* Manoj Srivastava <[EMAIL PROTECTED]> [060413 06:05]: > >> A colon is explicitly allowed if the version contains an epoch. In > >> this case, the build generate file names which contain a colon, > >> too, and I'd expect them to be rejected in the same manner (they > >> look essentially the same because the epoch is dropped). > > > > Sounds like Policy needs an update to me if we're not going to > > accept version numbers that contain a colon (other than the epoch > > colon). I'd be happy to second such a proposal. > > Why not fix the bug instead?
Policy has (besides other things) the job of documentating current practice. This discrepancy is there for a long time. It is even known for a long time. (compare e.g. http://lists.debian.org/debian-policy/2003/10/msg00110.html) No one uses colons in the upstream version, as it is not possible for a long time. So we can either adapt the policy to the rules actually in action, or the fix the rules in action. As noone needed it enough that it was fixed for years, and always disallowing colons makes the whole rule easier and brings us in a bug free state directly (some tools would not reject a explicitly invalid version then, but we know all programs work with valid versions and no other programs might cause problems when first hit by them) my suggestion is to adapt policy to practice. Hochachtungsvoll, Bernhard R. Link -- Sendmail is like emacs: A nice operating system, but missing an editor and a MTA. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
