On 7/4/2010 6:59 PM, Eli Wapniarski wrote: > On Sunday 04 July 2010 15:19:13 Eike Hein wrote: >> On 07/04/2010 06:53 AM, Eli Wapniarski wrote: >>> It is not unforseable that >>> me you or someone else, on purpose or by accident or for whatever reason >>> installs kdepim 4.5 on a production desktop. >> >> kde-unstable is not for production desktops. If kde-unstable were to >> cater to production desktops, it would not be able to fulfill the pur- >> pose it currently serves, which is to test software that is expected >> to be part of the next Fedora release. >> > > I beg to differ. kde-unstable is branched off into next release and current 2 > releases. It is there to test the schedualed to be released as part of the > current working versions of Fedora. Otherwise, there would only be a rawhide > version (currently F14) which is not the case. kdepim is schedualed never to > be released. We may see version 4.5.1. If things go well. I found this paragraph to be really hard to parse and am not sure if the following will answer the embedded questions, but I'll try: kde-unstable usually contains builds of rawhide (or rawhide-derived, if changes are necessary) specs for the current stable releases, thereby allowing users of those stable releases to test software that the KDE SIG is preparing for the upcoming Fedora release without having to test all of rawhide. Evidence for that is that kde-unstable frequently contains software for which no main repo update for the current release is intended to be submitted, e.g. KOffice 2.x for the longest time was only there. Further confirmation would be found in Kevin Kofler's mail in this thread (he's a KDE SIG member) or Thomas Janssen's (another KDE SIG member) first mail to his thread, where he interpreted your question about whether kdepim 4.5 would be fit for the release cycle as applying to rawhide. I mean, if multiple KDE SIG developers tell you you are wrong, isn't that cause to go "huh, maybe I got it wrong?" for you? I think that addresses your argument that if kde-un- stable were for rawhide testing there would only be a rawhide branch of it. > Eli > -- Best regards, Eike Hein
