On Sun, Mar 14, 2004 at 04:02:38PM -0500, you [seth vidal] wrote: > > > Hmm, I see. > > > > I don't suppose you have any plans to make this overridable by yum.conf > > options? (Something like repository priorities perhaps - if repository of > > lesser priority fails, the higher priority ones could still be upgraded, > > since any overlap would be solved by using the higher priority repository.) > > Feel free to write a specification up for how this should work and get > some consensus on it. It's not a trivial process to document. First, I'll admit that I haven't really thought this through - it was just an idea that sprang into mind while reading your reply. But anyway, I was thinking something as simple as: - repositories in yum.conf have priorities - priorities are defined either with priority=<number> or simply by repository order in yum.conf (the former is probably the better idea, it should be backwards compatible.) - if two or more repositories provide the same package, the one from the highest priority repository is used. - if a repository can't be connected to, the repositories with higher priority can still be considered without aborting the whole transaction. This way, one could have the distribution (say, Fedora Core) updates going even if the auxiliary repositories (livna, freshrpms) are down. I realize this hand-waving doesn't have much value, and regrettably I have no python programming experience. But feel free to shoot the idea down, if you care... -- v -- v@xxxxxx
