On Thu, Jun 10, 1999 at 01:22:26PM -0600, Marcelo E. Magallon wrote: > On Thu, Jun 10, 1999 at 02:02:35PM -0500, Chris Lawrence wrote: > > > I further propose that the use of bzip2 be mandatory for newly uploaded > > source files > > Upstream doesn't always provide .tar.bz2 packages. As said elsewhere, I think a source package can be regarded as "pristine" if it is md5-identical with the upstream version in *uncompressed* form. Even gzip does not compress deterministically. I learned this when fooling with the massive X source archives. I can take the same damn .tar file, gzip -9 it on different machines, and get different results. They're not different by much, say 14 bytes or so, but it is a difference. Therefore I am of the opinion that we should not in general take md5sums of compressed files anyway. In other news, you could always "encapsulate" the upstream pristine source archive in the Debian source archive, complete with the upstream compression method intact, if you use the approach that the X source packages you use. I think this still counts as "pristine source" because you can indeed *get* to the pristine source archive. So either way I do not see your objection as a big hurdle. But yeah, I'd be happy to compress the X source archives as aggressively as possible. bzip2 with the throttle wide open, or whatever. I wouldn't be worried about the hard hit taken by uncompressing it, since people without really manly machines won't be compiling X anyway. -- G. Branden Robinson | America is at that awkward stage. It's Debian GNU/Linux | too late to work within the system, but branden@ecn.purdue.edu | too early to shoot the bastards. cartoon.ecn.purdue.edu/~branden/ | --Claire Wolfe
Attachment:
pgpQikTKeu6Cm.pgp
Description: PGP signature