Subject: Re: Standing on multi-authored works
From: "Stephen J. Turnbull" <>
Date: Wed, 20 Apr 2005 17:06:59 +0900

This is nowhere near as hard as determining ownership of sandbars in
the Mississippi River.  :-)

Executive summary: the issues you identify are merely the problems of
distributed title in software.  It is just the way everybody lives in
the proprietary world.  Free software holds out the promise of cutting
these costs dramatically, but it can't do that unless you contribute
all of Bar, not just foo().

>>>>> "Ralph" == Ralph Corderoy <> writes:

    Ralph> This makes it awkward for me to pursue infringement in the
    Ralph> same way me retaining foo()'s copyright makes it awkward
    Ralph> for FSF with ping.

Not the same.  The FSF (a) wants all software to be free and (b) needs
to maintain a reputation for protecting its software.  So the FSF
wants title to foo() for its own sake, not just to protect ping.  And
if the FSF fails to go after a foo() infringement, that looks bad
since foo() is part of ping.

But in your case, going after foo() infringements is not a social
responsibility you have assumed; it's a privilege granted by law.
Once you have assigned copyright to the FSF, you have neither
responsibility nor privilege over foo().  (I guess it would be nice if
you notify the FSF of any infringements you observe.)

Nor does it impair your ownership of the-rest-of-Bar, as far as I can
see.  The FSF explicitly disclaims any interest in code you write that
you don't assign to the FSF and doesn't incorporate foo() (both in the
GPL and in the assignment form I signed).  The-rest-of-Bar is not a
derivative of foo().

The FSF does make a claim that distributing the-rest-of-Bar by itself
is equivalent to distributing Bar.  (I find this ethically rather
dubious and logically confusing, but evidently it's good law.)  So
they claim an interest in your _distribution_ of the-rest-of-Bar =>
distribution of Bar, because Bar is a derivative of foo().  But even
if the FSF may interfere with your distribution of the-rest-of-Bar[1],
you can still enforce your copyright in the-rest-of-Bar.

    Ralph> Also, when someone wants to buy Bar, Inc.  from me they may
    Ralph> not like that I don't hold all of Bar's copyright.

Sure.  Again, so what?  This is no different from you using any third-
party library in Bar, except that (1) the standard assignment gives
you a _very_ broad, permanent, non-exclusive, sublicensable license[2]
to foo(), and (2) both you and Bar, Inc.'s wannabe owner have
excellent assurance that foo() and its source will be available on
reasonably favorable terms forever.

[1]  Actually, they can't interfere with distribution under the
standard assignment, because of the "grant-back" clause.

[2]  I don't know whether this grant-back clause is _effective_; IANAL
and I've not yet seen discussion that looked authoritative on this
point.  However, the broad interpretation seems to be the clear intent
of that clause.

School of Systems and Information Engineering
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.