On Tue, Oct 27, 2009 at 3:42 PM, James Laska <jlaska@xxxxxxxxxx> wrote: [snip] > We have the infrequently used reactive response should a feature fail to > meet it's stated objectives. However, is there something we can offer > that's more proactive (as a compliment to test days)? In this case the failure appears to be resulting from the composition of packages: The screen casting tools using unreasonable settings which happened to get ignored by older versions of libtheora. The most important testing that Fedora can do is integration testing. Libtheora was extensively tested upstream (including for compatibility with several known popular tools), but I would never have thought of testing these screencasting utilities. Presumably upstream will generally handle testing their own packages, I think that for libraries the distributions can add the most value by looking for whatrequires it and making a point to test those things. On Tue, Oct 27, 2009 at 4:36 PM, Matthias Clasen <mclasen@xxxxxxxxxx> wrote: > xiph.org has its own community of developers and > testers, and theora has testsuites, etc. Indeed. 525155 has been open for over a month, even prior to the official libtheora 1.0 release; the issue turned out to not be with libtheora upstream, but as far as I can tell no notice was provided in that direction. ... If legit bugs are taking this long to make it upstream then that has to adversely impact the quality of Fedora. -- fedora-test-list mailing list fedora-test-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
