Brian White <[EMAIL PROTECTED]> writes: > But your promise in not the point. The author wants this promise > from everybody. It's the best way to be assured that improvements > get distributed to everyone and not just a select group.
What if the author decides to not accept a change? Say the author considers the intent of the change repugnant and simply will not accept patches, even if it's runtime configurable and defaults to off. (I'm thinking along the lines of adding MIME capabilities to a mailer written by Tom Christiansen :-) This is perfectly within the author's rights, and it would be unethical if not illegal for us to violate the authors wishes. If the license permits redistribution of modified binaries, then if our user base asks/demands that this feature be included, then we're able to keep our users happy. If not, we have a lot of users recompiling from source which largely defeats the purpose of having a binary distribution. And this isn't hypothetical, either; qmail is in this situation: in order to be compatible with Debian's locking scheme, one must apply the patches for .lock support, which Dan Bernstein has, I believe, declared will never go into the core. Mike. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
