On Tue, 2008-02-12 at 12:21 -0500, Bill Nottingham wrote: > Dave Jones (davej@xxxxxxxxxx) said: > > One thing that comes to mind is that when a compose breaks, we fix it > > up, and take the approach of 'maybe tomorrows will work'. > > Perhaps we need to stage things after a compose breakage, so that > > new builds can't add new breakage, and we lose a day of new builds > > in favour of getting a fixed tree out there over perpetual firefighting. > > IIRC, we don't build the combined stage1/stage2 isos for rawhide. Perhaps > we should do that, and always keep the 'latest' builds around. At the very least, I'm hoping to get us a nice script for building said iso given repos. That way even if the process fails on the builders, it's easier for a tester to make their own. And also hopefully for a developer to make their own given a git checkout of anaconda + a repo. Of course, I spent the morning today looking at a weird createrepo bug instead :( Jeremy -- fedora-test-list mailing list fedora-test-list@xxxxxxxxxx To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
