On Mon, May 28, 2001 at 07:20:22PM -0600, John Galt wrote:
> Having pointed out the compromise route, let me now go into why I think
> that "best effort" is not that onerous. Basically, making a one-off best
> effort to get the changes into the hands of one person is less onerous
> than providing changes to all comers for three years (GPL clause 3b). To
> be more exact, the license may try to restrict your use, but I believe
> that that is impossible _prima facie_: that implies some "click-wrap" or
> "shrink-wrap" license, and under the present law, that is just not
> feasable. Therefore this license must only deal with copying. Since it
> only deals with copying, this means you may only make copies of the
> program available if you try to provide the changes upstream. If you
> can't provide the changes upstream because of political reasons, what the
> HELL are you doing making them available to the general public?
This misses an important point, I think. The GPL merely requires
giving source to those who recieved binaries. There are, however,
situations where you would want to give neither source nor binaries to
the upstream author. For example, Red Hat modifies gcc to work on a
processor, under an NDA. The modified compiler is still under the
GPL, and anyone who has binaries can get source. But the only people
who have binaries are Red Hat and the client, and they sure as hell
aren't giving either to anyone else, least of all the FSF.
However, the MIT Scheme license requires making a best effort to send
the changes back to the author, something that would clearly be a
dealbreaker for this client.
sam th --- sam@uchicago.edu --- http://www.abisource.com/~sam/
OpenPGP Key: CABD33FC --- http://samth.dyndns.org/key
DeCSS: http://samth.dyndns.org/decss
Attachment:
pgpAPWICAO_ZQ.pgp
Description: PGP signature