skaller wrote on 28/07/2005 01:18: > On Thu, 2005-07-28 at 01:01 +0200, Nicolas Boullis wrote: >>As far as I know, such things already happen with conflicts: let foo >>conflict with bar. If you install foo first, everything is fine. Later, >>if you install bar, foo is broken by bar, while bar knows nothing about >>foo... Where's the difference? > > > Ouch! I see. Being a math type person I tried to see > if there were a proper extension. However I didn't > go back to consider whether Debian itself was broken. > > The assumption here is that Conflicts is a symmetric > relation: if A conflicts with B, then B conflicts with A. It can't be like that, take this example: When X enters the archive, Y doesn't even exist yet (let alone it being in the archive). X could be (for a real world example) lvm-common. Now when Y (udev) enters the archive, it conflicts with lvm-common. However: Why should a new package of lvm-common be created? It's quite sufficient that the conflict is defined in one direction _because_ the relation (in itself) is symmetric. > On that assumption, apt is broken and should be fixed > the way I suggested: if there is a conflict which > not currently true, a no-X package should be created > by the package manager to prevent subsequent installation. > > So it looks like 'no-X' is not simply needed to satisfy > an unusal request -- it is needed to repair a fundamental > bug in Debian. No. It is not a bug (apt doesn't behave in an unexpected way nor does it behave in an incorrect way). It is merely an inefficiency. cu, sven
Attachment:
signature.asc
Description: OpenPGP digital signature