Subject: Re: [was Cygnus] now Ghostscript and Aladdin
From: "Stephen J. Turnbull" <>
Date: Wed, 23 Jun 1999 16:07:48 +0900 (JST)

The real issue is not what you or I would do, except as examples.  (I
shouldn't have phrased things as though I were trying to change your
personal policy; I wasn't.)  The issues are first, do more people feel
one way or the other, and how strong is the bias?  Second (probably
more important), which type of person contributes more good code, and
how much more?  And finally, does the bias matter in the context of a
particular project?

>>>>> "Ian" == Ian Lance Taylor <> writes:

    Ian> If I build a program from source, and it has a bug, and it is
    Ian> a GPL or otherwise free program, particularly if it is from
    Ian> somebody I like and respect, I may spend the time to track
    Ian> down the bug.

For me personally, the "somebody I like and respect" is a lot more
important than the license.  Could be hard to separate, I guess.

    Ian>    From: "Stephen J. Turnbull" <>
    Ian>    Date: Wed, 23 Jun 1999 13:41:47 +0900 (JST)

    The author always has special rights to the code you do
    not have.  The author can always relicense under proprietary

    Ian> Theoretically, yes.  In practice, it's hardly worth worrying
    Ian> about.  I don't know of any case where a GPL program was
    Ian> taken proprietary by the author.

Ghostscript.  Taken "semi-free," anyway.

    Ian> I'm not inclined to worry about hypotheticals.

Sure you are.  That Peter would not publish your code under the
Aladdin license, or the GPL if you chose, is very much a hypothetical,
indeed, a counterfactual.

I would in fact worry about GPL programs being taken proprietary, as a
hypothetical; in theory, it is the most effective tool (except perhaps
for the Aladdin license) I can think of for encouraging user support
for a loss-leader program while protecting a proprietary advantage in
the "professional" version by starving the potential competitor of easy
revenues.  (Licenses asserting a property right in contributed code
seem to be not very effective.)  If as many of us argue open source is
in fact an effective development technology independent of its ethical
basis, I would expect "closed-source" developers to open much of their
code base, while perhaps retaining their flagship product with a
special algorithm or something as proprietary.

    Ian> I don't see it.  I see your footnote about the FSF
    Ian> assignment, but to me the assignment says the code will
    Ian> always be free for some definition of free.  In the binutils
    Ian> assignment form, the paragraph in question is this:

Omitted, looks familiar.

"Free?"  I don't see it.  It says that it will be redistributable
under the same terms that the FSF distributes.  It's not even clear to
me that the FSF is required by that language to let third parties
_compile_ the machine-readable source code (which it must distribute),
as long as everybody is allowed to redistribute machine-readable
source code that nobody is allowed to compile.  There is no definition
of "permitted use", and as long as the FSF needs the freedom to revise
the GPL, I suspect there cannot be a legally binding definition there.

I think it is plausible, although unlikely, that a future FSF board
could adopt an Aladdin-like license, and use the assignments to change
the entire corpus of FSF-assigned GNU software to AFPL.  I see nothing
in the legal language to prevent it.  The AFPL has the viral property,
and as far as I can tell that's all that para 4 requires, and only
para 4 (and 5, but that's just "machine-readability") puts
restrictions on the FSF's distribution.  And the AFPL might be
attractive to people who believe in "charity software", as long as the 
assignee was not in a position to profit from it---which the FSF is
not (by para 4).

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."