On Sunday 26 October 2003 22:23, Eran Tromer wrote:
> My first sentance ("The distiction is anything but simple.") refers
> to your first paragraph. The rest of my reply refers to your second
> paragraph. Indeed, I neglected to employ appropriate lexical
> constructs and quoting conventions, leading to the present
> misunderstanding.
Whatever :-)
>
> I must insist, however, that the definition of "derivative work",
> though indeed external to the GPL, is far from trivial in our case.
> Moreover, the GPL further muddies the water in its Section 2
> paragraph 5 (not paragraph 4 as I said earlier; that was an
> off-by-one):
Ok, maybe I should have said: as simple, or complicated, as the
distinction between derived work in any other case of copyright law :-)
>
> -------
> These requirements apply to the modified work as a whole. If
> identifiable sections of that work are not derived from the Program,
> and can be reasonably considered independent and separate works in
> themselves, then this License, and its terms, do not apply to those
> sections when you distribute them as separate works. But when you
> distribute the same sections as part of a whole which is a work based
> on the Program, the distribution of the whole must be on the terms of
> this License, whose permissions for other licensees extend to the
> entire whole, and thus to each and every part regardless of who wrote
> it. -------
>
There is nothing at all muddy about this. Copyright law has the concept
of aggregated work and collections as seperate and licensable works in
their own right apart (or in addition to) the license of their
components. Maddona has the copyright on "Song X", Big Puff Daddy is
the copyright holder of "Song Y" and Hed Artzi are the copyright
holders of the *collection* "Top Hitz 5".
> You could argue that the above is already implied by copyright law
> and its effect is thus null; however, I strongly doubt that holds for
> every country on the globe, and surely subtle differences of phrase
> can be put to great employ by those so inclined.
The GPL specifically states that it is governed by laws of Msc. It is
true though that whether or not you are bound by it or not depends on
the specific copyright laws of your locality - if you are required by
the local law to get a license for the work in question under locality
laws then the GPL is moot because there is nothing proving you have
accepted it's terms. The normal idea being that only by accepting the
GPL contract terms you can get the license to distribute the work in
question - if you can get the premissions to do so via other channles
(for example if your local copyright low does not recognise copyrights
of foreign nationals as the original Amercian copyright law was) then
ypu're off the hook.
>
> > However - since, as you mentioned, the original composer of the
> > copyright license publicly stated that they accept these heuristics
> > to be correct, at least in the general case, we can just as well
> > treat them as correct in the sense that anything that the FSF
> > considers to NOT be a derived work, isn't.
>
> Why should this holds when the copyright owners is not the FSF?
> By using the GPL the author does not grant the FSF any special status
> other than the ability to (formally!) issue new versions of the GPL.
> It could be argued that the ability to revise implies the ability to
> force an interpretation; this may have merit in some jursdictions,
> but is again far from simple.
As Shachar akready said - s/FSF/Copyright holder/
>
> > And please don't bring that stupid myth that the GPL is not 100%
> > "enforceable" because it was not tested in a courtroom - the fact
> > that it never GOT to a courtroom despite numerous incidents with
> > big multinational companies with loads of cash and hordes of
> > lawyers who mistakingly (or not) violated the license and settled
> > with the FSF (the single exception where the GPL got to a courtroom
> > is the MySQL case, where AFAIK the issue of whether the GPL holds
> > was never raised) speaks more loudly then a thousand courts - the
> > GPL is as "enforceable" as any contract can be.
>
> I didn't intend to bring up that myth (which, truth be said, doesn't
> seem very relevant to the discussion). Accepting your rebuttal,
> however, re-raises the opposite issue. Recall the hazy language used
> in the GPL, and consider -- if hordes of lawyers had so little doubt
> that linking is covered, then someone with lesser resources will have
> a hard time defending his use of (say) a pipe interface to a GPL
> program whose copyrights are owned by a non-FSF zealot.
The law uses the term of "common practice". What the FSF says about the
GPL is common practice, so unless you specifically and publicly stated
otherwise (see Linus et al) everyone assume that the common practice
holds.
Gilad
--
Gilad Ben-Yossef <[EMAIL PROTECTED]>
http://benyossef.com
=================================================================
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word "unsubscribe" in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]
