Subject: Re: "On Virus" -- get real
From: "Stephen J. Turnbull" <>
Date: Tue, 28 Sep 1999 13:53:34 +0900 (JST)

Sorry for following up to my own post, but I found this needs
clarification, and that led to further clarification of my own

>>>>> "rms" == Richard Stallman <> writes:

    rms> A free software package has some of the characteristics of a
    rms> "public good", but it may not precisely fit that category.

>>>>> "sjt" == Stephen J Turnbull <> writes:

    sjt> Stipulated.  I think it's clear why I abused that term in
    sjt> this way....

On second thought, it probably isn't.

    sjt>     As for "the spirit of all parties being equal," you will
    sjt> always be free ..., and I never will.  Not on your GPL code.

    rms> I think you have misunderstood the situation here, because it
    rms> is entirely symmetrical.  [...]

    sjt> No, you have misunderstood the situation.  Where in the GPL

This wasn't entirely fair.  Richard is looking at the global
situation, including all the factors that make up a good license, and
defining "symmetrical" in a way arguably appropriate for that
discussion.  I am only concerned with the freedoms explicit or
implicit in the GPL itself, not in the overall social system in which
it is embedded, and have defined "situation" appropriately for that.
It is that latter sense which Richard may have misunderstood.

    sjt> does it say that I must produce any free intellectual asset
    sjt> to receive a license to use free software?  Nowhere.  There
    sjt> is no axis of symmetry pertaining to "my" code vis-a-vis
    sjt> "his" code in the GPL.

This is where "public good" is important.  Once the "public good" is
produced, there is no distribution cost, and thus no intrinsic barrier
to my use of the code in any way, including resale under arbitrary
conditions.  There is no need to restrict such resale to the original
author; others may be able to do it "better" in some sense![1]

There are many productive ways to add value to GPL code other than
combining it with more code.  It is quite reasonable to suppose that
there can be "giants of free software" who produce important free
software products without ever producing a line of licensable code
(eg, a GNU/Linux distribution, which is at root a "mere aggregation").
Such a "giant" will still be on the short end of the stick vis-a-vis
any GPL author, in the sense described earlier in the thread.

Certainly any product that requires exercise of such rights would not
be "free" in the sense that "Copyleft" advocates mean the word.  But
debating the social value of permitting this is just the BSD vs.
Copyleft argument, and we needn't rehash that.

In the end, this issue about "partnership" and "all parties being
equal" resolves to the difference between Copyleft and non-Copyleft,
and I see use of these terms in the current context as a device by the
Copyleft forces to appropriate these terms and exclude their use in
describing non-Copyleft free licenses (let alone semi-free licenses).
(Remember this thread started out looking for a better term than
"viral", which is used precisely to decribe this difference.)

I think that is bad policy given that the two points of view
supporting the different kinds of license are both firmly held and
philosophically well-supported.

[1]  E.g., Red Hat has done a much better job of "marketing" GNU/Linux 
than Linus has done, in a narrow economic sense.  Of course giving
broader rights to relicense the code removes the "Copyleft" incentive
to license future code under free terms, and incidentally violates the
author's residual property rights, if you believe in such for
intellectual assets.  That's beside the current point.

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."