Skip site navigation (1)Skip section navigation (2)
Date:      17 Dec 2001 03:33:41 -0800
From:      swear@blarg.net (Gary W. Swearingen)
To:        Greg Lehey <grog@FreeBSD.ORG>
Cc:        Brett Glass <brett@lariat.org>, chat@FreeBSD.ORG
Subject:   Re: IBM's intentions with JFS (was: IBM suing (was: RMS Suing was [SUGGESTION] - JFS for FreeBSD))
Message-ID:  <qqheqq2i96.eqq@localhost.localdomain>
In-Reply-To: <20011217163427.A2885@monorchid.lemis.com>
References:  <3C1875D6.5DE4F996@mindspring.com> <20011213051012.Y56723-100000@turtle.looksharp.net> <3C186381.6AB07090@yahoo.com> <3C1875D6.5DE4F996@mindspring.com> <3C186381.6AB07090@yahoo.com> <20011214122837.O3448@monorchid.lemis.com> <3C19807D.C441F084@mindspring.com> <4.3.2.7.2.20011214175450.02da2a90@localhost> <4.3.2.7.2.20011215232233.00e74cc0@localhost> <4.3.2.7.2.20011216221810.031b6820@localhost> <20011217163427.A2885@monorchid.lemis.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Greg Lehey <grog@FreeBSD.ORG> writes:
> On Sunday, 16 December 2001 at 22:21:50 -0700, Brett Glass wrote:
> > At 12:18 AM 12/16/2001, Greg Lehey wrote:
> >>
> >>  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.

On the other hand (from the GPL):

    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.

Back on the first hand (from the GPL):

    In addition, mere aggregation of another work not based on the
    Program with the Program (or with a work based on the Program) on a
    volume of a storage or distribution medium does not bring the other
    work under the scope of this License.

These is clear as mud, which gives a licensor a good basis for taking
whichever side of the issue is important to him in a situation.  Who can
predict how a court would read this?  One can only ask the license if
he'd sue you (and hope he doesn't) or get a written clarification that
becomes a part of the contract (so it isn't the GPL).  More below.

FUD is very appropriate here, no matter how clear cut RMS makes it
sound.  Remember, this is the same guy who recommended using the GPL
(on CORBA interface code) in the same message he said he knew it wasn't
enforcible ("So the real question is, what result do you want?").

> > If they're part of the kernel, they're not separate works. RMS would
> > have the right to demand, TODAY, that the entire FreeBSD kernel be
> > licensed under the GPL. This is the danger of permitting the camel's
> > nose into the tent.

Well, IBM could; by threating suit to stop JFS distribution (silly since
it's GPLed) and collect damages and penalty (likely enough to make
people settle it IBM's way).  Except I wonder who they sue?  One can't
tell from freebsd.org who owns the copyrights on FreeBSD.  Is it WR,
one of several organizations of questionable legal status, or a cast of
hundreds?

> Well, why don't we ask him?

This flabbergasts me.  Can it BE that this issue hasn't been hashed out
long ago and the licensing policies well established among journeymen?

> On Sunday, 16 December 2001 at 22:50:40 -0700, Richard Stallman wrote:
> > When code is linked together, that is not "mere aggregation"; that is
> > making one program.  If you link some GPL-covered code into the
> > kernel, the GPL's conditions will apply to the kernel as a whole.

It's debatable whether linking is "mere aggregation".  We can't even be
sure what "mere aggregation" means.  Almost all aggregations form a new
work with it's own copyright, but maybe it could be interpretted as one
involving no creative authorship, like a phone number list.  Does
linking involve authorship?  The person linking or the makefile author?

Also note that all aggregations make one program in the terms of the
GPL which (in section 0) defines "program" as the work covered by the
GPL; it doesn't even have to contain executable or compilable code.
(This conforms with most legal language I've seen.  Law doesn't seem to
much care how works are used.)  If it's an aggregation and the BSDL part
is no not based on the GPL part, then contamination is not a problem
(if it weren't for the contradictory clauses, anyway).

Also, that "mere" is a mere editorial comment and doesn't effect the
meaning of the license clause; the word can be ignored.  This reflects
on the reasoning behind RMS's last sentence above, but doesn't mean that
it isn't still true.  We still need to interpret that quoted mess above.

> > I don't think that results in any legal difficulty.  The FreeBSD
> > kernel uses the revised BSD license, right?  That is compatible with
> > the GPL.  So you can link these things together.  The kernel code
> > released under the revised BSD license will continue to be under the
> > revised BSD license; it is only the *combination as a whole* that will
> > be covered by the GPL--if and when the GPL-covered code is included in
> > it.

Terry says the kernel has some non-GPL-compatible parts.

Even if it didn't, the GPL quote above says that the both parts (the
whole) must be distributed under the terms of the GPL.  This requires
a change of licensing (even if just to dual-license) on the kernel.
The GPL makes it a condition of the license that the kernel be licensed
for distribution under the terms of the GPL. (It doesn't prevent it from
also being licensed under the BSDL.)  But the important point is that it
requires a change of license, and only the copyright owners may do that.

IS there a single copyright owner who can do that?  Do the copyright
owners want the kernel to be dual-licensed?  It's a slippery slope.

> I interpret this to mean "after linking".  It would appear to be the
> kernel binary which falls under the GPL.  About the only obligation of
> the FreeBSD project would be to make the corresponding source code
> available.

But you could also be said to be distributing a work which consists of
the source code of both parts as a whole.  Maybe the ports scheme gets
around that, but then maybe the lawyers will call that a "work-around"
of the license and judge it the same as a statically linked program.

> > If someone links a kernel without that GPL-covered code, the GPL
> > won't apply to that kernel.
> >
> > The main consequence, legally, of including some GPL-covered code
> > would be that you could not *also* link in other code with
> > GPL-incompatible licenses.

Since the kernel has some incompatible code, that's a problem.
Actually, I consider it a good thing.  I don't want to see a GPL kernel
with a lot of BSDL code in it, which is what we're looking at here.

I believe that the aggregation clause allows even static linking without
contamination since that can be viewed as just a means of storing two
separate works of authorship which are not based on the other (one
doesn't contain parts of the other (as derivative is defined in the
GPL)) and it is a "mere" aggregation, etc.  But I wouldn't risk running
with an idea not shared by the licensors without some deep pockets.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-chat" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?qqheqq2i96.eqq>