On Thursday 22 January 2004 03:34 am, Michael Schwendt wrote: > On Wed, 21 Jan 2004 15:22:22 -1000, Warren Togami wrote: > > RH's policy is ANY PACKAGE THAT LEAVES YOUR LOCAL SYSTEM must always > > bump the release number. > > Uhm, that is certainly due to different requirements of their automated > build system. But if you do not bump the release number, you get into a situation of WHICH version of the release is really the release. The release number is just that; a serial number of which release it is from the build machine. Streamlining changes within a release number is a pain. If you must, for pre-releases use decimal releases. I try to do that upstream with the PostgreSQL packages; the first set off the press is release 0.1PGDG, up to the point I feel it is worthy of real release. In my case, all I can do for PostgreSQL is push these prereleases out to the ftp server, since there is no fancy release management system in place. When the QA has reached the level I desire (which means that it built and ran successfully on all target platforms, which for me goes all the way back to RH6.2), I bump it to a -1PGDG release. Sometimes that doesn't occur until the next PostgreSQL minor is released, so I release the -1PGDG in the next minor series. The decimal release number states what the intent is. Lining this up with the way the RH errata releases have been done is going to be interesting at best. But having two packages with identical E:V-R but different contents is terrible, even if it is within the QA queue. -- Lamar Owen Director of Information Technology Pisgah Astronomical Research Institute 1 PARI Drive Rosman, NC 28772 (828)862-5554 www.pari.edu
