On Wed, Jun 12, 2002 at 06:25:29PM -0500, Joe Wreschnig wrote: > Since the matter in question boils down to attribution, what if there > were sections that were marked such that if you make any changes, you > cannot use any other author's names without their permission, e.g. the > FSF could mark the GNU Manifesto as such, and if someone changed that > section, they would have to remove the FSF's name from it, or get their > permission? This solves my problems, and is significantly more free than > the FDL or the removal option. Here are my current thoughts on Endorsements: 1) Endorsements would go along with the copyright notice itself, not the license text. This is so that they are somewhat prominent. 2) Distributors are NOT compelled to *retain* endorsements in the copies of the document they distribute. If they want to trim the endorsement list, even to zero, for purposes of space, personal feud, or whatever, they may. 3) Endorsements *must* be removed when a document is modified in any way. Endorsers may wish to communicate to the world (via a Web Page), blanket permission to retain their endorsement under certain circumstances (e.g. "any typographical corrections" or "as long as the chapter entitled Funding Free Software is retained in its original form"). However, such communications are outside the scope of the DFCL. They are arrangements between the distributor and private entities. I.e., nothing in the DFCL will give you permission to retain an endorsement when you modify the document. Only your own understanding with the endorser can do that. 4) Anyone, not just the copyright holder of the document in question, can sign on as an endorser to a version of any DFCL-licensed document. Whether their endorsement is listed is up to the distributor (see 2 above). 5) There will be text, in the form of a brief notice, following the copyright notice, which mentions endorsements. Removing that text will not be permitted, unfortunately (yes, this is invariant text). The reason for this is so that everyone understands whether or not the author(s) and other parties actually approve of the edition of the document being distributed. E.g.: Copyright (C) 2002 Free Software Foundation. This edition of $DOCUMENT is distributed under the Debian Free Content License and is endorsed by the following parties. The absence of an endorsement by any party, including the copyright holder, may indicate that this edition does not meet with that party's approval, and is not representative of that party. And then you either would or would not have the Free Software Foundation listed. I *think* this would address the kind of problem RMS was having with people being able to remove certain sections from, e.g., the GNU Emacs Manual, though of course I have no idea if he would consider this solution "enough". Needless to say, the wording of that boilerplate could certainly use some tweaking. I'd like to make it as short as possible (because it's invariant text, and I hate that stuff, and it can compel people to fill a page with 6-point type) while still making it clear that "If you don't see the author's name below, he may completely hate and totally disagree with this edition of the document". We *could* require that an edition with no endorsements have an alternative notice that says so with scary language, but that introduces complexity and I'd rather not go that road. I like the idea of people being able to produce alternative editions of a DFCL-licensed document solely through deletion. (E.g., delete chapter 2, delete the endorsements, and go to press.[1]) [1] Of course, this is exactly what RMS is afraid of with respect to the GNU Emacs Manual, I think. It may be that RMS is impossible to please in this department, and that he can't bring himself to license the GNU Emacs Manual in a DFSG-free way, period. I say this only because I need to make it clear that I am not on a mission to keep the GNU Emacs Manual in main. I spent a lot of effort on that several months ago and it did not come to fruition. Yes, it may be in woody right now, but if it still has those invariant sections in it, then it is, in my opinion, in main illegitimately. By the way, this is a footnote for a reason. If you want to discuss it, please start a new thread. -- G. Branden Robinson | Debian GNU/Linux | kernel panic -- causal failure branden@debian.org | universe will now reboot http://people.debian.org/~branden/ |
Attachment:
pgprbnAcApcFb.pgp
Description: PGP signature