Re: Bug#17621: [PROPOSED]: About versions based on dates
>>"Joey" == Joey Hess <firstname.lastname@example.org> writes:
Joey> I prefer to take a "don't fix it until it breaks" approach.
You r approach below, with all due respect, is already broken
as a policy proposal. We need to be general, and consistent, with the
numbering scheme when we talk about making policy.
Joey> For example, my package, lambdacore, has had the following
Joey> upstream releases:
Joey> Now though this is obviously not a numbering scheme dpkg can
Joey> handle, as luck would have it, these version numbers have
Joey> compared correctly under dpkg so far.
Just because one package has been lucky so far is not grouds
for not changing a broken scheme.
Joey> If a new version comes out in 2 days, of course, it will not
Joey> version compare correctly, and so I'll then have to go to a
Joey> sane version numbering scheme. But why impose one before I
Joey> really need to?
Consistency. Setting a sane naming scheme as policy shall also
prevent unnecesary epochs (since just looking at the file names shall
no longer give a clue about prdering once epochs are in place). I
think that using the sane approach in the forst place saves a lot of
Woman was taken out of man -- not out of his head, to rule over him;
nor out of his feet, to be trampled under by him; but out of his
side, to be equal to him -- under his arm, that he might protect her,
and near his heart that he might love her. Henry
Manoj Srivastava <email@example.com> <http://www.datasync.com/%7Esrivasta/>
Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E