On Wed, 22 Jun 2005 09:58:07 -0700 (PDT), "Mark Rafn" <[EMAIL PROTECTED]> said: > Before going too far down this road, see > http://lists.debian.org/debian-legal/2003/03/msg00805.html > for some fundamental questions about where to draw the line on this > requirement. Unless you're willing to say "all changes must be > published, > even if never released", it's gonna be tough to find a free way to > require > that any publication in the absence of distribution can be required. >
Whoops, forgot to respond to that URL. My opinion on it is that /use/ rather than /distribution/ is what should be considered, so: a-d should not require release of source, because Joe is the user of the software, not the customer. e is sort of on the line, but basically over it. The customer is now the user and should be allowed to get the source from Joe (on request). f is back under the line, Joe is again the user of the software. I'm unclear on g - if Joe starts up the program and puts in the input, then that output is emailed automatically, Joe is still the user. If it actually grabs his email and such as an automated process, the customer is now the user. Same with f. Does it get translated as part of the process, or does Joe do it? (Let me clarify my opinions on g and f. The customer is the user because the customer is interacting with the program - that is, the actions of the customer are causing actions of the program. In other scenarios, the action of the customer does not cause action of the program, because Joe may decide to be a jackass that day and spit in the cusomer's face instead of running it through the program. In those scenarios, I would consider the customer to be a user of Joe-beta0.1.sh rather than the typesetting software) k and l are definitely over the line. Now the customer is the user, and should be able to see the code. (All my opinions, of course). By the way, as a response to having specific requirements for distribution. What if, rather than saying that the code must be distributed via the same network medium, it said that the code must be provided on request by any of the means allowed in the GPL (including such strange means as snail-mail, etc). Would that possibly push the usability issue to the other side of the line? - Gregor Richards -- Gregor Richards [EMAIL PROTECTED] -- http://www.fastmail.fm - A fast, anti-spam email service. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
