  Can you see that?



> I think I'd rather see a guaranteed increase of developers - even if
> it is only 10 - rather than hoping that the potential pool of 38
> actually follows through. Wouldn't you?

Yes.  How does this relate with the piece of the argument I've
proposed so far, or the whole argument I've posted before?

Answer: It doesn't.  At all.  You're just showing you didn't
understand the argument.  Which shows why I have to explain it piece
by piece.  Which suggests you shouldn't try to jump to conclusions.



Once again, now with clearer starting conditions (not intended to
match TiVo in any way, BTW; don't get into that distraction)


Vendor doesn't care about tivoizing, their business works the same
either way.

Vendor's employees will contribute the same, one way or another, so
their contributions are out of the equation.

Users get source code in either case, and they can modify it and share
it.  They're in no way stopped from becoming part of the community.


Given these conditions:

In a tivoized device, users will be unable to scratch their itches.
This doesn't stop them from contributing to the project, but they may
lack self-interest motivation to contribute, because they won't be
able to use their modifications in the device they own.

In a non-tivoized device, users can scratch their itches.  They can
contribute just as much as they would in a tivoized device, but since
they can use the changes they make to make their own devices work
better for them, this works as a motivator for them to make changes,
and perhaps to contribute them.  Therefore, they will tend to
contribute more.


Can you point out any flaw in this reasoning, or can we admit it as
true?

-- 
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 Tuesday 19 June 2007 02:44:32 Alexandre Oliva wrote:
> On Jun 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Tuesday 19 June 2007 01:51:19 Alexandre Oliva wrote:
> >> On Jun 19, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> >> > The GPLv2 is the one that allows more developers.
> >> >
> >> > The GPLv2 is the one that is acceptable to more people.
> >>
> >> Based on my understanding that the anti-tivoization provisions are
> >> *the* objectionable issue about GPLv3 for those of you who dislike
> >> GPLv3, this is circular reasoning:
> >>
> >> anti-tivoization is bad
> >> => we reject licenses with it
> >> => there are fewer developers willing to develop with such licenses
> >> => anti-tivoization is bad
> >
> > The logic is close to:
> >
> >    => License forbids X
> >    => developer has requirement for X in license, can't add to project
> >    => License forbidding X is bad
>
> I'm not sure it was clear that '=>' was meant as logical implication.
> Read it as "therefore".
>
> It's actually funny that what your inference sequence (in spite of the
> missing initial operand) rings so true about my impressions about some
> of the reactions I'm getting here.
>
>   GPLv3 forbids tivoization, therefore developer has requirement for
>   tivoization in the license, therefore GPLv3 forbidding tivoization
>   is bad.
>
> :-)

However, my argument is straight logic, nothing "circular" about it.  :)
Replacing "X" in my logic path above with "tivoization" and "license" 
with "GPLv3", as you've done, does produce a valid chain of logic.  

> >> > I haven't really seen a single one. Last I did the statistic, I asked
> >> > the top ~25-30 kernel developers about their opinion. NOT A SINGLE ONE
> >> > preferred the GPLv3.
> >>
> >> Wow, that's a really big sample among all Free Software and Open
> >> Source developers out there.  And not even a little bit biased at
> >> that.
>
> Sorry that I missed the <irony> markers.
>
> > Yes, the sample could be considered "biased" - jst as a sample taken
> > among the GCC developers could be considered "biased" towards the
> > other end of the spectrum.
>
> FWIW, I haven't taken such a sample, because I know my network of
> contacts would likely make it statistically useless.  I'd not try to
> make an argument based on that.

FWIW the Linux Kernel shouldn't be as homogeneous a population as it is. I'd 
expect it with an FSF run project, because they require copyright assignment 
in order to participate, but with a project like Linux, where everyone 
maintains the copyright to their contributions, should be a hell of a lot 
less homogeneous than Linus' numbers make it seem.

<snip>
> > Statistically the number of people that will even think of modifying
> > the code running on a "tivoized" device is minute
>
> Wait a minute, these figures you made up are for the tivoized hardware
> (no changes allowed to the GPLed software in it), or for the
> non-tivoized hardware (changes allowed to the GPLed software in it)?

Actually, any generic "TiVO"-like hardware - whether it is tivoized or not. 
Admittedly the numbers are significantly different for PC's (and other types 
of general purpose computing devices).

> > those who will contribute them back: 38 (25%)
>
> Regardless of what you meant, this is 38 developers *on top* of
> however many the company pays to work on that, unless you're jumping
> the gun and spoiling the multi-part argument.

38ppm is a fairly small amount, regardless.

> > What you are arguing is that people should abandon
>
> I'm not arguing any such thing.  Where's any such argument above?
>
> At this point, I'm only comparing a tivoized device with a
> non-tivoized device.  Nothing but it.

You've been making the argument the entire time you've been arguing that 
the "anti-tivoization" language in the GPLv3 is necessary. I think I'd rather 
see a guaranteed increase of developers - even if it is only 10 - rather than 
hoping that the potential pool of 38 actually follows through. Wouldn't you?

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 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Tuesday 19 June 2007 01:51:19 Alexandre Oliva wrote:
>> On Jun 19, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
>> > The GPLv2 is the one that allows more developers.
>> >
>> > The GPLv2 is the one that is acceptable to more people.
>> 
>> Based on my understanding that the anti-tivoization provisions are
>> *the* objectionable issue about GPLv3 for those of you who dislike
>> GPLv3, this is circular reasoning:
>> 
>> anti-tivoization is bad
>> => we reject licenses with it
>> => there are fewer developers willing to develop with such licenses
>> => anti-tivoization is bad

> The logic is close to:

>    => License forbids X
>    => developer has requirement for X in license, can't add to project
>    => License forbidding X is bad

I'm not sure it was clear that '=>' was meant as logical implication.
Read it as "therefore".

It's actually funny that what your inference sequence (in spite of the
missing initial operand) rings so true about my impressions about some
of the reactions I'm getting here.

  GPLv3 forbids tivoization, therefore developer has requirement for
  tivoization in the license, therefore GPLv3 forbidding tivoization
  is bad.

:-)

>> > I haven't really seen a single one. Last I did the statistic, I asked the
>> > top ~25-30 kernel developers about their opinion. NOT A SINGLE ONE
>> > preferred the GPLv3.

>> Wow, that's a really big sample among all Free Software and Open
>> Source developers out there.  And not even a little bit biased at
>> that.

Sorry that I missed the <irony> markers.

> Yes, the sample could be considered "biased" - jst as a sample taken
> among the GCC developers could be considered "biased" towards the
> other end of the spectrum.

FWIW, I haven't taken such a sample, because I know my network of
contacts would likely make it statistically useless.  I'd not try to
make an argument based on that.

>> > So I have actual *numbers* on my side. What do you have, except for a
>> > history of not actually understanding my arguments?
>> 
>> Which is worse, not understanding or repeatedly snipping out and
>> addressing unrelated points?

> In other words - you've done the same and more.

I've honestly tried not to.  I believe Linus has, too.  Many of us
have talked past each other, a lot.

That was actually the point behind breaking up the argument in small
pieces.

If Linus hadn't got the whole argument, a number of times, before,
this might be described as dishonest, but since he did, and he can
refer back to those messages, he can know where I'm going.

>> Dispute this:

>> non-tivoized hardware => users can scratch their itches => more
>> contributions from these users

>> tivoized hardware => users can't scratch their itches => fewer
>> contributions from these users

> Linus doesn't have to.

Of course he doesn't.  But he will.  Because he's always right, and he
wants to show that.  That this is a bait and he knows it won't stop
him.  He knows there's no hook, because he knows where I'm going with
the argument.  But it's going to be interesting to watch.

> Statistically the number of people that will even think of modifying
> the code running on a "tivoized" device is minute

Wait a minute, these figures you made up are for the tivoized hardware
(no changes allowed to the GPLed software in it), or for the
non-tivoized hardware (changes allowed to the GPLed software in it)?

> those who will contribute them back: 38 (25%)

Regardless of what you meant, this is 38 developers *on top* of
however many the company pays to work on that, unless you're jumping
the gun and spoiling the multi-part argument.

> What you are arguing is that people should abandon

I'm not arguing any such thing.  Where's any such argument above?

At this point, I'm only comparing a tivoized device with a
non-tivoized device.  Nothing but 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

[Nicolas Mailhot]

[-- Attachment #1: Type: text/plain, Size: 852 bytes --]

> > But you're not the user of the software on my laptop.  I am.
> 
> ahh, but by your own argument you aren't
> 
> the software on your laptop is owned by people like Linus, Al Viro, David 
> M, Alan Cox, etc.

More accurately according to the GPL v2 theory Linus advanced the laptop
you paid is a shared property of the laptop manufacturer (who can
install whatever he likes without your input thanks to DRM) and the
kernel developper (who gets the source code).

You as end-user are only there to give this technical elite the money
they need to continue their activities (I'm forcing the point? Find me
one anti-GPL-v3 message in this thread that acknowledged the buyer/user
had any right)

This thanks to DRM which was not provided for when GPL v2 was written
and has been retroactively declared kosher.

-- 
Nicolas Mailhot

[-- Attachment #2: Ceci est une partie de message numériquement signée --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

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

[Alexandre Oliva]

On Jun 18, 2007, Al Viro <viro@ftp.linux.org.uk> wrote:

> I.e. the phrase about similar spirit should be replaced with
> something far more explicit and very, very hard to miss.

This is a very good idea.  Would you please file it at
http://gplv3.fsf.org/comments?

> I don't think you need more proof that people *do* interpret it in
> very different ways, with quite unpleasant results.
 
Agreed.

>> Is it correct to say that you share Linus' opinion, that the only
>> problem with the GPLv3 is the anti-tivoization provision?

> No.

Thanks for your detailed analysis.  I wish I knew what to do with
it.  (I decline impolite suggestions, thanks ;-)

Would you like me to put you in touch with Richard Fontana, one of the
lawyers involved in GPLv3, that's regarded as a legalese compiler, to
discuss your issues about wording with him?  Or would you rather file
them (with a bit more detail) at gplv3.fsf.org/comments?

Aside from wording issues, which appear to dominate your comments, is
it fair to characterize that your objections to GPLv3 are
anti-tivoization provisions (strong) and Affero compatibility (weak?)?


I'm setting wording issues aside because these are easier to fix once
the problem is understood, rather than ideological differences, that
would probably be pointless to attempt to fix.

> 	7	-	if I want to give additional permissions, I don't
> 			want them stripped, for fsck sake!  There is a
> 			bog-standard mechanism for _that_ (dual-licensing),
> 			thank you very much.

additional permissions are indeed a form of dual-licensing, but one
that doesn't require one to create a copy of the GPL and add the
additional permissions to that copy.  Yes, it could be accomplished
with dual-licensing terms such as "you can follow the terms of the
GPL, with the following additional permissions".

I don't quite see the point of criticizing this.  This is more
informative than anything else.

The meat here is really in the few additional restrictions, and the
provisions to combat the practice of adding restrictions on top of the
GPL and claiming the software is available under the GPL, which has
made for a lot of confusion over time.


Thanks a lot for your feedback.

-- 
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 Tuesday 19 June 2007 01:51:19 Alexandre Oliva wrote:
> On Jun 19, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > The GPLv2 is the one that allows more developers.
> >
> > The GPLv2 is the one that is acceptable to more people.
>
> Based on my understanding that the anti-tivoization provisions are
> *the* objectionable issue about GPLv3 for those of you who dislike
> GPLv3, this is circular reasoning:
>
>   anti-tivoization is bad
>   => we reject licenses with it
>   => there are fewer developers willing to develop with such licenses
>   => anti-tivoization is bad

The logic is close to:
   => License forbids X
   => developer has requirement for X in license, can't add to project
   => License forbidding X is bad

When it comes to TiVO the reason the developers "required" the "tivoization" 
was because the company itself demanded it. The reason: providers of the 
content their device works with demanded it.

> > Face it, the "open source" crowd is the *bigger* crowd.
>
> I really don't know about that.  I can believe it may be so in LKML.

Actually, it is. Every "Free Software" developer that I personally know could 
care less about the FSF's motives - until they impact them. Since they don't 
care what the FSF does, publishes, etc... they cannot be termed "Free 
Software" developers (using the definition of the term "Free Software" 
provided by the FSF). In fact, almost all of them will state either: "I work 
on Open Source software" or "I develop FOSS stuff".

> > I haven't really seen a single one. Last I did the statistic, I asked the
> > top ~25-30 kernel developers about their opinion. NOT A SINGLE ONE
> > preferred the GPLv3.
>
> Wow, that's a really big sample among all Free Software and Open
> Source developers out there.  And not even a little bit biased at
> that.

Yes, the sample could be considered "biased" - jst as a sample taken among the 
GCC developers could be considered "biased" towards the other end of the 
spectrum.

> > So I have actual *numbers* on my side. What do you have, except for a
> > history of not actually understanding my arguments?
>
> Which is worse, not understanding or repeatedly snipping out and
> addressing unrelated points?

It's time to quote a very ancient source: "Don't point out the speck in your 
neighbors eye when you cannot see the log in your own"

In other words - you've done the same and more.

>
> Let's please try again.
>
> I'll try to keep it simple, since you can't seem to be able to grasp
> the entire argument, and keep disregarding essential parts, disputing
> unrelated points and jumping to the conclusions that you've disputed
> the point I was trying to make.
>
> I'll present it in parts, as an attempt to stop you from making this
> mistake, that I'm sure is not intentional.
>
> The first part is in this e-mail.
>
>
> Dispute this:
>
> non-tivoized hardware => users can scratch their itches => more
> contributions from these users
>
> tivoized hardware => users can't scratch their itches => fewer
> contributions from these users

Linus doesn't have to. Statistically the number of people that will even think 
of modifying the code running on a "tivoized" device is minute - at most 5% 
of the users of such a device. Of those people the ones with the skill to 
actually do the work is an even smaller number - figure 2.5 to 3% of them. Of 
those  with the skill, probably about 10% of them are actually *good* enough 
at it for their changes to be useful. Of that number, figure that only 25%, 
at most, will contribute the changes back.

Apply that to a sample case:
"tivoized" device total users: 1,000,000
people that think about modifying: 50,000 (5%)
people with skill: 1500 (3%)
people who are good enough for the changes to be useful: 150 (10%)
those who will contribute them back: 38 (25%)

Now a normal companies software department is anywhere from 10 to 50 people. 
Large companies can employ more - IBM employs hundreds.

What you are arguing is that people should abandon a firm set of developers 
that is proven to be large for the potential at adding, at most, 38 
developers per million users of the device. If that number was more than 1000 
per device I'd agree. The numbers don't support your argument.

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 19, 2007, Daniel Drake <dsd@gentoo.org> wrote:

> I realise that the latest GPLv3 draft would not pose restrictions
> here, as such devices would not be classified as consumer
> products.

And even if they were, there's always ROM.

I don't know whether hardware seals that state "once you break this
seal, law prohibits the use of this device with human patients".

Then the restriction is not being imposed by the manufacturer, only by
law, and this does make lot of a difference as far as software freedom
is concerned.

But then, law might not find this to be enough.  Software patents are
not the only stupid law that harms Free 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 19, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> Actually you are in error here. You are saying "More home users == More 
> Developers" when the ratio of home users to developers isn't all that high. 
> (small set of facts: "Hacker" == "Developer" (in most cases, where the term, 
> as defined in the Jargon File, can actually be applied), "Home User" * 0.10 
> (ie: 10%) == "Developer" (approximately, and the correlation may be 
> lower). "TiVO" == "Developers" (note the plural - they do employ more than 
> one person for development))

As I wrote in another e-mail, why makes this proportion different for
the other conditions determined by the GPL?

I.e., what is it that makes this particular condition so allegedly
harmful for bringing in more developers and contributions, when
compared with the requirements on passing on source code, licensing
necessary patents, not suing other users over patent infringement in
the software, not invoking anti-circumvention laws, not entering
discriminatory agreements?

> So "TiVO", even though they are walking all over the freedoms you love, means 
> more *guaranteed* developers than the potential pool from the users of their 
> boxes. (the pool of potential developers among the millions of TiVO users is 
> actually miniscule, despite the size of the sample)

> However, you do make a good argument. But when you look at the statistics[1] 
> they don't hold water.

Err...  I have no idea of the actual user base of TiVo, but if it's
really in the millions, and your 10% figure above is right, this makes
for hundreds of thousands of hackers that could be scratching their
itches and improving Linux on TiVo boxes.

How many thousand employees does TiVo have working on Linux?

(I realize a full-time employee is a lot more than a Joe Random
Hacker, but still, I'm keeping a ratio of 100:1 to make up for that)

> PS: I've beaten the addiction!

Good for you!

-- 
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 19, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> The GPLv2 is the one that allows more developers. 

> The GPLv2 is the one that is acceptable to more people. 

Based on my understanding that the anti-tivoization provisions are
*the* objectionable issue about GPLv3 for those of you who dislike
GPLv3, this is circular reasoning:

  anti-tivoization is bad
  => we reject licenses with it
  => there are fewer developers willing to develop with such licenses
  => anti-tivoization is bad



> Face it, the "open source" crowd is the *bigger* crowd.

I really don't know about that.  I can believe it may be so in LKML.

> I haven't really seen a single one. Last I did the statistic, I asked the 
> top ~25-30 kernel developers about their opinion. NOT A SINGLE ONE 
> preferred the GPLv3.

Wow, that's a really big sample among all Free Software and Open
Source developers out there.  And not even a little bit biased at
that.

> So I have actual *numbers* on my side. What do you have, except for a 
> history of not actually understanding my arguments?

Which is worse, not understanding or repeatedly snipping out and
addressing unrelated points?


Let's please try again.

I'll try to keep it simple, since you can't seem to be able to grasp
the entire argument, and keep disregarding essential parts, disputing
unrelated points and jumping to the conclusions that you've disputed
the point I was trying to make.

I'll present it in parts, as an attempt to stop you from making this
mistake, that I'm sure is not intentional.

The first part is in this e-mail.


Dispute this:

non-tivoized hardware => users can scratch their itches => more
contributions from these users

tivoized hardware => users can't scratch their itches => fewer
contributions from these 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 18, 2007, Jan Harkes <jaharkes@cs.cmu.edu> wrote:

> Not really, Tivo could simply sell you a box without any installed
> software.

Yes.

> The actual software is mailed to you on a credit card sized
> ROM when you activate service.

If that's a separate transaction, then yes, I believe it would not be
convered under the terms of GPLv3, but IANAL.  There might be some
catch about intent, and also about contractual obligations in the
hardware sale to provide the software (coupons anyone? :-) or some
such.


But then, that you sell specialized devices with say only MIT-licensed
software shouldn't stop you from selling CD-ROMs with GPLed software,
even if those CD-ROMs could possibly run on that device.


The GPLv3 won't remove every way in which people who want/need to stop
the user from making changes to the software could accomplishing this
(ROM).  It will just make this a bit more inconvenient, such that
vendors that have the option respect users' freedoms, and those that
find it too inconvenient respect the wishes of users who don't want
their software turned 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 18, 2007, "Kevin Bowling" <lkml@kev009.com> wrote:

> Legitimate laws and practices require that certain devices not be
> modified by end users.  Therefore TiVo fails and contributions
> cease.

I've never denied this possibility.

But how about all the other devices that are being tivoized that do
NOT require this?

Are people just blind to this possibility?  Or does it really not
exist, and I'm the first who ever though of it?

>> Yes.  This is one option that doesn't bring any benefits to anyone.
>> It maintains the status quo for users and the community, but it loses
>> the ability for the vendor to upgrade, fix or otherwise control the
>> users.  Bad for the vendor.

> And users.  Don't spin the facts.

How is it good that the vendor can downgrade the software behind the
user's back?

Oh, yeah, right, it could upgrade it too.

> You are advocating things which hurt the end user,

Actually, no.  I'm advocating for respect for users' freedoms.
Whoever choose not to do that gets slightly hurt in the process, as an
incentive for respecting users' freedoms.  And yes, when the users'
freedoms are not respected (the cases you mentioned), that's bad for
the user, no doubt.  And bad for the community as well.

-- 
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 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

> Why is the fact that only the root user can load a kernel module not a
> further restriction?

Because the user (under whose control the computer is, be it person or
company) set up the root password herself?

>> > The GPL was never, until GPLv3, about who gets to make
>> > authorization decisions.

>> I can agree with that.  As long as the authorization decisions are not
>> used as means to deprive users' of the freedoms that must not be
>> restricted, they can be whatever the distributor fancies.

> Right, which is the freedom to modify the software. The freedom to get the
> source code. The freedom to use the source code however you want, absent
> legitimate authorization decisions to the contrary.

What makes them lawful, given the "no further restrictions"?

> However, "you can't load your modified sofware on *MY* hardware" is
> not a further restriction.

As long as you didn't hand me the hardware along with the software,
for me to become a user of the software on that hardware, I agree.

>> >> Tivoizers say "hey, you can still modify and run the software, just
>> >> not on *this* hardware".

>> Tivoization is treating the hardware that comes along with the
>> software as if it was different from others.  But it isn't.

> Of course it is. They have the authorization right on that hardware, and
> they don't have that right on my laptop.

Ok, I stand corrected.  They have that right.

However, since they distribute GPLed software along with the hardware,
such that I'd be a user of the software on that hardware, they should
not impose further restrictions on my freedoms that the GPL stands to
defend WRT the GPLed software.  So, they must not use their
authorization right to deny me, the user of the software, in the
hardware that they meant me to use the software, the freedom to adapt
the software for my own needs, and run it for any purpose.

> That would mean it doesn't permit the distribute to state "BTW, you can't
> install, modify or run this software on *OUR* computers that run our
> corporate network".

No, because the user is not becoming a user of the software on their
own computers.  Only in the computer that was shipped along with the
software.

> Don't you see how obviously absurd that is?

Yes, it would be, if it were so.

>> The "no further restrictions" applies equally to all computers.
>> It's not just because you have some control over some particular
>> hardware that you deliver along with the software that you're
>> entitled to use that to limit the user's freedoms.

> I agree. However, that doesn't mean that people who own or control
> particular pieces of hardware can't put authorization barriers that
> prevent you from running whatever software you want on thos pieces
> of hardware.

That's correct, as long as they didn't give me that hardware with
GPLed software in it.  The moment they do, I become recipient and user
of GPLed software in that computer, and they should relinquish their
power to impose restrictions on my exercise of the freedoms WRT that
software.  And there's no reason whatsoever to exclude restrictions
such as those implemented by means of authorization.

>> > Which is a massive departure from the previous GPL spirit which
>> > was about being able to use the software on *ANY* hardware you
>> > controlled, not some special pieces more than others.

>> It doesn't make the sold hardware special.  How come you think it
>> does?

> Because it becomes the only piece of hardware in the entire universe on
> which the GPL gives you the right to run the software. On every other piece
> of hardware, you must obtain that right from whoever owns the right to
> decide what software runs on that hardware.

I see.  Good point.  Agreed.  That hardware is indeed special.  Per
the GPL, it's the only one in which the distributor must NOT exercise
any restraints whatsoever on my exercise of the freedoms.

Which in turns makes it non-special, in that, from the point of both
the distributor and the user, it becomes just like any other random
piece of hardware: the distributor doesn't limit the freedoms the user
had, and the user isn't limited in enjoying the freedoms she had.

>> It's exactly the opposite.  It just says the distributor can't
>> make the hardware special, so as to restrain the users' freedoms that
>> are inseparable from the software.

> Don't you see that the rule that "this one thing cannot be special" makes
> that one thing special since everything else *can* be special.

It took me several attempts to understand what you meant.  Yes.  It's
special in that it can't be made special.  Poof, there goes the
universe ;-)

>> > Because that is not a right the vendor chooses to give to the user.

>> As in, the vendor can turn to the user and sue her for patent
>> infringement, after distributing GPLed software to her, just because
>> the use of the patent is not a right the vendor chooses to give to the
>> user?

> I don't know what patent you are talking about.

I'm talking about the implicit patent licenses that arise from
distributing software under GPLv2.

>> > You may dislike this decision, but it's not irrational.

>> I never said it was irrational.  I just said it's a further
>> restriction on the exercise of the freedoms that must accompany the
>> software wherever it goes.

> No more than having to be 'root' to load a kernel module. You are free to
> remove it from any hardware for which you have the right to choose what
> software runs.

Yup.  And I get that right (because the distributor must not stop me)
when I receive software under the GPL along with the computer in which
I'm expected to use 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

[Tim Post]

On Mon, 2007-06-18 at 23:21 -0400, Daniel Drake wrote:
> Let's take a certain class of medical devices into account: ones that 
> are absolutely definitely for medical treatment, but are not life 
> threatening if they fail.
> 
> Say, a dental treatment device -- if the device produces a crown or 
> bridge that doesn't fit properly, the dentist says "nope" and throws it 
> away. No harm done.

I've done quite a bit of research, I'm not nearly done.

These regulations (from what I can tell) seemed to follow suit with the
National Electric Code (NEC) [latest] when dealing with mandatory
isolated ground devices and special cabling methods when it comes into a
device touching a patient. If that remains consistent, this won't be so
bad.

If the patient never comes in contact with it, its not regulated as much
and (from what I've seen) has no requirement for tamper proofing. I
point out again, I am not _nearly_ done with my research.

I think of nothing else, anyone with an interest should closely monitor
how these devices are being regulated by the FDA as more of them begin
to look like penguins.

I won't argue one way or another as to the presence of benevolent intent
in those laws-to-come, I'm simply pointing out the questionable
technical competency of those who will be writing them and their need
for guidance when doing so.


Best,
--Tim

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

[Al Viro]

On Mon, Jun 18, 2007 at 08:46:44PM -0700, Linus Torvalds wrote:
 
> So I have actual *numbers* on my side. What do you have, except for a 
> history of not actually understanding my arguments?

Why do I suddenly have an image of Palin as Ximenez doing the answer?

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

[Linus Torvalds]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:
>
> > "More Developers" (either "Free Software" or "Open Source") == "More 
> > Contributions"
> 
> No, seriously.  Linus is disputing the equation above, dismissing my
> various attempts to show it to him, whenever it appears in teh context
> of tivoization, apparently because it doesn't match his moral belief
> that tivoization ought to be permitted on his moral grounds.

No. Linus is not AT ALL disputing the equation above.

But you are too f*cking stupid to admit that I *accepted* the

 - "More developers" == "More contributions" == good

equation, but I was claiming that your *other* part was totally broken.

You try to claim that the GPLv3 causes "More developers", and that, my 
idiotic penpal, is just crazy talk that you made up. 

But since you cannot follow a logical argument, and cannot make one up 
on your own, you instead make up some *other* argument, and try (like 
above) to try to say that I made that claim.

The GPLv2 is the one that allows more developers. 

The GPLv2 is the one that is acceptable to more people. 

Face it, the "open source" crowd is the *bigger* crowd. The FSF crowd is 
vocal and opinionated, but it's largely made up of people who _talk_ more 
than they actually _code_. 

Hot air doesn't make the world go round. Real code does.

Look at the kernel developers who claim that the GPLv2 is better. Not just 
me. Then look at the people who actually GET THINGS DONE. 

There's a big overlap there.

Now, look at the people who try to sell the GPLv3 as the best thing since 
sliced bread. How many of those are people who actually get things *done*?

I haven't really seen a single one. Last I did the statistic, I asked the 
top ~25-30 kernel developers about their opinion. NOT A SINGLE ONE 
preferred the GPLv3.

So I have actual *numbers* on my side. What do you have, except for a 
history of not actually understanding my arguments?

		Linus

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

[Daniel Drake]

Let's take a certain class of medical devices into account: ones that 
are absolutely definitely for medical treatment, but are not life 
threatening if they fail.

Say, a dental treatment device -- if the device produces a crown or 
bridge that doesn't fit properly, the dentist says "nope" and throws it 
away. No harm done.

Alexandre Oliva wrote:
> There may be business models that require the ability to make changes.

I'd say that its sensible for the manufacturer to attempt to retain this 
ability in every case. You never know what's going to go wrong, so it's 
a plus to have this option so that you can roll out some types of fixes 
without going bankrupt.

Now, for medical devices, this is tricky stuff: medical devices require 
all sorts of certifications, so modifying your product after you have 
certified it has it's complications. However, despite all the 
regulations it's realistic to be able to do this, and it does happen. 
Hell, windows-based devices in this field download new antivirus 
definitions and run windows update every few days.

> Then it's fair to enable the user to make changes as well, such that
> they don't become dependent on the vendor

Now this is where the regulations get really heavy. If the user is 
offered the ability to modify the device, theres *no way* it would get 
certified. Your business is dead - you do not have a product you can 
sell. In such case, the license has completely excluded free software 
from the market and everyone is forced to use completely closed systems.


I realise that the latest GPLv3 draft would not pose restrictions here, 
as such devices would not be classified as consumer products. That said, 
talking purely in terms of business models and fairness: there ARE 
decent reasons for manufacturer lockdown in some industries.

Daniel

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

[Daniel Hazelton]

On Monday 18 June 2007 22:06:57 Alexandre Oliva wrote:
> On Jun 18, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:
> > On Monday 18 June 2007 19:31:30 Alexandre Oliva wrote:
> >> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> >>
> >> Actually, just think of how many times you've heard the argument "I
> >> can't give you the source code for this driver/firmware/etc under the
> >> GPLv2 because the law says so."
> >
> > Sorry to tell you this, but anyone that makes a modification to GPLv2
> > covered code and distributes that modification is bound by the license.
>
> Of course I know that.  I'm not the one making those arguments.
> And then, not all of those pieces of code are indeed moficiations of
> GPLv2-covered code, so your objection is off target.

I had a parsing error with your statement. My mind made the jump to "they are 
doing it with GPLv2 code" - meaculpa.

> >> > b) I think you're simply wrong in your math. I think more people
> >> > like the middle-ground and not-frothing-at-the-mouth spirit of "open
> >> > source" over the religious dogma of "free software".
> >>
> >> It looks like the math you're talking about is in no way related with
> >> anything I've argued about.  You seem to be thinking about the number
> >> of people who claim to be on the "free software" or "open source"
> >> sides, but I can't fathom in what way this is related with whether you
> >> get more or less contributions from users as a consequence of users'
> >> being permitted to tinker with the free software in their own devices.
> >
> > "More Developers" (either "Free Software" or "Open Source") == "More
> > Contributions"
> >
> > That equation is very simple to understand - claiming its wrong is
> > impossible.
>
> YES!  Thank you!  This is exactly the point I'm trying to make.
>
> Now can you please explain this to Linus in terms that his brain won't
> dismiss as "coming from a fundamentalist"?

No need. Linus already understands the equation, and also the secondary fact 
that most home users are not developers. However, companies like TiVO do 
employ developers. This is why the equation works.

> > Apparently because you can't admit that a good reason *IS* a good reason
> > when it conflicts with your belief that the FSF is correct.
>
> No, seriously.  Linus is disputing the equation above, dismissing my
> various attempts to show it to him, whenever it appears in teh context
> of tivoization, apparently because it doesn't match his moral belief
> that tivoization ought to be permitted on his moral grounds.

Actually you are in error here. You are saying "More home users == More 
Developers" when the ratio of home users to developers isn't all that high. 
(small set of facts: "Hacker" == "Developer" (in most cases, where the term, 
as defined in the Jargon File, can actually be applied), "Home User" * 0.10 
(ie: 10%) == "Developer" (approximately, and the correlation may be 
lower). "TiVO" == "Developers" (note the plural - they do employ more than 
one person for development))

So "TiVO", even though they are walking all over the freedoms you love, means 
more *guaranteed* developers than the potential pool from the users of their 
boxes. (the pool of potential developers among the millions of TiVO users is 
actually miniscule, despite the size of the sample)

However, you do make a good argument. But when you look at the statistics[1] 
they don't hold water.

> > PS: I know I've said I'm done with this conversation, but this is like a
> > bad habit. I just couldn't help myself.
>
> You've helped me a lot while at that.  Thanks!
>
> I hope this helps others fundamentalist anti-fundamentalists :-) see
> reason too.

I love that phrase!

But seriously, all I did was stop trying to give fully reasoned counters 
(complete with examples) and state the simple truth.

DRH
[1] "There are three types of lies - lies, damn lies and statistics" - 
Attribution uncertain (Benjamin Disraeli, Mark Twain and Alfred Marshal are 
all said to have issued this famous quotation)

PS: I've beaten the addiction! This post was to clarify some things that were 
either misunderstood or stood a chance of being twisted to mean something 
other than what I intended. (not that anyone did the latter)

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

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

[Michael Poole]

David Schwartz writes:

>> David Schwartz writes:
>
>> >> First, end users buy and use the hardware in question.  It does not
>> >> belong to Tivo, so the analogy to his laptop fails there.
>
>> > No, this is incorrect. They buy *some* of the rights to the
>> > hardware but not
>> > all of them. Specifically, they do not buy the right to choose
>> > what software
>> > runs on that hardware. That right is still owned by TiVo.
>
>> Do you have a reference to the contract establishing that cession of
>> rights from the buyer to Tivo?
>
> No, and I submit that this is at least arguably something wrong that TiVo is
> doing. Note that Microsoft does this too when you buy an Xbox. It has
> nothing to do with the GPL.

There is a significant difference between what is a legally recognized
right and what no one has litigated over.  I tend to not recognize the
latter as the former until I see specific backing for the idea that
the purported right has been recognized by law, a court, or all
involved parties.

>> To the extent that some contract
>> purports to restrict the user in ways contrary to the GPL, I suspect
>> Tivo might have a hard time defending it in court.
>
> I agree, however, this doesn't restrict the user in ways contrary to the
> GPL. The GPL does not say that you have to be allowed to modify the Linux
> running on some particular piece of hardware because that is a legitimate
> authorization decision. TiVo not letting you change the software is the same
> as me not letting you change the software on my laptop.

I disagree that the two are the same -- for the fundamental reason
that you have not distributed the software on your laptop to me.  Tivo
has distributed the software on Tivo DVRs to their customers.  The act
of distribution is governed by (in the case of Linux) copyright law
and the GPL.

>> > You can argue that TiVo is being dishonest, breaking the law,
>> > being immoral,
>> > or whatever in retaining this right or in failing to disclose that they
>> > retain it. But you cannot coherently deny that TiVo retains
>> > this right when
>> > they sell certain other rights to the hardware.
>
>> By the first sale doctrine, someone who buys an item has practically
>> unlimited rights to deal with it or dispose of it as the buyer wishes.
>
> This is solely a right against copyright claims. You would be correct if
> TiVo were going to sue you for violating some copyright they hold in the
> hardware or software if you modified the software.

Do you propose that Tivo would (or could) sue a customer for some
non-copyright tort if the customer were to run a Linux kernel that has
not been authorized by Tivo on a Tivo-manufactured DVR?  As far as I
can tell, the legal concerns in question are all copyright issues.

>> The only things that would restrict that are statute or a contract
>> entered as part of the sale -- most likely a EULA or other shrink-wrap
>> agreement.  Given that most such recognized agreements deal with
>> software or services rather than hardware, I am not sure a court would
>> recognize a hardware EULA as being binding.  (I suspect this is the
>> direction you were heading with the paragraph below.)
>
> Yep, but that has nothing whatsoever to do with the GPL. The exact same
> argument applies with the Xbox. It's about whether authorization to modify a
> device should or must come with buying that device.
>
> The GPL was never about allowing you to load modified software onto hardware
> where the legitimate creators/owners of that hardware say, "no, you may not
> modify the software running on this hardware".

True.  The GPL always about allowing someone to modify software that
they received from someone else.  Tivo's Linux kernel images qualify
both as softare that they distribute to others and software that is
loaded onto hardware that they created.  The concern at hand is not
about hardware that Tivo owns or software that Tivo never distributes
-- except where it is also source code for software that they *do*
distribute.

Michael Poole

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

[Alexandre Oliva]

On Jun 18, 2007, david@lang.hm wrote:

> On Tue, 19 Jun 2007, Johannes Stezenbach wrote:
>>> But since the software is good, and moving to another software
>>> would be costly in various dimentions, the vendor has an incentive
>>> to stick with the software they have.

> but if regulations or other contracts require tamper-resistant
> hardware they have no choices other then to fork the existing GPLv2
> versions or switch to alternate options for anything that switches to
> GPLv3

Where by "alternate options" I hope you mean non-copyleft or
tivoizable (copyleft by definition) software, or (newer versions of)
the same software they already use, but in ROM.  (I point this out
because people keep forgetting all the available options when they
claim to enumerate all available options ;-)

-- 
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 18, 2007, Johannes Stezenbach <js@linuxtv.org> wrote:

> On Mon, Jun 18, 2007, Alexandre Oliva wrote:

>> People talk a lot about TiVo here, but do they the faintest idea of
>> how the conversations with TiVo are proceeding?  I thought so...

> Oh, if you know something we don't, could you please fill us in?

Honestly, I don't know either.  But I get an impression that there are
conversations underway.

> And who was it who coined the "Tivoization" term, thus putting
> TiVo into focus?

AFAIK TiVo invented the practice, did they not deserve the credit?

> Hm, you only talk about people who already use free software,
> but I tried to make you aware of the importance of
> _promoting_ free software, i.e. winning new people and
> companies for the free software idea.

Aah, I see.  Indeed, I'd missed that aspect.  Sorry about that.

My take on it is that bringing free loaders in doesn't help us much,
and bringing them in in a way that they don't learn the essential
aspects of the community will hurt the community in the long run.

So they must become aware that respecting others' freedoms is not only
the right thing to do, from a moral and ethical standpoint, but also
that this is precisely what enables our community to thrive, and to
enable everyone to get the best out of the software we cooperate to
develop.

> I think the majority of embedded devices still run proprietary
> RTOSes, and the majority of desktops still run Windows or Mac OS.
> Don't you want to change that?

Sure.  But getting those companies to adopt Free Software in a way
that turns it into non-Free Software doesn't change that in any way.

Of course we might get some additional contributions here and there,
but then more and more users would still be stuck, unable or limited
in the ways and incentives they have to participate in our community.
Permitting this is very short-sighted.  It might bring us apparent
advantages in the short run, but the more such disrespects there are,
the more there will be, and the fewer users will be able to become
developers.  In the end, this may kill the whole process, in a tragedy
of the commons.  In the article linked below, I argue this very point,
comparing how the demand for respecting users' freedoms is what keeps
the free-loaders away and makes the GPL the most cost-effective
license for software development, compared with permissive licenses
and non-Free licenses.  The very same arguments apply to a comparison
between a license that permits tivoization and one that doesn't,
because the latter is more likely to have more contributors to share
the load, and both equally reduce the likelihood of unmergeable forks.
http://www.lsd.ic.unicamp.br/~oliva/papers/free-software/BMind.pdf

> if you raise the entry barrier too high, they won't get started at
> all.

I acknowledge this argument, but I hear the same arguments against the
GPLv2, claiming the barrier is too high, and it's not from people who
believe that tivoization is already prohibited.

> They are aware of the trend towards Linux, but are afraid that the
> obligations of the GPL might be impractical for them.  Then they
> only have the choice to not use Linux, or to use "creative
> workarounds".

Or to respect users' freedoms, enabling/motivating those users to
become developers in our community.

> It's true that what these companies do might have little direct
> benefit for users buying their products, however the long term
> benefits of getting the people in these companies exposed to free
> software ideas, and in contact with the free software community, can
> only be positive

As long as they understand how the community works, be it from the
moral and ethical standpoint, be it from the pragmatic standpoint.  In
both cases, the end result is that they learn that, when they share
and cooperating, respecting users freedoms (enabling and providing
incentive for them to improve the software), everybody wins,
themselves included.

>> So you see, the picture of anti-tivozation is not as bleak as people
>> try to frame it.  In fact, it's not bleak at all.  If one out of 10,
>> maybe even 1 out of 100 vendors start respecting users' freedoms, when
>> faced with anti-tivoization provisions, the community will already win
>> big time, because each vendor is likely to have thousands of
>> customers, some of which will use the freedoms to serve the goals of
>> the community, in the very terms the community claims to care about.

> Does this multiplicator also apply to new companies
> which start using free software for their products?

Of course, even more so!  Then you win not only the contributions from
the user, but also from the company itself, which you didn't have
before.

-- 
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]

> This is a very limited reading of the GPL that leaves out one of its
> most important provisions: the bit about "no further restrictions".

Why is the fact that only the root user can load a kernel module not a
further restriction? Simple -- anyone who is bothered by that restriction
can remove it on any hardware for which they have the right to load modified
software. Anyone who does not have the right to load modified software on
some hardware simply does not have the right to change it.

Why is that not a "further restriction"? It means that I can't load kernel
modules on hardware that I don't have the right to load modified software
on.

> > The GPL was never, until GPLv3, about who gets to make
> > authorization decisions.

> I can agree with that.  As long as the authorization decisions are not
> used as means to deprive users' of the freedoms that must not be
> restricted, they can be whatever the distributor fancies.

Right, which is the freedom to modify the software. The freedom to get the
source code. The freedom to use the source code however you want, absent
legitimate authorization decisions to the contrary.

> > You are taking my claim out of contect. I am distinguishing
> > legal obstacles
> > from *authorization* obstacles, not technical obstacles.
>
> It doesn't matter how elaborate the excuse to disrespect the freedoms
> of the user is.  If there are further restrictions to them, then this
> violates the spirit, if not the letter, of the GPL.

I agree. However, "you can't load your modified sofware on *MY* hardware" is
not a further restriction. If it was, we get absurdities.

> >> Someone else's hardware is just a distraction.  You're not a user of
> >> software on someone else's hardware.  You have no rights over that.
>
> > You are. In the case of TiVo, the hardware (specifically the right
> > to decide what software runs on that hardware) is someone
> > else's. That is part of the bundle of rights that owning a piece of
> > hardware includes. That is a right you simply do not have with TiVo.

> Ah, ok, so I was sloppy above and you caught that.

> If someone else places hardware on your home for you to use, even if
> they still own it, then you can be a user of someone else's hardware.

Definitely.

> And at that point the GPL kicks in, because the software was
> distributed to you (even if the hardware wasn't sold), and with the
> distributed software come the freedoms, which, per the GPL, the
> distributor must not disrespect.

Absolutely.

> >> Tivoizers say "hey, you can still modify and run the software, just
> >> not on *this* hardware".

> Tivoization is treating the hardware that comes along with the
> software as if it was different from others.  But it isn't.

Of course it is. They have the authorization right on that hardware, and
they don't have that right on my laptop. For any piece of hardware, there
has to be someone who decides who can and can't choose what software runs on
that hardware.

> > Exactly. The GPL is about rights that apply to *all* hardware,
> > not some one
> > specific piece.

> Exactly!  Just like the GPL doesn't permit the distributor to state
> "BTW, you can't install or run this software on your mother's
> computer", it doesn't permit the distributor to state "BTW, you can't
> install or run this software on this computer I'm selling you".

That would mean it doesn't permit the distribute to state "BTW, you can't
install, modify or run this software on *OUR* computers that run our
corporate network". Don't you see how obviously absurd that is?

Someone has to be authorized to decide what software runs on some particular
piece of hardware. The GPL cannot mean that other people get to modify and
run software on that particular piece of hardware.

> The
> "no further restrictions" applies equally to all computers.  It's not
> just because you have some control over some particular hardware that
> you deliver along with the software that you're entitled to use that
> to limit the user's freedoms.

I agree. However, that doesn't mean that people who own or control
particular pieces of hardware can't put authorization barriers that prevent
you from running whatever software you want on thos pieces of hardware.

> > Which is a massive departure from the previous GPL spirit which
> > was about
> > being able to use the software on *ANY* hardware you
> > controlled, not some
> > special pieces more than others.

> It doesn't make the sold hardware special.  How come you think it
> does?

Because it becomes the only piece of hardware in the entire universe on
which the GPL gives you the right to run the software. On every other piece
of hardware, you must obtain that right from whoever owns the right to
decide what software runs on that hardware.

> It's exactly the opposite.  It just says the distributor can't
> make the hardware special, so as to restrain the users' freedoms that
> are inseparable from the software.

Don't you see that the rule that "this one thing cannot be special" makes
that one thing special since everything else *can* be special.

> > That's a nonsensical comparison. You can run it on any hardware
> > for which
> > you have the right to say what software runs.

> And why don't I have the right to say what software runs on the
> hardware I received along with the GPLed software?  Because the
> tivoizer doesn't want me to.  The tivoizer is placing barriers such
> that I cannot adapt the GPLed software included in that device to my
> own needs.  How is that not a further restriction to the four
> freedoms?  How is that not making that hardware special?

You can adapt it to your own needs, you just can't run it on hardware you
don't fully own. You do not fully own the TiVo because you do not own the
right to run modified software on it. It is just like *my* laptop -- you
don't own the right to choose what software runs on it. Someone has to have
that right, and in the case of TiVo, it's not you.

> > Because that is not a right the vendor chooses to give to the user.

> As in, the vendor can turn to the user and sue her for patent
> infringement, after distributing GPLed software to her, just because
> the use of the patent is not a right the vendor chooses to give to the
> user?

I don't know what patent you are talking about.

> > You may dislike this decision, but it's not irrational.

> I never said it was irrational.  I just said it's a further
> restriction on the exercise of the freedoms that must accompany the
> software wherever it goes.

No more than having to be 'root' to load a kernel module. You are free to
remove it from any hardware for which you have the right to choose what
software runs.

DS

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

[David Schwartz]

> David Schwartz writes:

> >> First, end users buy and use the hardware in question.  It does not
> >> belong to Tivo, so the analogy to his laptop fails there.

> > No, this is incorrect. They buy *some* of the rights to the
> > hardware but not
> > all of them. Specifically, they do not buy the right to choose
> > what software
> > runs on that hardware. That right is still owned by TiVo.

> Do you have a reference to the contract establishing that cession of
> rights from the buyer to Tivo?

No, and I submit that this is at least arguably something wrong that TiVo is
doing. Note that Microsoft does this too when you buy an Xbox. It has
nothing to do with the GPL.

> To the extent that some contract
> purports to restrict the user in ways contrary to the GPL, I suspect
> Tivo might have a hard time defending it in court.

I agree, however, this doesn't restrict the user in ways contrary to the
GPL. The GPL does not say that you have to be allowed to modify the Linux
running on some particular piece of hardware because that is a legitimate
authorization decision. TiVo not letting you change the software is the same
as me not letting you change the software on my laptop.

> > You can argue that TiVo is being dishonest, breaking the law,
> > being immoral,
> > or whatever in retaining this right or in failing to disclose that they
> > retain it. But you cannot coherently deny that TiVo retains
> > this right when
> > they sell certain other rights to the hardware.

> By the first sale doctrine, someone who buys an item has practically
> unlimited rights to deal with it or dispose of it as the buyer wishes.

This is solely a right against copyright claims. You would be correct if
TiVo were going to sue you for violating some copyright they hold in the
hardware or software if you modified the software.

> The only things that would restrict that are statute or a contract
> entered as part of the sale -- most likely a EULA or other shrink-wrap
> agreement.  Given that most such recognized agreements deal with
> software or services rather than hardware, I am not sure a court would
> recognize a hardware EULA as being binding.  (I suspect this is the
> direction you were heading with the paragraph below.)

Yep, but that has nothing whatsoever to do with the GPL. The exact same
argument applies with the Xbox. It's about whether authorization to modify a
device should or must come with buying that device.

The GPL was never about allowing you to load modified software onto hardware
where the legitimate creators/owners of that hardware say, "no, you may not
modify the software running on this hardware".

DS

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

[Alexandre Oliva]

On Jun 18, 2007, Daniel Hazelton <dhazelton@enter.net> wrote:

> On Monday 18 June 2007 19:31:30 Alexandre Oliva wrote:
>> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

>> Actually, just think of how many times you've heard the argument "I
>> can't give you the source code for this driver/firmware/etc under the
>> GPLv2 because the law says so."

> Sorry to tell you this, but anyone that makes a modification to GPLv2 covered 
> code and distributes that modification is bound by the license.

Of course I know that.  I'm not the one making those arguments.
And then, not all of those pieces of code are indeed moficiations of
GPLv2-covered code, so your objection is off target.

>> > b) I think you're simply wrong in your math. I think more people
>> > like the middle-ground and not-frothing-at-the-mouth spirit of "open
>> > source" over the religious dogma of "free software".
>> 
>> It looks like the math you're talking about is in no way related with
>> anything I've argued about.  You seem to be thinking about the number
>> of people who claim to be on the "free software" or "open source"
>> sides, but I can't fathom in what way this is related with whether you
>> get more or less contributions from users as a consequence of users'
>> being permitted to tinker with the free software in their own devices.

> "More Developers" (either "Free Software" or "Open Source") == "More 
> Contributions"

> That equation is very simple to understand - claiming its wrong is impossible.

YES!  Thank you!  This is exactly the point I'm trying to make.

Now can you please explain this to Linus in terms that his brain won't
dismiss as "coming from a fundamentalist"?

> Apparently because you can't admit that a good reason *IS* a good reason when 
> it conflicts with your belief that the FSF is correct.

No, seriously.  Linus is disputing the equation above, dismissing my
various attempts to show it to him, whenever it appears in teh context
of tivoization, apparently because it doesn't match his moral belief
that tivoization ought to be permitted on his moral grounds.

> PS: I know I've said I'm done with this conversation, but this is like a bad 
> habit. I just couldn't help myself.

You've helped me a lot while at that.  Thanks!

I hope this helps others fundamentalist anti-fundamentalists :-) see
reason too.

-- 
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 box could even be sold by third party vendors, I think they may even
> have started off that way, my old Series 1 had a big Philips logo on it.
> So now we make sure that this hardware refuses to boot any unsigned
> code, but it wasn't shipped containing GPLv3 software, so it's license
> terms simply does not apply.
>
> The software is shipped on a ROM card which can no longer be modified by
> the manufacturer or any third party, so it would seem to comply with the
> GPLv3. I can even imagine that the hardware is really general purpose
> but the ROM is encrypted so that only the BIOS/bootloader can unlock it.
>
> So the GPLv3 seems to fall short on actually preventing tivoization. It
> just requires an extra layer of indirection, ship hardware seperately
> from software.
>
> Jan

If this flaw is still the latest GPLv3 drafts, it should be fixed. It's a
simple, technical error that can easily be rectified.

I don't have the latest draft handy, but my recollection is that it talks
about software that is "transferred along with" hardware. What it should say
is object code that contains activation logic for hardware.

That is, if I ship the Linux kernel (assuming for the moment the Linux
kernel was under GPLv3) with special activation logic to run on platforms X,
Y, and Z, then the source code should have to include how to make my changes
have that same special activation logic for those same platforms.

If the GPLv3 doesn't do that, it's broken.

Note that I'm not opining on whether this is a good thing or a bad thing.
But people who choose the GPLv3 because they want to prohibit tivoization of
their software should in fact get that prohibition.

DS

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

[Alexandre Oliva]

On Jun 18, 2007, Joshua David Williams <yurimxpxman.lkml@gmail.com> wrote:

> The Open Source Definition

... derived from the Debian Free Software Guidelines, engineered to
reflect the Free Software definition ...

> wrote:

>> 9. License Must Not Restrict Other Software

>> Yes, the GPL is conformant with this requirement. Software linked
>> with GPLed libraries only inherits the GPL if it forms a single work,
>> not any software with which they are merely distributed.

> The way I understand it, programs licensed under the GPLv3 are *not* open 
> source software. FSF is so caught up in their own agenda that they're 
> forgetting the whole point - the freedom of choice.

Err...  Excuse me?  Whole point for whom?

Free Software is not about freedom of choice.  That's an OSI slogan
for "if you like, you can shoot your own foot, regardless of whether
the shrapnel hurts people around you".
http://www.fsfla.org/?q=en/node/139#1

Free Software is about respect for the four freedoms.


I don't think the FSF is at all concerned whether GPLv3 complies with
the OSD.  They couldn't care less.  It was OSI that tried to create a
definition that matched exactly the meaning of the Free Software
definition under "more objective criteria".  We already know they
failed, since the Reciprocal Public License is accepted as an OSS
license, but it's a non-Free Software license.  There may be other
examples.


That said, since a number of people already understand the GPLv2
prohibits tivoization, your argument means that either the comment in
the OSD is wrong, and GPLv2 already fails to match the OSD, or that
GPLv3 complies with it in just the same way.

-- 
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 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> > Any number of ways. For example, you probably don't connect the
>> > serial ports
>> > to a device I have access to.

>> But you're not the user of the software on my laptop.  I am.

> Even when I get web pages from your web server?

Yes.  I'm (hypothetically) running the web server such that it serves
web pages to you or anyone else.  You don't become user of a software
just because you establish a network session with it.  You clearly
need more than that.

That said, the precise threshold isn't clear.  For complex web
applications that run part on the client and part on the server, or
even almost entirely on the server, one can argue that a user is
indeed using the software even though it runs mostly on the server.
Some people call this kind of situation the ASP loophole in the GPL.
I'm not sure I want to disturb users of this list with the details
about this, but I'd be pleased to discuss this off the list.

>> The requirements as to "installation information" apply to conveying
>> the program along with a user product.

> In other words, the GPLv3 *compels* a critical authorization decision to
> follow the physical possession of the device. Do you see that, as far as the
> GPLv2 is concerned, this is from outer space?

Not really.  When a user receives a copy of the software, there's
distribution going on, and that's when the user can start having any
expectations of having her freedoms respected as to that software.

>> > How exactly does the GPLv3 specify who should and should not be able to
>> > change the software on a particular physical machine?

>> IANAL, but my understanding is that (paraphrasing), when you convey
>> the software along with a user product, you must permit the recipient
>> of the software to install and run modified versions of the software
>> in the user product as well.

> Which is totally alien to everything in the GPLv2, word and spirit. It never
> required any authorization decisions be made any particular way, nor even
> hinted that authorization decisions were within its scope.

It's the authorization decisions that are alien to GPLv2.  That's just
yet another form of denying users the freedoms that they ought to
receive along with the software.

> What does the spirit of the GPLv2 say about who is authorized to modify the
> software on some particular piece of hardware?

It doesn't.  Why should it have to?  Whether someone is authorized or
not is a direct consequence of the freedoms.  The moment the software
was distributed to you, you're entitled to the freedoms.  Imposing
restrictions on them is a violation of the spirit, if not the letter,
of the license.

>> What if the authority that controls the use of the hardware is
>> forbidding from restricting this possibility by law?  By contractual
>> provisions?  By a patent license?  By a copyright license?

> Those kinds of things are totally alien to the GPL, which was about getting
> the source code and being able to modify it and use it on any hardware for
> which you were authorized to do so.

This is a very limited reading of the GPL that leaves out one of its
most important provisions: the bit about "no further restrictions".

> The GPL was never, until GPLv3, about who gets to make
> authorization decisions.

I can agree with that.  As long as the authorization decisions are not
used as means to deprive users' of the freedoms that must not be
restricted, they can be whatever the distributor fancies.

> You are taking my claim out of contect. I am distinguishing legal obstacles
> from *authorization* obstacles, not technical obstacles.

It doesn't matter how elaborate the excuse to disrespect the freedoms
of the user is.  If there are further restrictions to them, then this
violates the spirit, if not the letter, of the GPL.

>> Someone else's hardware is just a distraction.  You're not a user of
>> software on someone else's hardware.  You have no rights over that.

> You are. In the case of TiVo, the hardware (specifically the right
> to decide what software runs on that hardware) is someone
> else's. That is part of the bundle of rights that owning a piece of
> hardware includes. That is a right you simply do not have with TiVo.

Ah, ok, so I was sloppy above and you caught that.

If someone else places hardware on your home for you to use, even if
they still own it, then you can be a user of someone else's hardware.
And at that point the GPL kicks in, because the software was
distributed to you (even if the hardware wasn't sold), and with the
distributed software come the freedoms, which, per the GPL, the
distributor must not disrespect.

>> > And I think they change it utterly by treating one piece of hardware
>> > different from others for GPL purposes.

>> No, it's tivoization that does this.

> How so?

Like this:

>> Tivoizers say "hey, you can still modify and run the software, just
>> not on *this* hardware".

Tivoization is treating the hardware that comes along with the
software as if it was different from others.  But it isn't.

> Exactly. The GPL is about rights that apply to *all* hardware, not some one
> specific piece.

Exactly!  Just like the GPL doesn't permit the distributor to state
"BTW, you can't install or run this software on your mother's
computer", it doesn't permit the distributor to state "BTW, you can't
install or run this software on this computer I'm selling you".  The
"no further restrictions" applies equally to all computers.  It's not
just because you have some control over some particular hardware that
you deliver along with the software that you're entitled to use that
to limit the user's freedoms.

>> GPLv3 says you must make this artificial distinction.  You must not
>> place barriers on the freedoms of the user WRT to the GPLv3 software
>> they use on the hardware you sold/rented/leased/lent/gave them along
>> with the GPLv3 software you meant them to use.

> Which is a massive departure from the previous GPL spirit which was about
> being able to use the software on *ANY* hardware you controlled, not some
> special pieces more than others.

It doesn't make the sold hardware special.  How come you think it
does?  It's exactly the opposite.  It just says the distributor can't
make the hardware special, so as to restrain the users' freedoms that
are inseparable from the software.

>> You can't waive your hands to escape your obligations saying "you can
>> run it elsewhere", in just the same way you can't escape your GPLv2
>> obligations to provide source code saying "you can download it
>> elsewhere"

> That's a nonsensical comparison. You can run it on any hardware for which
> you have the right to say what software runs.

And why don't I have the right to say what software runs on the
hardware I received along with the GPLed software?  Because the
tivoizer doesn't want me to.  The tivoizer is placing barriers such
that I cannot adapt the GPLed software included in that device to my
own needs.  How is that not a further restriction to the four
freedoms?  How is that not making that hardware special?

> Because that is not a right the vendor chooses to give to the user.

As in, the vendor can turn to the user and sue her for patent
infringement, after distributing GPLed software to her, just because
the use of the patent is not a right the vendor chooses to give to the
user?

> You may dislike this decision, but it's not irrational.

I never said it was irrational.  I just said it's a further
restriction on the exercise of the freedoms that must accompany the
software wherever it goes.

>> > More importantly, the change in scope to claim rights over things
>> > that are not derivative works and do not include any GPL'd code is
>> > so massive that it's a change in spirit, IMO.

>> Show how patents whose licenses are implicitly granted under GPLv2 are
>> derivative works and your argument might begin to make sense.

> The GPL does not claim any control over those patents. If it included
> mandatory licensing of them, then you would have a point.

It doesn't because the US law makes that implicit.  GPLv3 makes it
explicit because it was found that it wasn't like this everywhere.

>> Oh, and user products that GPLv3 talks about *do* include GPLv3 code,
>> otherwise the license is irrelevant for them, since GPLv3 code is not
>> being conveyed.  I guess you meant something else when you wrote "do
>> not include any GPL'ed code".

> The TiVo loader does not include any GPL'ed code. The TiVo signing keys do
> not contain any GPL'ed code. If you are not claiming the GPLv3 exerts any
> control over the loader or the keys, then what is left to assure the user
> can replace the software on his TiVo?

Absence of disrespect for users' freedoms.  Any measure taken by the
vendor to disrespect them is a failure to comply with the obligations
imposed by the spirit, if not the letter, of the license.

And, just in case, 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

[Alexandre Oliva]

On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> First, end users buy and use the hardware in question.  It does not
>> belong to Tivo, so the analogy to his laptop fails there.

> No, this is incorrect. They buy *some* of the rights to the hardware but not
> all of them.

Wow, really?  I thought TiVo actually sold the computer.

Not that it would make a difference as far as GPLv3 is concerned.
It's still a user product, and it still contains GPLed software, and
TiVo distributes that software to other users.

> But you cannot coherently deny that TiVo retains this right when
> they sell certain other rights to the hardware.

Heh.  I mis-parsed "sell rights to the hardware".  How can the
hardware buy something?


Whatever rights TiVo wants to retain or keep from the user is of
little concern here, as long as this doesn't get in the way of the
user's exercise of the freedoms that the GPL stands to defend.  If it
wants to retain more rights than that, then it may have to refrain
from using GPLed software, or face the risk of a court finding it
couldn't have done that in the first place.

-- 
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

[Jan Harkes]

On Mon, Jun 18, 2007 at 08:31:30PM -0300, Alexandre Oliva wrote:
> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > In the GPLv3 world, we have already discussed in this thread how you can 
> > follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM 
> > instead of using a flash device.
> 
> Yes.  This is one option that doesn't bring any benefits to anyone.
> It maintains the status quo for users and the community, but it loses
> the ability for the vendor to upgrade, fix or otherwise control the
> users.  Bad for the vendor.

Not really, Tivo could simply sell you a box without any installed
software. The actual software is mailed to you on a credit card sized
ROM when you activate service. When they want to (or need to) update the
software they send out a new ROM card, maybe yearly as part of the
service subscription renewal.

The box could even be sold by third party vendors, I think they may even
have started off that way, my old Series 1 had a big Philips logo on it.
So now we make sure that this hardware refuses to boot any unsigned
code, but it wasn't shipped containing GPLv3 software, so it's license
terms simply does not apply.

The software is shipped on a ROM card which can no longer be modified by
the manufacturer or any third party, so it would seem to comply with the
GPLv3. I can even imagine that the hardware is really general purpose
but the ROM is encrypted so that only the BIOS/bootloader can unlock it.

So the GPLv3 seems to fall short on actually preventing tivoization. It
just requires an extra layer of indirection, ship hardware seperately
from software.

Jan

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

[Michael Poole]

David Schwartz writes:

>> First, end users buy and use the hardware in question.  It does not
>> belong to Tivo, so the analogy to his laptop fails there.
>
> No, this is incorrect. They buy *some* of the rights to the hardware but not
> all of them. Specifically, they do not buy the right to choose what software
> runs on that hardware. That right is still owned by TiVo.

Do you have a reference to the contract establishing that cession of
rights from the buyer to Tivo?  To the extent that some contract
purports to restrict the user in ways contrary to the GPL, I suspect
Tivo might have a hard time defending it in court.

> You can argue that TiVo is being dishonest, breaking the law, being immoral,
> or whatever in retaining this right or in failing to disclose that they
> retain it. But you cannot coherently deny that TiVo retains this right when
> they sell certain other rights to the hardware.

By the first sale doctrine, someone who buys an item has practically
unlimited rights to deal with it or dispose of it as the buyer wishes.
The only things that would restrict that are statute or a contract
entered as part of the sale -- most likely a EULA or other shrink-wrap
agreement.  Given that most such recognized agreements deal with
software or services rather than hardware, I am not sure a court would
recognize a hardware EULA as being binding.  (I suspect this is the
direction you were heading with the paragraph below.)

Michael Poole

> I do in fact argue that there are things that are wrong with TiVo doing
> this. But they are not GPL-related things. I would make these same arguments
> if the TiVo contained no GPL'd software and I in fact do make them about
> products like the Xbox.
>
> DS

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

[Alexandre Oliva]

On Jun 18, 2007, david@lang.hm wrote:

> Ok, next question, could you do the same thing if you used a CD
> instead of a ROM?

Yes, I believe the very same reasoning applies.

> what makes a blob delivered via a network inherently different from
> the same blob delivered via a plugin ROM or CD?

Err...  Nothing, really, except that delivery over the network doesn't
even have a physical support that you might wish you could possibly
modify but Mother Nature won't let you.

What could make a difference, as far as GPLv3 is concerned, is that
conveyance over the network might be considered more easily separate
from the transaction involving the user product than receiving the CD
or the ROM chip along with it.

But then, a lot of it would depend on the the precise contractual
terms surrounding the transference of the user product, on intent and
how that's interpreted by courts, and on how good a justice money can
buy ;-)

-- 
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

[Kevin Bowling]

> On 6/18/07, Alexandre Oliva <aoliva@redhat.com> wrote:
> > On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> > I just want software back. I think it is *wrong* for me to ask for
> > anything else. It's literally my personal "moral choice": I think the
> > hardware manufacturers need to make their _own_ choices when it comes to
> > _their_ designs.
>
> >  - I think that *technical*quality* is more important than *quantity*.
>
> This argument fails to make the point you're trying to make.

No, it has been countered many times and you are simply proving your ignorance.

>   you trade the potential contributions of all those users for the
>   contributions of tivoizers, apparently assuming that all tivoizers
>   would simply move away from the community, taking their future
>   contributions away from your community, rather than moving to a
>   position in which you'd get not only the contributions from the
>   company itself, but also from all their users
>
> and you say "oh, I don't care about quantity, I care about quality",
> as if this somehow related with the above.

Being strictly pragmatic - what makes you think TiVoland is some
fledgling grounds of /expert kernel developers/ that are otherwise
deprived of contributing, unless they can illegally modify their TiVo?

Under GPLv2, we have the kernel modifications and can include them in
our software.  If you don't agree with TiVo, purchase an open product.
 Download their kernel source and use it on your open product.  Pure
consumerism and capitalism at work.

The GPLv3 is a solution in search of a problem.  Worse, it creates
problems outside the grasp of your understanding.

> Just do the math.  Hypothetically, Linux goes GPLv3, without
> permission to tivoize.  TiVo has to decide among:
>
> - switching to another kernel, no further contributions from them
Bad for us, bad for users.
> - sticking with old version, no further usable contributions from them
Bad for us, bad for users.
> - switching to ROM, still the same contributions from TiVo
Bad for users.
> - no more tivoization, contributions from TiVo and users

Bad for us, bad for users.  Legitimate laws and practices require that
certain devices not be modified by end users.  Therefore TiVo fails
and contributions cease.

> So, you see, in no case do you get more contributors while at the same
> time losing TiVo's quality contributions.

No.  Outside of this FSF {academic,religious} diatribe, in the real
world, things aren't as you see them.  The fact is that these people
can get the code and contribute.  It just won't run on TiVo.  So don't
buy TiVo but use the code.  Problem solved, free software in action.

> > In the GPLv3 world, we have already discussed in this thread how you can
> > follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM
> > instead of using a flash device.
>
> Yes.  This is one option that doesn't bring any benefits to anyone.
> It maintains the status quo for users and the community, but it loses
> the ability for the vendor to upgrade, fix or otherwise control the
> users.  Bad for the vendor.

And users.  Don't spin the facts.

You are advocating things which hurt the end user, which the "Spirit"
should seek to help.  The GPLv3 is here to stroke the FSF ego because
they don't like how somebody has legally found a way to use free
software in a way they don't agree with.  Sort of like dynamite being
used violently...

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

[Bron Gondwana]

On Mon, Jun 18, 2007 at 01:03:40PM -0700, David Schwartz wrote:
> So you're arguing two sides of no argument at all.

Yeah, pretty much.  I take back my arguments in the previous
couple of my posts up this thread.  They don't actually hold
together!  Sorry for wasting your time correct me.

Bron.

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

[Linus Torvalds]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:
> 
> technically, it asks you to pass on (!= give back) access to the
> software (not to the hardware that contains it).

That "technically" is just another way of saying "if you look cross-eyed 
at it, and don't look too closely".

> No.  The reason, again, is the portion you snipped out.
> 
> Could you try again?

No.

Not worth my time. You have shown yourself unable to learn and understand. 

Just live with it. The GPLv2 is better for the kernel.

			Linus

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

[Daniel Hazelton]

On Monday 18 June 2007 19:31:30 Alexandre Oliva wrote:
> On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:
<snip>
>
> > With the GPLv2, you need to give your software modifications back, but
> > the
>                                    ^^^^  BZZT!
> > GPLv2 never *ever* makes any technical limitations on the end result.
>
> Actually, just think of how many times you've heard the argument "I
> can't give you the source code for this driver/firmware/etc under the
> GPLv2 because the law says so."

Sorry to tell you this, but anyone that makes a modification to GPLv2 covered 
code and distributes that modification is bound by the license. If a law 
makes following the license illegal, then they can't use any rights granted 
by the license. They are breaking the law by refusing to follow the license.

<snip>
> > The GPLv2 requires that you give source code out.
>                  ^^^^^^^^ BZZT ;-)
> > But if you want to make your hardware in a way that it only runs
> > signed versions, because of some reason like an FCC rule, or banking
> > rule, or just because you damn well want, the GPLv2 doesn't stop
> > that.
>
> And then, the user is stopped from making appropriate technical
> decisions.

You marked the "requires" as an error. Technically it is. Practically, 
however, it is rare for a modification to not fall under the "distribution" 
part of the license, making the "release the source" requirement active 
almost all the time.

<snip>
> > b) I think you're simply wrong in your math. I think more people
> > like the middle-ground and not-frothing-at-the-mouth spirit of "open
> > source" over the religious dogma of "free software".
>
> It looks like the math you're talking about is in no way related with
> anything I've argued about.  You seem to be thinking about the number
> of people who claim to be on the "free software" or "open source"
> sides, but I can't fathom in what way this is related with whether you
> get more or less contributions from users as a consequence of users'
> being permitted to tinker with the free software in their own devices.

"More Developers" (either "Free Software" or "Open Source") == "More 
Contributions"

That equation is very simple to understand - claiming its wrong is impossible.
<snip>
> > See? Those are three totally different reasons why I think the GPLv2 is
> > the right license for me, and for the kernel.
>
> Ok, the only one that stands is the moral reason.  

Apparently because you can't admit that a good reason *IS* a good reason when 
it conflicts with your belief that the FSF is correct. (The same as 
the "Science can't be right because it conflicts with the bible" I hear from 
all kinds of Xtians these days)

DRH
PS: I know I've said I'm done with this conversation, but this is like a bad 
habit. I just couldn't help myself.

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

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

[david]

On Tue, 19 Jun 2007, Johannes Stezenbach wrote:

> On Mon, Jun 18, 2007, Alexandre Oliva wrote:
>>
>> People talk a lot about TiVo here, but do they the faintest idea of
>> how the conversations with TiVo are proceeding?  I thought so...
>
> Oh, if you know something we don't, could you please fill us in?
> And who was it who coined the "Tivoization" term, thus putting
> TiVo into focus?

what conversations are going on?

Tivo checked years ago and were told that what they were going to do was 
Ok. I don't know thatanyone is talking too Tivo about anything. they are 
just screaming about how evil Tivo is at every public opportunity.

David Lang

>> But since the software is good, and moving to another software would be 
>> costly in various dimentions, the vendor has an incentive to stick with 
>> the software they have.

but if regulations or other contracts require tamper-resistant hardware 
they have no choices other then to fork the existing GPLv2 versions or 
switch to alternate options for anything that switches to GPLv3

David Lang

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

[Alexandre Oliva]

On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> On Mon, 18 Jun 2007, Alexandre Oliva wrote:
>> 
>> 1. I asked you why GPLv2 is better, and you said it was because it
>> promoted giving back in kind.

> Where I explained that "in kind" was about *software*.

Yes, we'd already established that.

>> 2. I asked you what you didn't like about GPLv3, and you said it was
>> Tivoization.

> Right. The GPLv3 asks you to give back *money*.

> That's like the Microsoft license agreements. I don't like them either.

> Oh, and replace "money" with "access to hardware", to make that thing 
> technically correct.

technically, it asks you to pass on (!= give back) access to the
software (not to the hardware that contains it).

>> 3. Then I argued that, since Tivoization enables tivoizers to remove
>> some motivation for potential developers (= their customers)

> That's simply not my *reason* for doing "tit-for-tat".

No.  The reason, again, is the portion you snipped out.

Could you try again?

> I just want software back. I think it is *wrong* for me to ask for 
> anything else. It's literally my personal "moral choice": I think the 
> hardware manufacturers need to make their _own_ choices when it comes to 
> _their_ designs.

It's comforting to see that you're not the pure-pragmatics-no-morals
character that some people (yourself included) try to paint you as.
Thank you for this.  I can relate with that.  I can easily respect
that, as much as I think that (poor attempt at humor follows, no
offense really intended) standing up for the freedoms of the poor
hardware manufacturers against the evil software developers who want
to control the ways they can use to control their customers serves the
common good or even your own stated interests.

> But if you actually want to discuss "number of developers" and their 
> motications, I actually have another few arguments for you:

>  - I just personally think your math is bogus. I think more people think 
>    like I do, than people think like you and the FSF does.

> But I don't even depend on that. Because:

>  - I think that *technical*quality* is more important than *quantity*.

This argument fails to make the point you're trying to make.  I wrote:

  you trade the potential contributions of all those users for the
  contributions of tivoizers, apparently assuming that all tivoizers
  would simply move away from the community, taking their future
  contributions away from your community, rather than moving to a
  position in which you'd get not only the contributions from the
  company itself, but also from all their users

and you say "oh, I don't care about quantity, I care about quality",
as if this somehow related with the above.

Just do the math.  Hypothetically, Linux goes GPLv3, without
permission to tivoize.  TiVo has to decide among:

- switching to another kernel, no further contributions from them

- sticking with old version, no further usable contributions from them

- switching to ROM, still the same contributions from TiVo

- no more tivoization, contributions from TiVo and users

So, you see, in no case do you get more contributors while at the same
time losing TiVo's quality contributions.

The argument is not about quality vs quantity, it's about getting
lots of additional contributions along with what's already in place,
VS ending up without some quality contributions you get today.

> With the GPLv2, you need to give your software modifications back, but the 
                      ^^^^  BZZT!
> GPLv2 never *ever* makes any technical limitations on the end result.

Actually, just think of how many times you've heard the argument "I
can't give you the source code for this driver/firmware/etc under the
GPLv2 because the law says so."

> In the GPLv3 world, we have already discussed in this thread how you can 
> follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM 
> instead of using a flash device.

Yes.  This is one option that doesn't bring any benefits to anyone.
It maintains the status quo for users and the community, but it loses
the ability for the vendor to upgrade, fix or otherwise control the
users.  Bad for the vendor.

As another option, the vendor can respect users' freedoms, and then
everybody wins big.  That's the option that anti-tivoization provides
economic incentive for vendors to take.  Sure, they may still prefer
the alternative above, or stick with an older version (which has its
costs), or move to different software (which also has its costs), but
it's unreasonable to claim that I'm advocating for vendors to move to
ROM.

I'm saying they have this option.  I'm advocating for them to respect
users' freedom.  And if that's incompatible with their business model,
well, so what?  GPLv2 and Free Software in general are incompatible
with a number of business models too, and who's complaining?  Heck,
even using slave work-forces was part of legitimate business models at
some point in time.

> The GPLv2 requires that you give source code out.
            ^^^^^^^^ BZZT ;-)
> But if you want to make your hardware in a way that it only runs
> signed versions, because of some reason like an FCC rule, or banking
> rule, or just because you damn well want, the GPLv2 doesn't stop
> that.

And then, the user is stopped from making appropriate technical
decisions.

> The GPLv3 doesn't stop it *either*, but the GPLv3 requries that you
> make the INFERIOR TECHNICAL CHOICE.

This is a lie (by which I don't mean it's malicious).

What it requires is the vendor to decide between making the inferior
technical choice and respecting users' freedoms to make their own
technical choices.

> b) I think you're simply wrong in your math. I think more people
> like the middle-ground and not-frothing-at-the-mouth spirit of "open
> source" over the religious dogma of "free software".

It looks like the math you're talking about is in no way related with
anything I've argued about.  You seem to be thinking about the number
of people who claim to be on the "free software" or "open source"
sides, but I can't fathom in what way this is related with whether you
get more or less contributions from users as a consequence of users'
being permitted to tinker with the free software in their own devices.

>    I think Linux has pretty much proved my point. Look at Hurd, then
>    look at Linux. Am I *that* much better than the Hurd developers
>    (yes, of course I am, but let's assume not). Or is it just that
>    my approach of being more _pragmatic_ about things rather than
>    talking about those "four freedoms" all the time was just much
>    easier for people to accept?

I could argue that Hurd took a different approach that proved to be
far more difficult, and that the urgency for the development of a Free
Software kernel by the GNU project disappeared with the relicensing of
Linux under a Free Software license (thanks!).  But I guess you'll
just dismiss that on whatever reasons move you, and I don't really
care about these particular historical issues to spend time discussing
them.

> See? Those are three totally different reasons why I think the GPLv2 is 
> the right license for me, and for the kernel.

Ok, the only one that stands is the moral reason.  That's a good one,
but it contradicts the stated reasons as to why you prefer GPLv2 over
GPLv3, stating it was a purely pragmatic decision, based on getting
more software contributions back, based on an assumption, that so far
lacks any material evidence, that permission for tivoization somehow
gets you more of that than getting just as many contributions from the
former-tivoizer, plus whatever any of their users decide to
contribute.

-- 
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

[Johannes Stezenbach]

On Mon, Jun 18, 2007, Alexandre Oliva wrote:
> 
> People talk a lot about TiVo here, but do they the faintest idea of
> how the conversations with TiVo are proceeding?  I thought so...

Oh, if you know something we don't, could you please fill us in?
And who was it who coined the "Tivoization" term, thus putting
TiVo into focus?

> So, you see, when people who oppose anti-tivoization measure the
> outcome for the community, they only look at the second possibility,
> assuming the vendor would immediately switch to some other software.
> As if that was easy for the vendor, and as if the software sucked so
> much that the vendor was just looking for a reason to switch.
> 
> But since the software is good, and moving to another software would
> be costly in various dimentions, the vendor has an incentive to stick
> with the software they have.

Hm, you only talk about people who already use free software,
but I tried to make you aware of the importance of
_promoting_ free software, i.e. winning new people and
companies for the free software idea.

I think the majority of embedded devices still run proprietary
RTOSes, and the majority of desktops still run Windows or Mac OS.
Don't you want to change that?

There are dozens of proprietary RTOSes, and along with
them dozens of proprietary toolchains, development environments
and trace/debug tools. Companies which worked in this field
for decades have invested money to create proprietary software
on top of them, and to train their staff to use them. Those companies
won't switch to Linux lightly. And it won't be a singular event,
but a process. They might start low, and maybe (hopefully) might
become well-playing free software contributors. But if you raise
the entry barrier too high, they won't get started at all.

OK, I don't have experience talking to big companies, but
I have talked to people working for smaller ones. They are
aware of the trend towards Linux, but are afraid that the
obligations of the GPL might be impractical for them.
Then they only have the choice to not use Linux, or to use
"creative workarounds".

It's true that what these companies do might have little direct benefit
for users buying their products, however the long term benefits of
getting the people in these companies exposed to free software ideas,
and in contact with the free software community, can only be
positive -- I think it's more important to spread the general
idea of free software into as many minds as possible than to
ensure that few follow the pure spirit of the FSF free software
definition in every detail.

> So you see, the picture of anti-tivozation is not as bleak as people
> try to frame it.  In fact, it's not bleak at all.  If one out of 10,
> maybe even 1 out of 100 vendors start respecting users' freedoms, when
> faced with anti-tivoization provisions, the community will already win
> big time, because each vendor is likely to have thousands of
> customers, some of which will use the freedoms to serve the goals of
> the community, in the very terms the community claims to care about.

Does this multiplicator also apply to new companies
which start using free software for their products?

I think the FSF strategy is suboptimal. The Linux
strategy works better for promoting free software.

In the end I want my devices to be open and hackable, too,
and I'm sure it will take an effort to convince companies
to open up. But I'm not convinced that the GPLv3 is a
step in the right direction towards that goal.


Johannes

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

[Alan Cox]

> the software on your laptop is owned by people like Linus, Al Viro, David 
> M, Alan Cox, etc.

Not quite that simple. An easier way to think about this one is books.
You own the book but you don't own the right to reproduce the words
within. You can however boil the book, use it as bog roll or read it (not
in that order I suggest)

Alan

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

[David Schwartz]

> First, end users buy and use the hardware in question.  It does not
> belong to Tivo, so the analogy to his laptop fails there.

No, this is incorrect. They buy *some* of the rights to the hardware but not
all of them. Specifically, they do not buy the right to choose what software
runs on that hardware. That right is still owned by TiVo.

You can argue that TiVo is being dishonest, breaking the law, being immoral,
or whatever in retaining this right or in failing to disclose that they
retain it. But you cannot coherently deny that TiVo retains this right when
they sell certain other rights to the hardware.

I do in fact argue that there are things that are wrong with TiVo doing
this. But they are not GPL-related things. I would make these same arguments
if the TiVo contained no GPL'd software and I in fact do make them about
products like the Xbox.

DS

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

[David Schwartz]

> > But you're not the user of the software on my laptop.  I am.

> ahh, but by your own argument you aren't

Let's not confuse owner with user and let's not confuse ownership of
copyrights with ownership of particular copies.

> the software on your laptop is owned by people like Linus, Al Viro, David
> M, Alan Cox, etc.

No. The copyright to the software is owned by those people. But particular
copies of copyrighted items can be owned by other people.

> they have the right to put a license on that software that would require
> you to give them access to your hardware (after all, that's the argument
> that you are useing to justify requireing Tivo to give you access
> to their hardware)

That's right, they do have that right so long as they condition it on the
exercise of something I could not do without their permission. (Ignoring for
the moment the fact that the software is a derivative work of GPL'd
software.)

I'm not sure whether you think this disagrees with or refutes anything I've
said.

DS

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

[Joshua David Williams]

On 6/17/07, Linus Torvalds <torvalds@linux-foundation.org> wrote:

>  Everybody else just cares about the legal reasons.

> The "legal terms" is the only reason a license *exists*. That's what a
> license *is*, for crying out loud!

>  If you don't care about the legal side, go and read the free software
> manifesto. That's the paper you're really arguing about.

> If you want to argue about the GPLv2 *license*, then you'd better start
> caring about the legal issues. Because that is what the license is: a
> _legal_ document.

IMHO, free and open source software seem to differ on one key point:

The Open Source Definition wrote:

> 9. License Must Not Restrict Other Software
> 
> The license must not place restrictions on other software that
> is distributed along with the licensed software. For example,
> the license must not insist that all other programs distributed
> on the same medium must be open-source software.
> 
> Rationale: Distributors of open-source software have the right
> to make their own choices about their own software.

> Yes, the GPL is conformant with this requirement. Software linked
> with GPLed libraries only inherits the GPL if it forms a single work,
> not any software with which they are merely distributed.

The way I understand it, programs licensed under the GPLv3 are *not* open 
source software. FSF is so caught up in their own agenda that they're 
forgetting the whole point - the freedom of choice. The GPLv2 may 
be "conformant with this requirement", but it goes against the ethics of the 
FSF, so we can't expect each new version of the GPL to comply to this right.

Attacking this so-called "tivozation", IMO, finally draws a distinct line 
between "free" and "open source". We, as open source developers, are not 
politicians or philosophers; we write software, and we wish to publish our 
code under a certain set of ten rights.

Yes, the GPL is a legal document, but it was written in order to compliment 
the GNU Manifesto by setting legal parameters for which they could publish 
their code under.

Until now, the GPL (v2) has had the same *legal* paramaters the open source 
developers need in order to do the job we need it to do. We have clearly went 
our separate ways now, so I think it's time to drop the GPL. (See my other 
thread about writing an open source license.)

Anyways, that's my $0.02.

-- 
It's a commonly known fact that most intruders come in through Windows.

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

[Al Viro]

On Mon, Jun 18, 2007 at 02:09:13PM -0700, Linus Torvalds wrote:

> Quite frankly, I don't *want* to attract develpers that are not 
> technically "up to snuff". And if you think that making the technically 
> worse decisions is the "rigth decision", then hey, you're clearly not in 
> the same technical quality range as I am, or Al Viro is.

You are confusing being generally fucked in head (aka being a true
believer) with being willing to go for technically worse decisions
*when* *it* *comes* *to* *making* *them*.  Which would be in gcc-related
work in this case.  I don't see any evidence of the latter...

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

[Al Viro]

On Sun, Jun 17, 2007 at 02:56:24AM -0300, Alexandre Oliva wrote:
> Can you please acknowledge that it doesn't, such that I can feel I've
> fulfilled my goal of dispelling the myth that the GPLv3 changes the
> spirit of the GPL?

No.  I don't do metaphysics.  This thread alone has shown that the
notion is not well-defined *at* *all*, to the point of being useless
and seriously misleading.  I.e. the phrase about similar spirit
should be replaced with something far more explicit and very, very
hard to miss.  I don't think you need more proof that people *do*
interpret it in very different ways, with quite unpleasant results.
 
> > GPLv3, with your involvement in its development or not, sucks rocks,
> > thanks to what you call anti-tivoization section.
> 
> Is it correct to say that you share Linus' opinion, that the only
> problem with the GPLv3 is the anti-tivoization provision?

No.  If you want a basic splitup by sections compared to GPLv2,
	1	-	at least not better; attempts at being precise
			end up creating a no-common-sense-land *and*
			turn out to leave serious unanswered questions
			in that area.
	2	-	no opinion on actual changes
	3	-	more or less an improvement
	4,5	-	about on par with v2, modulo wording in (5)
	6	-	much worse
	7	-	if I want to give additional permissions, I don't
			want them stripped, for fsck sake!  There is a
			bog-standard mechanism for _that_ (dual-licensing),
			thank you very much.  I.e. that section looks like				a pile of dishonest PR games, pardon the redundance.
	8	-	on par
	9	-	on par, modulo piss-poor attempt to define "modify"
			backfiring here (e.g. prelinking constitutes
			modification according to it, so does running rdev(8),
			etc., etc.)
	10	-	no opinion on actual changes
	11	-	improvement
	12	-	on par (aside of basic bad writing, but there are
			much worse problems *not* with wording, so that's
			not interesting)
	13	-	special-case kludges are fun, aren't they (specifically
			"linking"?), but in any case, that's secondary.
			FWIW, I'm not fond of ideas behind Affero, so if
			anything, that's a point against v3.
	14	-	... and thank you very much for keeping such a lovely
			source of periodic clusterfucks in v3 as well.
			I think it's painfully obvious for everyone in this
			thread that reference to "spirit" is a recipe for
			massive disagreements down the road.  If you want the
			words you are using to be interpreted your way, use
			ones that have commonly agreed upon meaning.  The
			measure is "do other people read it differently?",
			not "how sure I am in deriving the meaning I want from
			the words I've used?".  Related problem is that
			version choice rules _must_ be stated in maximally
			unambiguous and hard to miss way.  Look through
			Bernd-produced parts of this thread and you'll see
			the reason why it is needed.
			Moving that into terms and conditions is a good step,
			but it's still not enough.  E.g. you really want
			to be explicit on the form (in)sufficient to specify
			the version of license.
	the rest	on par.

Overall: definitely worse than v2.  v2 + (3) + (11) would be an improvement,
provided that v2 section 9 is cleaned up.

> To make this more concrete, if there was a hypothetical GPLv2.9,
> consisting of GPLv3dd4 minus the "installation information"
> requirements for user products, (i) Would you consider it a better
> license than GPLv2?
	Negative, see above
 (ii) Better for Linux?
	Negative, for kernel as well as for userland
(iii) Enough to go through the trouble of switching?
	See above.

In other words, I don't see any chance for v3 to be a good choice
for anything I write, kernel or userland.  If I end up sending patches
to v3 projects, I'll put the patches under BSDL and let them convert
on merge.

Note that this is *not* about the problems with wording; those also exist,
of course (_that_ is a final draft?), but that's a separate story and it
interests me only inasmuch as it is caused by inherent problems with meaning
of section in question.

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

[Theodore Tso]

On Mon, Jun 18, 2007 at 12:48:13PM -0700, Greg KH wrote:
> > > So, why would we want to waste our time filling out web forms after
> > > that?
> > 
> > If you're adamantly favorable to permitting any form of Tivoization
> > whatsoever, don't bother.
> 
> For the record, I completely feel that what Tivo did was both legally
> correct[1], and the correct thing to do for their system, and would
> fight _very_ hard any attempt to change the Linux kernel's license that
> would prevent this usage model.
> 
> So I will not bother anymore.

Linus has spoken, Greg K-H has spoken, many other people spoken ---
and yet Alexandre keeps on speaking, and speaking, and speaking....

It's pretty clear no one is convincing anyone, and that everyone
understands their position, and are happy with it, and so all we're
doing now is wasting bandwidth.

Can we please end this thread?  

Please?

						- Ted

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

[david]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:

> On Jun 18, 2007, david@lang.hm wrote:
>
>> On Mon, 18 Jun 2007, Alexandre Oliva wrote:
>>> On Jun 18, 2007, david@lang.hm wrote:
>>>
>>>> they want to prevent anyone from modifying the credit card machine to
>>>> store copies of all the card info locally.
>>>
>>> I see.  Thanks for enlightening me.
>>>
>>>> you don't really answer this issue. since these boxes are required to
>>>> be sealed and physically anti-tamper, changing the ROM is not
>>>> acceptable.
>>>
>>> Given the ROM exception in GPLv3, I guess you could seal and
>>> anti-tamper it as much as you want, and leave the ROM at such a place
>>> in which it's easily replaceable but with signature checking and all
>>> such that the user doesn't install ROM that is not authorized by you.
>
>> 'sealed, but easy to replace ROM containing the programming' is a
>> contridiction.
>
>> if a local person can easily replace the programming it doesn't meet
>> the PCI requirements and therefor you just cannot use GPLv3 code for
>> this sort of application.
>
> How can someone easily replace the programming if there's signature
> checking and all?
>
> The sealing of the ROMmed software is accomplished by other means, but
> it's there.  I shall mention that I'm not endorsing or recommending
> this practice, it might very well be copyright infringement even under
> GPLv1 and v2.

Ok, next question, could you do the same thing if you used a CD instead of 
a ROM?

what makes a blob delivered via a network inherently different from the 
same blob delivered via a plugin ROM or CD?

David Lang

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

[Michael Poole]

david@lang.hm writes:

> On Mon, 18 Jun 2007, Alexandre Oliva wrote:
>
>> On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:
>>
>>>>> Sure, and you use the hardware to stop me from modifying the
>>>>> Linux on your
>>>>> laptop.
>>
>>>> Do I?  How so?
>>
>>> Any number of ways. For example, you probably don't connect the serial ports
>>> to a device I have access to.
>>
>> But you're not the user of the software on my laptop.  I am.
>
> ahh, but by your own argument you aren't
>
> the software on your laptop is owned by people like Linus, Al Viro,
> David M, Alan Cox, etc.

To be pedantic, the *copyrights* for certain software on his laptop
are owned by those people.  (Fortunately, they have been friendly
enough to engage in software quid-pro-quo with those rights.)

> they have the right to put a license on that software that would
> require you to give them access to your hardware (after all, that's
> the argument that you are useing to justify requireing Tivo to give
> you access to their hardware)

Even as straw men go, that is pretty incoherent.

First, end users buy and use the hardware in question.  It does not
belong to Tivo, so the analogy to his laptop fails there.

Second, the important access is not to the hardware, but to the bits
used to build the version of Linux that is distributed by Tivo.  This
is purely software.

Third, such a license would be neither a free software nor an open
source license.  No one argues it would be.

Michael Poole

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

[Alexandre Oliva]

On Jun 18, 2007, david@lang.hm wrote:

> On Mon, 18 Jun 2007, Alexandre Oliva wrote:
>> On Jun 18, 2007, david@lang.hm wrote:
>> 
>>> they want to prevent anyone from modifying the credit card machine to
>>> store copies of all the card info locally.
>> 
>> I see.  Thanks for enlightening me.
>> 
>>> you don't really answer this issue. since these boxes are required to
>>> be sealed and physically anti-tamper, changing the ROM is not
>>> acceptable.
>> 
>> Given the ROM exception in GPLv3, I guess you could seal and
>> anti-tamper it as much as you want, and leave the ROM at such a place
>> in which it's easily replaceable but with signature checking and all
>> such that the user doesn't install ROM that is not authorized by you.

> 'sealed, but easy to replace ROM containing the programming' is a
> contridiction.

> if a local person can easily replace the programming it doesn't meet
> the PCI requirements and therefor you just cannot use GPLv3 code for
> this sort of application.

How can someone easily replace the programming if there's signature
checking and all?

The sealing of the ROMmed software is accomplished by other means, but
it's there.  I shall mention that I'm not endorsing or recommending
this practice, it might very well be copyright infringement even under
GPLv1 and v2.

-- 
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 Mon, 18 Jun 2007, Linus Torvalds wrote:
> 
>  a) I don't personally feel like I have the "moral authority" to require
>     hardware designers to give access to their hardware to me.
> 
>     I can tell them that I *like* open hardware more than closed hardware, 
>     but they designed the hardware, and as a result I think it's *their* 
>     choice.
> 
>     In contrast, I _do_ think I have the moral authority to ask for 
>     modifications to the _software_ back. Because they didn't design it, 
>     they just improved on it.

Btw, this may explain the differences between rms/FSF and me.

Richard Stallman is out to "improve the world". He thinks he has not just 
the moral authority and right, but the _obligation_ to tell others how to 
live their lives.

So rms feels he has to "spread his message", the same way those annoying 
people come to your door on Saturday in order to "save you". rms literally 
is saving you from a sin.

Me, not so much. I'm not out to "improve the world". I'm not saving you 
from any sin. I just have this project I started, and I think that if you 
improve on the project, you should let me use your improvements.

And I don't ask you use that license anywhere else. For your other 
projects, you should follow *their* licenses. I'm just asking you to honor 
my choice of license for those projects I started, and I designed, and 
where I conned other people into helping me under the same rules.

				Linus

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

[David Schwartz]

> > Any number of ways. For example, you probably don't connect the
> > serial ports
> > to a device I have access to.

> But you're not the user of the software on my laptop.  I am.

Even when I get web pages from your web server?

> > I'm sorry, who is "the user"? Who exactly is supposed to be
> > able to install
> > and run modified versions? How does the GPLv3 specify who is
> > supposed to be
> > authorized to do this?

> Aah, good question.  Here's what the draft says about this:
>
>   Mere interaction with a user through a computer network, with no
>   transfer of a copy, is not conveying.
>
> The requirements as to "installation information" apply to conveying
> the program along with a user product.

In other words, the GPLv3 *compels* a critical authorization decision to
follow the physical possession of the device. Do you see that, as far as the
GPLv2 is concerned, this is from outer space?

> > How exactly does the GPLv3 specify who should and should not be able to
> > change the software on a particular physical machine?

> IANAL, but my understanding is that (paraphrasing), when you convey
> the software along with a user product, you must permit the recipient
> of the software to install and run modified versions of the software
> in the user product as well.

Which is totally alien to everything in the GPLv2, word and spirit. It never
required any authorization decisions be made any particular way, nor even
hinted that authorization decisions were within its scope.

In fact, there were many discussions where it was made clear that GPLv2
specifically allowed you to make the authorization decisions any way you
want, because it permitted anyone else to remove them. For GPLv3,
apparently, that anyone can remove them is not good enough.

> >> A condition that is
> >> arguably already encoded in the "no further restrictions to the rights
> >> granted" by the license" and to the requirement for complete
> >> corresponding source code to accompany the binary.
>
> > Except that the "right" to upload the software on some
> > particular piece of
> > hardware was *never* a right granted by the GPL, nor could it be.

> It is a restriction on adapting the software installed in the machine,
> and a restriction on running the software on that machine.  You can
> argue these are not granted by GPLv2.  You may be right.  But per the
> spirit of the GPL, they should be protected, and so GPLv3 fixes the
> legal conditions such that they are.

What does the spirit of the GPLv2 say about who is authorized to modify the
software on some particular piece of hardware? This is not per the spirit of
the GPL, it's totally alien to the spirit of the GPL. It has always been
explicitly clear (I can dig up the old discussions if needed) the the GPL
stayed totally away from authorization. Otherwise, you could argue that the
fact that a non-root user can't install a modified kernel "violates the
spirit of the GPL".

> > That *HAS* to be a right granted by whatever authority controls the
> > use of that hardware.

> What if the authority that controls the use of the hardware is
> forbidding from restricting this possibility by law?  By contractual
> provisions?  By a patent license?  By a copyright license?

Those kinds of things are totally alien to the GPL, which was about getting
the source code and being able to modify it and use it on any hardware for
which you were authorized to do so.

> > It's totally obvious that who gets to install what software on a
> > given piece of hardware is determined by the person who creates/owns
> > that hardware and they have to authorize anyone else to change it.

> If who creates and who owns are different people, who gets to decide it?

That's a question on which I would likely agree with you, but it has *ZERO*
to do with the GPL. The GPL was never, until GPLv3, about who gets to make
authorization decisions.

> > It is not. The GPL was never about who was allowed to modify the
> > software on particular pieces of hardware. It was about the lack of
> > *legal* obstacles to your doing so.

> GPL has never been concerned *only* about *legal* obstacles.  In fact,
> the only obstacle GPLv1 addressed by name was not a legal, but a
> technical obstacle: denying access to source code.  Your distinction
> is flawed.

You are taking my claim out of contect. I am distinguishing legal obstacles
from *authorization* obstacles, not technical obstacles. Tivoization is
about authorization even though that authorization is enforced by technical
means.

> >> Both are means to disrespect users' freedoms.
>
> > The freedom to control what software runs on someone else's hardware?!

> Freedom to control the software you use on the hardware you use it.

But that's not a freedom, that's an authorization right that belongs to
someone. Someone gets to choose what software runs on what hardware.

> Someone else's hardware is just a distraction.  You're not a user of
> software on someone else's hardware.  You have no rights over that.

You are. In the case of TiVo, the hardware (specifically the right to decide
what software runs on that hardware) is someone else's. That is part of the
bundle of rights that owning a piece of hardware includes. That is a right
you simply do not have with TiVo.

With respect to control over what software runs on it, your TiVo is someone
else's.

> > And I think they change it utterly by treating one piece of hardware
> > different from others for GPL purposes.

> No, it's tivoization that does this.

How so?

> Tivoizers say "hey, you can still modify and run the software, just
> not on *this* hardware".

Exactly. The GPL is about rights that apply to *all* hardware, not some one
specific piece. That's a massive change in the spirit of the GPL. (Special
rights to one piece of hardware.)

> GPLv3 says you must make this artificial distinction.  You must not
> place barriers on the freedoms of the user WRT to the GPLv3 software
> they use on the hardware you sold/rented/leased/lent/gave them along
> with the GPLv3 software you meant them to use.

Which is a massive departure from the previous GPL spirit which was about
being able to use the software on *ANY* hardware you controlled, not some
special pieces more than others.

> You can't waive your hands to escape your obligations saying "you can
> run it elsewhere", in just the same way you can't escape your GPLv2
> obligations to provide source code saying "you can download it
> elsewhere"

That's a nonsensical comparison. You can run it on any hardware for which
you have the right to say what software runs.

> > GPL was always about equal freedom to use the software on *ALL*
> > hardware, not special rights to use it on one piece of hardware.

> Exactly.  But tivoizers are making these distinctions, trying to frame
> their hardware as somehow special, even though the users that receive
> the hardware with the software become users of the software on that
> very hardware, and that's why they must be able to enjoy the freedoms
> on that very hardware.  Not being able to enjoy them elsewhere could
> defensibly be not the vendor's fault.  Not being able to enjoy them on
> that hardware is obviously the result of choices made by the vendor,
> since the vendor *could* put the software there and get it to run.
> Why couldn't the user?

Because that is not a right the vendor chooses to give to the user. You may
dislike this decision, but it's not irrational. There are any number of
reasons you might want a device to be "trusted".

> > More importantly, the change in scope to claim rights over things
> > that are not derivative works and do not include any GPL'd code is
> > so massive that it's a change in spirit, IMO.

> Show how patents whose licenses are implicitly granted under GPLv2 are
> derivative works and your argument might begin to make sense.

The GPL does not claim any control over those patents. If it included
mandatory licensing of them, then you would have a point.

> Oh, and user products that GPLv3 talks about *do* include GPLv3 code,
> otherwise the license is irrelevant for them, since GPLv3 code is not
> being conveyed.  I guess you meant something else when you wrote "do
> not include any GPL'ed code".

The TiVo loader does not include any GPL'ed code. The TiVo signing keys do
not contain any GPL'ed code. If you are not claiming the GPLv3 exerts any
control over the loader or the keys, then what is left to assure the user
can replace the software on his TiVo?

DS

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

[david]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:

> On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:
>
>>>> Sure, and you use the hardware to stop me from modifying the
>>>> Linux on your
>>>> laptop.
>
>>> Do I?  How so?
>
>> Any number of ways. For example, you probably don't connect the serial ports
>> to a device I have access to.
>
> But you're not the user of the software on my laptop.  I am.

ahh, but by your own argument you aren't

the software on your laptop is owned by people like Linus, Al Viro, David 
M, Alan Cox, etc.

they have the right to put a license on that software that would require 
you to give them access to your hardware (after all, that's the argument 
that you are useing to justify requireing Tivo to give you access to their 
hardware)

David Lang

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

[Alexandre Oliva]

On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> > Sure, and you use the hardware to stop me from modifying the
>> > Linux on your
>> > laptop.

>> Do I?  How so?

> Any number of ways. For example, you probably don't connect the serial ports
> to a device I have access to.

But you're not the user of the software on my laptop.  I am.

> I'm sorry, who is "the user"? Who exactly is supposed to be able to install
> and run modified versions? How does the GPLv3 specify who is supposed to be
> authorized to do this?

Aah, good question.  Here's what the draft says about this:

  Mere interaction with a user through a computer network, with no
  transfer of a copy, is not conveying.

The requirements as to "installation information" apply to conveying
the program along with a user product.


> How exactly does the GPLv3 specify who should and should not be able to
> change the software on a particular physical machine?

IANAL, but my understanding is that (paraphrasing), when you convey
the software along with a user product, you must permit the recipient
of the software to install and run modified versions of the software
in the user product as well.

>> A condition that is
>> arguably already encoded in the "no further restrictions to the rights
>> granted" by the license" and to the requirement for complete
>> corresponding source code to accompany the binary.

> Except that the "right" to upload the software on some particular piece of
> hardware was *never* a right granted by the GPL, nor could it be.

It is a restriction on adapting the software installed in the machine,
and a restriction on running the software on that machine.  You can
argue these are not granted by GPLv2.  You may be right.  But per the
spirit of the GPL, they should be protected, and so GPLv3 fixes the
legal conditions such that they are.

> That *HAS* to be a right granted by whatever authority controls the
> use of that hardware.

What if the authority that controls the use of the hardware is
forbidding from restricting this possibility by law?  By contractual
provisions?  By a patent license?  By a copyright license?

> It's totally obvious that who gets to install what software on a
> given piece of hardware is determined by the person who creates/owns
> that hardware and they have to authorize anyone else to change it.

If who creates and who owns are different people, who gets to decide it?

> It is not. The GPL was never about who was allowed to modify the
> software on particular pieces of hardware. It was about the lack of
> *legal* obstacles to your doing so.

GPL has never been concerned *only* about *legal* obstacles.  In fact,
the only obstacle GPLv1 addressed by name was not a legal, but a
technical obstacle: denying access to source code.  Your distinction
is flawed.

>> Both are means to disrespect users' freedoms.

> The freedom to control what software runs on someone else's hardware?!

Freedom to control the software you use on the hardware you use it.

Someone else's hardware is just a distraction.  You're not a user of
software on someone else's hardware.  You have no rights over that.

> And I think they change it utterly by treating one piece of hardware
> different from others for GPL purposes.

No, it's tivoization that does this.

Tivoizers say "hey, you can still modify and run the software, just
not on *this* hardware".

GPLv3 says you must make this artificial distinction.  You must not
place barriers on the freedoms of the user WRT to the GPLv3 software
they use on the hardware you sold/rented/leased/lent/gave them along
with the GPLv3 software you meant them to use.

You can't waive your hands to escape your obligations saying "you can
run it elsewhere", in just the same way you can't escape your GPLv2
obligations to provide source code saying "you can download it
elsewhere"

> GPL was always about equal freedom to use the software on *ALL*
> hardware, not special rights to use it on one piece of hardware.

Exactly.  But tivoizers are making these distinctions, trying to frame
their hardware as somehow special, even though the users that receive
the hardware with the software become users of the software on that
very hardware, and that's why they must be able to enjoy the freedoms
on that very hardware.  Not being able to enjoy them elsewhere could
defensibly be not the vendor's fault.  Not being able to enjoy them on
that hardware is obviously the result of choices made by the vendor,
since the vendor *could* put the software there and get it to run.
Why couldn't the user?

> More importantly, the change in scope to claim rights over things
> that are not derivative works and do not include any GPL'd code is
> so massive that it's a change in spirit, IMO.

Show how patents whose licenses are implicitly granted under GPLv2 are
derivative works and your argument might begin to make sense.


Oh, and user products that GPLv3 talks about *do* include GPLv3 code,
otherwise the license is irrelevant for them, since GPLv3 code is not
being conveyed.  I guess you meant something else when you wrote "do
not include any GPL'ed 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

[Linus Torvalds]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:
> 
> 1. I asked you why GPLv2 is better, and you said it was because it
> promoted giving back in kind.

Where I explained that "in kind" was about *software*.

> 2. I asked you what you didn't like about GPLv3, and you said it was
> Tivoization.

Right. The GPLv3 asks you to give back *money*.

That's like the Microsoft license agreements. I don't like them either.

Oh, and replace "money" with "access to hardware", to make that thing 
technically correct. But the point is, that's what I don't like about the 
GPLv3.

> 3. Then I argued that, since Tivoization enables tivoizers to remove
> some motivation for potential developers (= their customers)

That's simply not my *reason* for doing "tit-for-tat". My basic reason for 
"tit-for-tat" was not about "lots of potential developers", but simply 
because I think it's the right choice for me!

Can you not understand that? I simply DO NOT LIKE TO CONTROL PEOPLE!

I just want software back. I think it is *wrong* for me to ask for 
anything else. It's literally my personal "moral choice": I think the 
hardware manufacturers need to make their _own_ choices when it comes to 
_their_ designs.

I feel that I have the moral right to ask for modifications to the kernel 
(because I started it), but I *personally* am very unhappy about asking 
people to also give their hardware access. That's *their* choice.

Is that really so hard to understand? I ask you to respect _my_ choice wrt 
license for my software, but the same way I expect others to respect my 
choices, I also myself need to respect *their* choices. 

So to me, it's the hardware manufacturers choice to to select the license 
for their hardware, exactly the same way it was *my* choice to select it 
for my software. I believe in basically *one* freedom: the freedom to make 
our own choices!

But if you actually want to discuss "number of developers" and their 
motications, I actually have another few arguments for you:

 - I just personally think your math is bogus. I think more people think 
   like I do, than people think like you and the FSF does.

But I don't even depend on that. Because:

 - I think that *technical*quality* is more important than *quantity*.

   And I think you have already proven a point: the GPLv3 seems to attract 
   people who make the wrong *technical* decisions.

Put another way: I'll much rather attract one Al Viro to the project, than 
a hundred rabid FSF followers.

See? Because I think that one Al Viro will make *more* of a difference 
than a hundred people who think that their *religion* is more important 
than making the technically correct choice is!

With the GPLv2, you need to give your software modifications back, but the 
GPLv2 never *ever* makes any technical limitations on the end result.

In the GPLv3 world, we have already discussed in this thread how you can 
follow the GPLv3 by making the TECHNICALLY INFERIOR choice of using a ROM 
instead of using a flash device.

Quite frankly, I don't *want* to attract develpers that are not 
technically "up to snuff". And if you think that making the technically 
worse decisions is the "rigth decision", then hey, you're clearly not in 
the same technical quality range as I am, or Al Viro is.

Am I elitist? HELL YES! I think some people are simply *better* engineers 
than other people. I've met my share of outstanding engineers, and I've 
met average engineers. 

I am firmly of the opinion that one of the signs of an outstanding 
engineer is making the right technical choices. The GPLv2 is ok with that. 
The GPLv3 is not. The GPLv3 makes *limits* what you can do from a 
technical angle, in a way that the GPLv2 does not.

The GPLv2 requires that you give source code out. But if you want to make 
your hardware in a way that it only runs signed versions, because of some 
reason like an FCC rule, or banking rule, or just because you damn well 
want, the GPLv2 doesn't stop that.

The GPLv3 doesn't stop it *either*, but the GPLv3 requries that you make 
the INFERIOR TECHNICAL CHOICE.

In other words: the GPLv3 is for people who care more about the opinions 
of the FSF than about the technology.

And why the hell should I trust people like that to make the right 
technical choices in *other* matters? They have already shown themselves 
to make bad technical choices.

> Do you understand now why I feel you haven't answered the 'why'?

Ok, so now I have. I have three *different* and independent answers for 
you:

 a) I don't personally feel like I have the "moral authority" to require
    hardware designers to give access to their hardware to me.

    I can tell them that I *like* open hardware more than closed hardware, 
    but they designed the hardware, and as a result I think it's *their* 
    choice.

    In contrast, I _do_ think I have the moral authority to ask for 
    modifications to the _software_ back. Because they didn't design it, 
    they just improved on it.

 b) I think you're simply wrong in your math. I think more people like the 
    middle-ground and not-frothing-at-the-mouth spirit of "open source" 
    over the religious dogma of "free software".

    I think Linux has pretty much proved my point. Look at Hurd, then look 
    at Linux. Am I *that* much better than the Hurd developers (yes, of 
    course I am, but let's assume not). Or is it just that my approach of 
    being more _pragmatic_ about things rather than talking about those 
    "four freedoms" all the time was just much easier for people to 
    accept?

 c) Even if you're not wrong in the math, I've seen the kind of people who 
    argue for it, and quite frankly, I think they are making bad technical 
    decisions. You arguing for a ROM over a flash is an excellent example. 

    You seem to never even have given a second _thought_ to the fact that 
    you actually advocated what is technically the inferior choice.

See? Those are three totally different reasons why I think the GPLv2 is 
the right license for me, and for the kernel.

			Linus

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

[Alexandre Oliva]

On Jun 18, 2007, "David Schwartz" <davids@webmaster.com> wrote:

>> On Jun 17, 2007, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
>> 
>> >> I don't know any law that requires tivoization.
>> 
>> > In the USSA it is arguable that wireless might need it (if done in
>> > software) for certain properties. (The argument being it must be
>> > tamperproof to random end consumers).
>> 
>> But this is not tivoization.

>> Tivoization is a manufacturer using technical measures to prevent the
>> user from tampering (*) with the device, *while* keeping the ability
>> to tamper with it changes itself.

> You're splitting those hairs might finely.

And I was wrong.  Please see the "mea culpa on the meaning of
tivoization" thread.

> So when you ask whether there's any law that "requirse tivoization",
> you won't accept a law that creates a situation where the only
> practical solution is tivoization?

I guess it amounts to what you mean by "*only* practical solution".

"I can't fit the corresponding sources in this CD, so you won't get
them." is no excuse to disrespect users' freedoms, why should this be
different?

>> Taking it further, do you know whether any such law requires
>> *worldwide* tivoization, as in, applying the restrictions in the law
>> even outside its own jurisdiction?

> A law that requires certaint things be tamper-proof, where engineering
> realities requires that they be controlled by software and the software be
> upgradable (for security reasons and for support of future protocol
> revisions) isn't good enough for you?

"engineering realities" is the weak point of your argument, see above.
Is ROM still software?  Is replaceable ROM still 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

[david]

On Mon, 18 Jun 2007, Alexandre Oliva wrote:

> On Jun 18, 2007, david@lang.hm wrote:
>
>> they want to prevent anyone from modifying the credit card machine to
>> store copies of all the card info locally.
>
> I see.  Thanks for enlightening me.
>
>> you don't really answer this issue. since these boxes are required to
>> be sealed and physically anti-tamper, changing the ROM is not
>> acceptable.
>
> Given the ROM exception in GPLv3, I guess you could seal and
> anti-tamper it as much as you want, and leave the ROM at such a place
> in which it's easily replaceable but with signature checking and all
> such that the user doesn't install ROM that is not authorized by you.

'sealed, but easy to replace ROM containing the programming' is a 
contridiction.

if a local person can easily replace the programming it doesn't meet the 
PCI requirements and therefor you just cannot use GPLv3 code for this sort 
of application.

David Lang

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

[Alexandre Oliva]

On Jun 18, 2007, Linus Torvalds <torvalds@linux-foundation.org> wrote:

> I care about one thing, and one thing only: I care that you respect my 
> choice of license for the projects _I_ started. Nothing more.

I do.  Really.

Once the issue about the spirit of the GPL is (hopefully) settled with
all concerned about it, my job would have been done if it hadn't been
for my having got interested in this other issue:

> I think the GPLv2 is superior to the GPLv3.  That is simply not something 
> you can argue against. You can just say "ok, it's your choice". You can 
> ask me *why*, and I've told you at length, but in the end, it doesn't 
> matter. 

Let me explain why I don't see that you've told me at length why you
consider GPLv2 superior to GPLv3.

1. I asked you why GPLv2 is better, and you said it was because it
promoted giving back in kind.

2. I asked you what you didn't like about GPLv3, and you said it was
Tivoization.

3. Then I argued that, since Tivoization enables tivoizers to remove
some motivation for potential developers (= their customers) to
contribute, you trade the potential contributions of all those users
for the contributions of tivoizers, apparently assuming that all
tivoizers would simply move away from the community, taking their
future contributions away from your community, rather than moving to a
position in which you'd get not only the contributions from the
company itself, but also from all their users.

This last piece of the theorem that proves that GPLv2 is more aligned
with your stated goals than GPLv3 is the one that is missing, and so
far you've dodged that portion entirely.  That's the 'connecting the
dots' that I mentioned earlier.  You haven't even acknowledged its
existence, going back to points 1. and 2. as if they were enough, as
if 3. didn't show a contradiction between them.

Now, it may be that 3. is wrong, or that you think it is wrong.  But
you've never said so, or explained why you think so.  You've simply
disregarded that point entirely.

Do you understand now why I feel you haven't answered the 'why'?

-- 
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]

> > Sure, and you use the hardware to stop me from modifying the
> > Linux on your
> > laptop.

> Do I?  How so?

Any number of ways. For example, you probably don't connect the serial ports
to a device I have access to.

> >> You don't use the software in my laptop.  The laptop is not yours.
> >> You have no claims whatsoever about it.

> > Exactly. And I have no *GPL* claims to my laptop either. The GPL
> > doesn't talk about who owns what hardware and it would be insane for
> > it to do so.  Even though the TiVo hardware is yours, you have no
> > more *GPL* claims to it than you do to someone else's laptop. The
> > GPL does not talk about who owns what hardware.

> This is absolutely correct.

> What it does is impose conditions for whoever wants to distribute the
> software.  And GPLv3 makes it explicit that one such condition is to
> permit the user to install and run modified versions of the program in
> the hardware that ships with the program.

I'm sorry, who is "the user"? Who exactly is supposed to be able to install
and run modified versions? How does the GPLv3 specify who is supposed to be
authorized to do this?

The TiVo control over updating the software is a specific access control
measure. It says, "X is authorized to replace the software on this machine
but Y is not". Now, somebody has to make that decision. It's clearly chaos
if anyone can change the software on any machine.

How exactly does the GPLv3 specify who should and should not be able to
change the software on a particular physical machine?

> A condition that is
> arguably already encoded in the "no further restrictions to the rights
> granted" by the license" and to the requirement for complete
> corresponding source code to accompany the binary.

Except that the "right" to upload the software on some particular piece of
hardware was *never* a right granted by the GPL, nor could it be. That *HAS*
to be a right granted by whatever authority controls the use of that
hardware. It seems utterly nonsensical to argue otherwise.

> That you disagree with it doesn't make you right.

Anyone can disagree over anything. If I'm not right just because people
disagree with me, then nobody is ever right.

It's totally obvious that who gets to install what software on a given piece
of hardware is determined by the person who creates/owns that hardware and
they have to authorize anyone else to change it.

> But that it is within the spirit of the GPL defined by its authors
> (which is all I'm trying to show here), it is.

It is not. The GPL was never about who was allowed to modify the software on
particular pieces of hardware. It was about the lack of *legal* obstacles to
your doing so. It wasn't about *authorization* obstacles imposed on the
creators/owners of hardware. These are night and day different categories.

> > The GPL (at least through version 2) is about