Subject: Re: Standing on multi-authored works
From: Ralph Corderoy <>
Date: Tue, 19 Apr 2005 17:27:50 +0100


Off-topic, but hopefully not uninteresting.

DV Henkel-Wallace <> wrote:
> On 18 Apr 2005, at 09:14, Marshall W. Van Alstyne wrote:
> > Interesting point, which sounds correct.  It suggests that a good
> > property of F/OSS licensing, in any form, might be to grant
> > (non-exclusive) standing to the original (c) holder or the
> > institution, such as FSF, responsible for managing the software.
> Why be so complicated?  Just assign it to them.  The assignment for
> (or the blanket assignment form, which is just a modification I made
> to the regular one) allows you to "take back" a copy of what you
> assigned and do something else with it.

That assignment form seems more tricky than the first few reads apply
AFAICS.  Imagine

    I write foo() in project Bar in 1990.
    foo() is large enough, and expressive enough, to be copyrightable.
    I see foo() is useful to GNU ping and contribute it in 2005.
    FSF send copyright assignment form for contributions to ping.

The FSF want to be the sole copyright holders for GNU ping so they can
pursue infringement.  Fine.  I'm happy to sign over copyright on the
foo() I've contributed to ping.

However, if Bar is taken to be a derivative work of foo() then Bar is a
work "based on the Works".  Consequently, I'm no longer the sole
copyright holder on Bar, I hold the copyright on Bar less foo() and FSF
hold copyright over Bar's foo().  This makes it awkward for me to pursue
infringement in the same way me retaining foo()'s copyright makes it
awkward for FSF with ping.  Also, when someone wants to buy Bar, Inc.
from me they may not like that I don't hold all of Bar's copyright.

That Bar existed 15 years before foo's contribution to ping, or the
existence of ping, doesn't matter since apparently "timing is irrelevant
for a determination of whether one work is a "work based on the work.""

I doubt most people signing the form realise the implications, assuming
I've got it right.  (I have been getting some questions answered by the

Lastly, given the assignment, once signed, is in effect the instant I
create a copyrightable work, if I put other bits of Bar into my own
private copy of ping it's covered by the assignment, there's no act of
contribution, e.g. email a patch, required.  So I have to have a very
good memory and not put anything into ping without careful
consideration.  In fact, the easiest solution is to give up contributing
anything in case I make a slip up.  :-(

Could a `one off' assignment form be done, and could the paper-shuffling
overhead to the FSF this would cause be lessened by some kind of
automatic patch-signing process?