On Tuesday 10 November 2009 17:30:32 Dale wrote: > Alex Schuster wrote: > > Dale writes: > >> I am running into the same thing. I'm updating everything else then > >> going to tackle this. In theory, you have to unmerge device-mapper > >> then let it update. Something like: > >> > >> emerge -C device-mapper && emerge -uvDN world > >> > >> That's my plan at least. Unless someone else posts a better one. > > > > I just unmerged device-mapper and updated lvm2 right after that. So in > > case of a crash or something during a longer world update I still have an > > intact system. I'm not sure if it would boot with the old lvm2 and > > without device-mapper, probably yes, but I did not want to take any risk. > > > > Um, shouldn't the new portage have dealt auto-magically with this > > blocker? > > > > Wonko > > Here's something more interesting, I don't need or use lvm. I was > trying to track it down with equery but I was getting drunk going in > circles. > > I also assume that is why portage can't figure out a way to handle > this. It got drunk too. After a while a drunk just falls down. lol > > Thought about using the -et options with emerge to figure this out.
The bits of device-mapper you use are now in lvm2, not in a separate package anymore. You don't have to use lvm2 just because you merged it, it simply contains some tools you use. It's a case of the ebuild package being larger than you strictly need. -- alan dot mckinnon at gmail dot com
