Python 2.0b1 is released!

Gareth McCaughan Gareth.McCaughan at pobox.com
Fri Sep 8 18:17:52 EDT 2000


As I understand it, the situation is this.

  1 CNRI wants the applicable law specified, to make sure
    they don't get sued in unpredictable ways.

  2 RMS wants the applicable law not specified, because
    as it stands it would have to apply to the GPL as
    well, and that's (allegedly) something that can't
    be done.

So what's wrong with these possibilities?

  a Amend the licence so that it says "The disclaimer
    of liability in section 1.2.3.4 shall be interpreted
    according to the law of the state of Virginia" or
    whatever.

  b Amend the licence so that it says "This license[1]
    shall be interpreted according to the law of the
    state of Virginia, except that this shall not apply
    to the provisions of the GNU General Public Licence
    unless it would do so in the absence of this paragraph."

  c Some other intermediate stipulation, like "Paragraphs
    1 through 12 of this license shall be interpreted
    according to the law of the state of Virginia" where
    those are in fact all the paragraphs of the licence
    (the point being to make it more explicit that there's
    no attempt to have the GPL interpreted under Virginian
    law).

The wording in each case is certainly broken; I'm
not a lawyer. But I think the point stands, namely
that the stuff CNRI want to be protected (by being
interpreted under Virginian law) and the stuff RMS
wants to be protected (by not having whose law it's
to be interpreted under spelt out) are disjoint.
So (he said, naively) it ought to be possible to
protect both at once by being careful about exactly
what the specification applies to and what it doesn't.

Now, the people involved in this are not stupid, so
presumably there's a good reason why this sort of thing
can't be made to work. Can someone explain what that
good reason is?


[1] I'd prefer "This licence", but one must be realistic. :-)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



More information about the Python-list mailing list