ESR: "We Don't Need the GPL Anymore"

[Michael K. Edwards]

On 7/12/05, Rick Moen <rick@linuxmafia.com> wrote:
> Quoting Michael K. Edwards (m.k.edwards@gmail.com):
> > Precisely.  And in fact the GPL is an excellent solution to the
> > problem that it originally was designed to solve -- assuring the
> > recipient of the code that he can modify it to fit his needs, and pass
> > it (modified or not) to others with similar needs, while assuring the
> > upstream author that her code base, enhanced or ported by an outside
> > vendor, will not become a tool for vendor lock-in against her own
> > organization.
> 
> Lest we forget:  It also, as part of a multi-licensing model, can be a
> marketing tool for selling a proprietary-licensed instance of the
> codebase:  Customers can see the merit of using and building on the GPL
> (or other copyleft) instance, and thereby be motivated to pay for the
> proprietary one to permit third-party uses that the copyleft one doesn't
> allow.

I don't agree with that "doesn't allow" part based on the actual terms
of the GPL and applicable law, which is part of why I say that
propaganda and intimidation uses of the GPL have long overshadowed its
meaning as a legal instrument.  The use you describe is known in the
literature as "viral marketing" (I believe the term predates the use
of "viral" to refer to the GPL), and is the answer to at least one of
the why-does-X-publish-under-the-GPL questions that I posed to the
reader.  Sometimes it's also referred to as "the second hit will cost
you", and is something less than socially revolutionary.

> > People outside the industry -- and as far as I am concerned that
> > includes ESR, who does not appear ever to have run a software or
> > services business -- are unfamiliar with the actual economics of any
> > of these situations, and in any case tend to think exclusively in
> > terms of the consumer software sector.  For ESR to pontificate about
> > the superior economics of the GPL, when his interviewer knows more
> > about modern developments with respect to the GPL and trademarks and
> > patents than he does, is pretty funny.
> 
> I'm probably not competent to speak to software/services microeconomics,
> but would like to mention, in case it's relevant, that Eric _does_ have
> a demonstrated excellent grasp of licence mechanics and applicable law
> (including trademark and patents).  Please see, for example, his and
> Catherine Olanich Raymond's Licensing HOWTO
> (http://www.catb.org/~esr/Licensing-HOWTO.html).

To my eye, that document indeed appears on first reading to reflect a
firmer grasp on the applicable law than most commentators can claim. 
For one thing, he actually identifies a few of the relevant appellate
precedents, and paraphrases them more or less correctly (according to
my equally non-lawyer and uncredentialed understanding).  I disagree
with him on a number of points, but at least I can say with a straight
face that I _respectfully_ disagree with him.  :-)  However, that says
little or nothing about his sophistication about the economics, on
which the secondary evidence (such as his ignorance of Red Hat's
trademark games prior to the interview) casts some doubt.

> I'm disinclined to play credentialism.  Whether this has anything to do
> with my lack of credentials, you be the judge.  ;->

I'll contest that lack of credentials with you.  ;-)

> > Andrew Tridgell knew what he was doing, as did Larry McVoy.  There is
> > considerable prior art on vendor lock-in and its discontents (some of
> > it created by Andrew and Larry themselves), and I didn't see anything
> > particularly insightful in that part of ESR's comments.
> 
> He's one of very few who have clearly and unambiguously stated that
> Tridge did nothing wrong, and explained why in a fashion that could be
> understood by any layman.  I find that valuable.

What do you mean by "wrong" in this context?  As far as I'm concerned
none of McVoy's or Tridgell's conduct skirted any particularly
important ethical or legal boundary, and it all had more to do with PR
maneuvering and deliberate provocation of a crisis that was coming
anyway than with any particular engineering outcome.  IIRC the
extraction of BitKeeper metadata was independently reverse engineered
at various times by several people on LKML, and I don't think Linus
wound up using Tridge's work in the course of migrating to git. 
Everyone involved got exactly what they paid for, in one currency or
another.

> > If you ask me, if he wants to retain celebrity status (which has, I
> > think, been of some economic use to him), he's going to have to work a
> > little harder.
> 
> "You got money off being famous" objections have always struck me as
> more than a little facile, not to mention the supposition that Eric's
> obliged to work at making you or me happy.

Oh, I have no objection to that at all; I'm just commenting on his
claim that the best use he can think of for notoriety is to point out
unpopular truths.  Perhaps he considers that an excellent use, as have
other celebrities (Albert Einstein comes to mind); but it's hardly the
principal use that ESR has made of it.  :-)

Cheers,
- Michael