Subject: Re: New angle on the patent problem
From: "Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp>
Date: Tue, 14 Sep 1999 14:12:40 +0900 (JST)

>>>>> "shapj" == shapj  <shapj@us.ibm.com> writes:

>>>>> "me" == "Stephen Turnbull" <turnbull@sk.tsukuba.ac.jp> writes:

    me> No.  The "patent-poison-pill" license is only good if it is (a)
    me> viral and (b) widespread.  That combination makes experiments
    me> essentially impossible: people who adopt this kind of license
    me> are betting the future of their software.

I would like to reiterate my point that there is no way to "try out"
this clause with respect to a given piece of software, unless you
follow the FSF path and insist that all copyrights relating to a
"poison-patent-licensed" Work be assigned to a single agent (and even
then, it's not clear to me that individual authors lose their own
right to enforce under this clause when they assign the copyright; I
think it could very plausibly be argued that changing that clause can
only be done with the permission of all authors, although the current
FSF assignment papers do give very broad powers to the FSF).  It's as
permanent as the GPL itself once many hands have touched the Work.

Of course, you could try it out with Linux and see how it goes, then
apply to Emacs and gcc if it works.  But I don't see how Linux could
ever revert.  I'd have to hear a lawyer's opinion about FSF-
administered copyrights, maybe the FSF could revert gcc.  Emacs might
be difficult, as the FSF's legal situation with Mule is complicated.

    shapj> The *current* state of affairs is that GPL'd software (I'll
    shapj> use GPL here, but I think any widely used viral license
    shapj> --free or not -- will do) can be killed by a patent.  There
    shapj> exists, at present, no protection against this.

This is true of all software.  It applies to graduate students doing
research (although I doubt they'd get caught), to two hackers in a
loft above a garage, and to Microsoft.  There is no defense for any of 
them.  Only the possibility of bargaining with the patent holder.

The two differences between free software and (some) others are that
(a) there are no free software developers with stocks of patents to
barter with others, and (b) current business models don't permit
extracting license fees from most users.[1]  These make the strategic
position for OSS advocates difficult, but that's part of our moral
superiority, is it not?

    shapj> In practice, those companies engaging in patent extortion
    shapj> (er, um, I mean "enforcement" of course) do not appear to
    shapj> be free software companies. [...]

    shapj> Adding the "patent poison pill" provides some protection,
    shapj> but as far as I can see has no downside to the producers of
    shapj> free software.

Aren't you contradicting yourself here?  AFAICT the "some protection"
and the "no downside" are actually of the same order of magnitude.

Anyway, for starters, a Work covered by such a license is _not free
software_ any more.  IMHO.  There are now conditions---completely
unrelated to redistribution of the Work itself, which have no direct
relationship to any of the "four freedoms" of the Work itself---
imposed on usage.  (This is not true of the Jikes license, etc; the
poison pill is triggered only by action against the Work itself,
AFAIK.)

And the remedy is imposed, if at all, mostly on people who are not
directly relevant to the offensive behavior, and in fact may never
have heard of the Work in question.  This is clearly in violation of
the definition of free software:

"In order for these freedoms to be real, they must be irrevocable as
long as you do nothing wrong; if the developer of the software has the
power to revoke the license, even though you have not given cause, the
software is not free."  (http://www.gnu.org/philosophy/free-sw.html)

How many _programmers_ (who may very well be independently using
Linux, gcc, or Emacs in their daily work; I'm not talking about say
Sony where the official Playstation development platform is Linux, I'm
talking about _me_ using Linux at my desk, surrounded by the Madding
Crowd of Windose users) at Unisys (or wherever the LZW patent
currently resides) agree with the enforcement of that patent?

How about the students at a university which takes it upon itself to
enforce a software patent against an OSS project?[2]  Can they use
Linux to do homework?

OK, it took a while, but now I've figured out what I don't like about
this proposal.


Footnotes: 
[1]  At least, I would assume that a license clause, applied to soemthing
up for grabs on a public FTP site, that says "this software is
provided as is under the condition that the user must negotiate a
license fee with XYZ Corp, the holder of patent 23456789 which applies
to the frobbaz algorithm in the foo-processing routine" would be
laughed at by a court, and the distributor (OSS developer) would be
held liable for the relevant fees (at a minimum, cease and desist from
distribution).

[2]  Of course, this is really the only way for this clause to have
much effect.  But I want people to be aware of just who is likely to
be heavily represented in the "collateral damage" in such an action.
This is nuclear war, not a Stealth bomber attack.


- 
University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
__________________________________________________________________________
__________________________________________________________________________
What are those two straight lines for?  "Free software rules."