Subject: Re: A few here may have an opinion on this
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Fri, 25 Oct 2002 11:38:44 +0900

>>>>> "Don" == Don Marti <dmarti@zgp.org> writes:

[Aside:  Don't you just love the way GPL advocates always end up
talking about _failing_ to distribute software as if that were a good
thing?  I know, "omelets and eggs", but it is striking.]

    Don> What proprietary derived works?  The GPL at least puts a kill
    Don> switch on distributing the software

Not if the patent and copyright holders are the same entity, which is
what I was thinking about.

So you're thinking about third party patents in derivative works?
Then you're not talking about _preserving_ freedom of the
government-funded software, you're talking about _extending_ freedom
to the independently-funded, independently-developed, patented
feature.  In this case the usual GPL advocate's plea that "it's based
on my idea, so it essentially _is_ my idea that I'm protecting" won't
fly; if that's true, the funded software constitutes prior art and no
patent should be issued.

The usual university license advocate's argument that the proprietary
derivative doesn't affect the freedom of the government-funded ideas
seems to be unassailable in this case.  So while a private entity has
every right to put that condition on its/his/her creation, I don't
think that (in a country containing both GNU GPL advocates and
"university" license advocates) the government should require it.

I conclude that you want the GPL here because it's viral and extends
freedom where it otherwise wouldn't go.  The patent is just an excuse
you can use to confuse the university license advocates.  I think
that's bad government policy, but if you want to argue it, be my
guest.  Just argue it on the merits---"Resolved: that the government
should aggressively take measures to ensure that government-funded
software is not neutral toward third party developers, but rather
strongly encourages them to use free licenses."

    Don> until the patent mess is straightened out.

There's no "mess" here that you can't "straighten out" simply by using
a university license with an appropriate research contract, unless you
really do want to restrict the freedom of independent developers.  You
do, and for good reasons, I'm sure.  I do not think those reasons can
be sufficient for the U.S. government to impose such restrictions.

If you're talking about the fact that the whole patent system is
crooked and inefficient, then _please_ fix or abolish the patent
system; don't burden the world with more contradictory regulations.

The kind of half-baked remedy you are proposing will not affect purely
proprietary organizations.  It will force many neither-fish-nor-fowl
businesses to make a choice between ugly alternatives: retreat to the
all-proprietary stance, or to become dependent on government handouts,
losing all the discipline and value-measurement benefits that come
from interacting with markets.

    Don> University "technology transfer" departments can snap
    Don> something up and get it into the patent process before
    Don> publication or code release.

Of course they can.  But then they are both patent and copyright
owner, and the GPL terms don't apply to their distribution activities.
They are bound only by their contract with the government (which is
not a license because the government doesn't own IP).  Once again, fix
the contract to deal with patents directly, don't put conditions on
the copyright license in order to deal with patents.

You like the GPL because you like the GPL.  That's fine.  But it
doesn't have inherent advantages here, in the sense of preserving
university license-style freedom better than a university-style
license could.

Advocate the GPL _because_ it's viral.  That's the whole point!


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py