On Fri, May 08, 2009 at 09:46:36AM -0700, Jesse Keating wrote: > On Fri, 2009-05-08 at 12:26 -0400, James Laska wrote: > > One thought was ... in the absence of an Alpha, QA can schedule several > > test days. However, experience shows that not all test days have the > > same starting conditions. During F10 and F11, several test days > > experienced 'launch failure'. That is, we either couldn't generate a > > usable live image, or the steps required to prepare the test environment > > required enough forks in the road that it became a barrier to > > participation. > > Having more visibility and more coordinated planning on upcoming test > days can help us focus to get the required bits in good shape for those > days, be it anaconda, be it livecd-tools, whatever. These are lower > cost then grinding everything to a halt (as far as the freeze is > concerned) for a week or more. > > > > > Another investment for QA would be to provide more data on the health of > > rawhide. This doesn't directly influence quality, but more a foot in > > the door when it comes to measuring quality. > > > > Without the Alpha, how do we entice the distro to come together? How do > > we ensure that we aren't shifting the bugs we find in an Alpha now, to > > the Beta? Should we hold the F12 beta to the same/higher/lower > > standards that we hold the current Beta? > > Yes, we should hold Beta to the same or higher standard. What standard should that be? Are there some metrics we could use to measure against that standard? There must be some factors we can identify that constitute a "successful" Alpha vs. Beta (or Beta vs. Preview, or...). For instance, you could say a successful test release results in filing of bugs. When it comes to those bugs, what would interest us to know about their distribution between one test phase and the next, or from test release to GA? -- Paul W. Frields http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 http://redhat.com/ - - - - http://pfrields.fedorapeople.org/ irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
