Please tone down the GPL3 paranoia

[Rob Landley]

On Thursday 14 September 2006 8:13 pm, Bruce Perens wrote:
> To make this a more concrete statement, I do not trust HP, IBM, and OSDL's
> board to refrain from entering a settlement that would preserve their
> business and avoid further legal cost but would be to the disadvantage of
> Open Source / Free Software in general.

Because when SCO tried to shake IBM down and IBM could have made SCO go away 
for less money than they spend on catering, IBM immediately caved rather than 
spending tens of millions of dollars making an example out of them.

> That they have not done so yet in existing 
> cases is not from altruism, but is driven by a strategy to deter further
> suits.

I'm not expecting altruism.  I'm expecting enlightened self-interest.  I've 
been getting it from IBM.  I haven't been getting it from the FSF.

I've worked for IBM back on OS/2 and despite that I _still_ trust them more 
than I trust the FSF.

> In contrast, I trust FSF to fight to the death for your GPL rights and
> never compromise.

Cool!  By staying GPLv2, I hope to take advantage of this.

Of course I wouldn't mind if the FSF stopped caring about GPLv2 projects after 
GPLv3 projects ship.  It's only fair; the BusyBox list doesn't provide as 
much support for older releases once we've shipped new ones (after all, it's 
quite likely we already fixed whatever bug you're encountering if your 
version is old enough).

Still, we don't pester people into upgrading from older versions.  We still 
provide the old source tarballs in the downloads/legacy directory.  If 
somebody wants to use 0.60 in a new project, that's their perogative.  (We 
may disagree, but once we've expressed that disagreement we're happy to go 
our separate ways.)

> And I am sure that they can find funding and pro-bono 
> legal resources to do so. Aren't you taking advantage of those resources
> right now?

Not that I'm aware of, no.  (I asked the Software Freedom Law Center if they 
were willing to represent a GPLv2 only project.  They said yes.  I can ask 
them for permission to quote their response publicly, if you'd like.)

> > No, I was talking about IBM, HP, SGI, Sony, Toshiba, Samsung...
> >   
> Rob, I spent a time in Europe fighting those very companies while they
> were lobbying for globaly-enforcible software patents that would likely
> have been used against us. W3C rules do not allow to tell you who I was
> fighnting when W3C had a draft policy that would have put software
> patents in web standards. Don't assume that those folks are willing to
> stand behind you.

I'm not.  I assume the Linux kernel developers have enough political capital 
stored up with them to convince them it's in their best interests to side 
with the kernel developers on matters of immediate and pressing concern.

And I trust them to know what immediate and pressing concern really IS more 
than I trust the FSF.

> >> I think you need to leave the option to progress to GPL3
> >>     
> >
> > It is.  For versions 1.2.2 and earlier.
> In other words, your GPL strategy is that you will discard all work that
> the project does from next week until whenever, if it is necessary to go
> to GPL3. This seems hollow.

My strategy is that if the Linux kernel goes GPLv3, I'll ask them how they 
managed it.  And as long as the Linux kernel is under GPLv2, it is a viable 
license pretty much by definition.

> > The license says "end of terms and conditions", and that's where the 
actual license ends.
> >   
> If you look at my own copyright grant in the file "LICENSE" of older
> busybox releases, it has my copyright statement at the top, and on the
> bottom of the file says "This program suite may be distributed under the
> GNU General Public License." You are changing those terms. This is
> regardless of the text of the GPL, which is a separate document from my
> copyright grant and was indeed not included in the tarball at that time.

I have not removed any copyright statements from the code that I am aware of.  
(If I even trim in-file changelogs, I put them in the SVN commit comment for 
that change.)  And I try to name each contributor of a patch in the 
changelog.  I am _way_ more rigorous about licensing than anybody else on the 
list, which is how this whole topic started.

Are you saying the GPLv2 is not a compatible license to "The GNU General 
Public License"?  (A grant issued in 1999 when GPLv2 was the most recent 
version of the GNU General Public License?)

Let's see, download the oldest version in the legacy directory (0.25)...

mkswap
Copyright 1991 Linus Torvalds

That was GPLv2 only.  Linus made clear in 2000 that he's never licensed any 
code under an "or later" permission grant to license code under GPLv3.

So you're pointing out that three was tained back under your tenure.  Wheee...

And in fact none of this explictly says "or later".  You didn't use the 
recommended boilerplate, your permission grant is vague and the _default_ 
reading would be the version of the license that was then current (I.E. 
GPLv2, only).  At least we _bothered_ to say "or later".

> Now, you may attempt to prove that everything I've written has been
> filtered out over 6 years. But that would be your only excuse for
> changing the terms of my copyright grant without my permission.

Do you even understand the difference between copyright and licensing?  In 
order to transfer a copyright you need a written instrument of conveyance.  
(That's why the FSF has that file folder full of papers with signatures on 
them.)  A License is totally different, it's a permission statement giving 
_access_ to a copyright.  One copyright can have multiple licenses, and you 
only NEED one valid license containing a compatible subset of the terms of 
the two licenses.

Please, go talk to a lawyer before you embarass yourself further.  Over the 
years I've worked with lawyers, worked for lawyers (they pay quite well), 
known lawyers socially, lived with a laywer for several months, and studied 
the law myself.  I don't claim to be an expert but intellectual property is 
an area of special interest for me and I bothered to at least learn the 
basics.

And I repeat, nobody's changing the license terms on old code.  I'm saying 
that the new code I write, and the new code I merge, is GPLv2, and that GPLv2 
is a license that has a compatible subset with "GPLv2 or later", and that 
compatible subset is the GPLv2.  If I combine GPLv2 only code with GPLv2 or 
later code, the result can be distributed under the terms of GPLv2.

Ask a lawyer.  I did.

> >> only for a 
> >> future in which GPL2 may no longer be viable because major license
> >> circumventions are found acceptable by some court and thus gain the
> >> strength of law.
> >>     
> >
> > And you're calling _me_ paranoid?  (Yes you did, check the title.)
> >   
> We would not be developing GPL3 if there wasn't a problem with changes
> in the law and in software that potentially weaken GPL2.

Potentially.

> I've certainly 
> seen people cite Nintendo v. Goloob as an excuse for their GPL
> violations on Busybox.

And how many cases did you lose trying to enforce the license against them?

> > How?  The new DRM provisions are a restriction that GPLv2 didn't require
> But surely would have, if DRM existed back then. The world is changing,
> the license has to keep up.

The old code can still be used under the terms of the GPLv2.  If people found 
a legal way to use GPLv2 as if it were BSD, then they would still be able to 
use the old code that way.  That includes every line of code you ever wrote 
back when you were a programmer.

You're attempting to dictate to me how I should write _new_ code.  I refuse.  
Go fork the project.

I would REALLY have preferred to have this flamewar off-list, but you're the 
one who chose to turn a list you'd never posted on before into your soapbox, 
so...

> Busybox had an independent release on the system where most Linux
> utilities were distributed from when I created it: sunsite.unc.edu (now
> called iBiblio) and had Debian-independent users before Erik came along.
> Busybox it was released before the formation of Lineo and Monte Vista,
> before there was any commercial interest in embedded Linux. Lineo was an
> early embedded linux commercializer and Erik's employer.
> > Erik took what you had lost interest in
> Rather than lose interest, I finished the project and went on to other
> stuff

And you are welcome do dig up your old "finished" version and re-license it as 
GPLv3 any time you feel like it.

> > Have you ever even posted to this list before?
> >   
> There has been no need until now. Erik was doing fine and he and I
> corresponded directly when necessary.

Why didn't you email me off list to express your concerns?

> > But you don't seem to think it's ok to drop GPLv3 and go with GPLv2 only.
> I don't think it's necessary, and I think it's bad strategy.

I disagree.

> > I do not consider that position internally consistent.
> It seems that you are requiring my decision to be idempotent before you
> will consider it to be logically consistent. That's bad logic in that it
> rejects any possible reason for progress.

No, I'm talking about what's legally allowed, and what's morally allowed.  
(Things you seem to be combining here.)  You're insisting that I'm allowed to 
drop GPLv2, but not allowed to drop GPLv3.  This is different from whether or 
not I _should_.

> >   This is aside 
> > from the fact that GPLv3 doesn't even exist yet and GPLv2 is the only 
actual 
> > license the project has ever been under.  That doesn't matter: if it's 
> > morally ok to drop GPLv2 and go to GPLv3 at some point in the future, then 
it 
> > must be ok to drop GPLv3 and stick with GPLv2 now.
> >
> > And thus I stuck by my original decision.
> >
> > Thank you for expressing your opinion.  I did take it into consideration.  
I 
> > disagree.
> >   
> So, you're forcing my hand. You're forcing me to fork the project for
> software that I created. I'd rather this hadn't happened.

Go for it.

Rob
-- 
Never bet against the cheap plastic solution.