On 2003/10/26 21:33, Gilad Ben-Yossef wrote:
On Sunday 26 October 2003 20:15, Eran Tromer wrote:The distiction is anything but simple. [snip]My answer was given in the form of two separate paragraphs and such a choice of lexical structure usually denotes two separate subjects are discussed. Indeed, such was the case with my answer but you seemed to have missed this completely.
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.
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):
-------
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.
-------
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.
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.
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.
Eran
================================================================= 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]
