On Mon, Feb 7, 2011 at 3:46 AM, Mihai Popescu <mihai...@gmail.com> wrote: > So the process I thought about it's not true. Better to remove the > SHA256 then, what purpose can it serve if it is not syncronised?
Practice for release, perhaps? If they cause too much of an uproar in snapshots, then they'll likely be killed completely, including from the releases, because trying to get them right in the release without the installer bits being regularly tested would be unsustainable. Sure, it would work fine in the 4.9 release, but with no testing between the 4.9 release and the 4.a^H^H^H5.0 release there's a not insignificant chance they wouldn't work then, and the probabilities get worse from there. > I still don't figure out why this checksum missmatch is ( on the same > server, not among servers). No x* entries in the separate SHA256 file, so you obviously grabbed the files between when a base build finished and when the following xenocara build did. As for other claimed mismatches: insufficient data. Philip Guenther