On Mon, 2006-07-17 at 14:06 -0400, Jesse Keating wrote: > On Monday 17 July 2006 09:47, seth vidal wrote: > > error handling will have to be maintained, too. > > > > > > You're more-or-less proposing: > > What if someone hoses their box, will we be hosing it further in the > > process of trying to unhose it? > > I was creating a hyperbole. > > Yum update starts, but doesn't finish. Has a journal of what should have been > done. > > User makes further changes to rpmdb, either by removing packages w/ rpm or > adding packages with rpm. > > User envokes yum, yum tries to replay journal, boom? Right - and I was responding that the only way we can tell is by figure out how to 1. know if the rpmdb has changed in the intervening time 2. know how to bail on brokenness. -sv -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
