Re: Another job for the Gnome Foundation ?



> Sender: foundation-list-admin@gnome.org
> From: Brian Behlendorf <brian@collab.net>
> Date: Mon, 21 Aug 2000 19:10:29 -0700 (PDT)
> To: Russell Steinthal <rms39@columbia.edu>
> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>, foundation-list@gnome.org
> Subject: Re: Another job for the Gnome Foundation ?
> -----
> On Mon, 21 Aug 2000, Russell Steinthal wrote:
> > Since I think that developers should be able to control the future
> > uses of their code, I would oppose a mandatory assignment regime.
> 
> Totally reasonable.  But does the Gnome Foundation exist to protect
> individual developers, or the foundation as a whole?  If part of the point
> of the Foundation is to shield developers from legal liability associated
> with the code, it can't do that on code it doesn't have the copyright on,
> and having third-party code on the site becomes both a liability for the
> project and additional complexity.

Big companies are perfectly capable of defending copyright themselves.

I can see circumstances that individuals might want to assign copyright
to enable enforcement or avoid liability (but if they wrote the code from
scratch, short of brain-damage like the DMCA, there is little they can
be liable for).

> 
> Why complexity?  If code is stored in the Gnome CVS tree and says
> "Copyright Russell Steinthal", if Miguel goes in and modifies the code,
> does it become "Copyright Russell Steinthal, Miguel de Icaza"?  Add in
> fifty more changes by different developers, does it become fifty specific
> names?  At a certain point, was it fifty individuals who wrote the code,
> or collectively the Gnome foundation?  If that's where it ends up, why not
> start there?

Yup.  Not a problem.  There can be many people with copyright.  As to starting
there, see below..

> 
> Additionally, having the ability to modify the license when the need
> arrives helps make very thorny legal situations much easier to handle in
> the long run.  The GPL has not yet been tested in court, and it could be
> that there's something broken in it.  The GPL has a forwards-competibility
> clause, sure, but then only Stallman (or anyone else who can
> authoritatively claim to be publishing a "future version of the GPL") can
> change it. 
> 
> > There's nothing wrong with the GNOME Foundation being able to accept
> > assignments from willing authors, however, and a module maintainer
> > should still be able to require patches to conform to a particular
> > assignment regime before they are integrated.
> 
> I'm in agreement there, but I'd also say there shouldn't be an obligation
> for the Gnome foundation to provide code hosting/development/management
> etc for non-Gnome-assigned code.
> 

You are all missing another important point:

A large company may make code available, under the GPL or X style license,
or other free license, but wish to continue to use it internally without
issues (including under other licenses).

If you force copyright assignment, you won't get such donations.

With the X Window System, by not assigning copyright, the original authors
continued to hold the copyright and be able to do what they wished with
it.  This makes it much easier for companies to donate code.  It also meant
that there was community control of the aggregate, which turned out in
the long run to be important.

All donors were asked to sign was a letter stating what the X consortium
was allowed to distribute, that the copyrights on the code be no more
restrictive than the standard X copyright, and that they had developed
the code.

This does not mean that the Gnome foundation should not be willing to accept
copyrights, but I worry about the single point of failure.  Yes, in theory
the fact that the code is all X or (L)GPL protects against failures, but
in practice, the agony that can happen if the successor organization goes
south is very large: just remember back a few years to the X case.
				- Jim


--
Jim Gettys
Technology and Corporate Development
Compaq Computer Corporation
jg@pa.dec.com






[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]