On Thu, Jun 11, 2009 at 06:06:09PM +1000, Trent W. Buck wrote: > The most obvious way, I suppose, would be for mount to skip the test > if nfs-common is not installed. If nfs-common isn't installed, then > presumably any NFS entries it "sees" in /etc/mtab or /proc/mounts are > either stale or not relevant to the running environment.
And that would specifically be the wrong answer, since it is the absence of nfs-common that is the failure case. The situation is: Upgrades of a machine with NFS mounts that do not have nfs-common installed result in a machine that cannot mount NFS mounts, quite possibly necessary in order to have the machine function, including installing packages. Traditional Depends: would result in every machine on the planet having portmapper and nfs-common installed, which would cause the security team to justifiably kill me. The check is one that we should be able to drop for squeeze, since it only matters going from etch to lenny. > If this is a bad idea, I'd appreciate explanatory details from the > maintainer(s) -- currently it looks like this has just been tagged > "wontfix" without any discussion. There was a great deal of discussion on the original bug report, iirc, which was closed when the package was uploaded. lamont -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
