owner 536613 ! quit Hi!
On Sat, Jul 11, 2009 at 11:31:07PM +0200, Francesco Poli wrote: > On Sat, 11 Jul 2009 20:09:18 +0200 max wrote: > > [...] > > Hi > > Hi! > > > > > the command > > > > # apt-listbugs list apt-listbug > > > > lists > > > > Reading package fields... Done > > Reading package status... Done > > Retrieving bug reports... Done > > Parsing Found/Fixed information... Done > > > > and the return value is 0, ie. Success > > You are right, I can confirm that apt-listbugs behaves like this (I've > just checked with version 0.1.0, currently in unstable and testing). > > > > > but there is no apt-listbug package and so the listing should be > > something like: > > > > ... > > ... > > WARNING: There is no maintainer for apt-listbug > > WARNING: There is no record of the apt-listbug package > > > > > > and the return value should not be 0. > > Well, I am not sure that this proposed new behavior (i.e.: exit with > non-zero status when a package does not exist) is a good idea. > > For instance, when you pin a package because of some bug, the pinning > is written to /etc/apt/preferences; a cron.daily job checks the package > to see if it still has the bug (see /etc/cron.daily/apt-listbugs, which > uses /usr/share/apt-listbugs/aptcleanup) and drops the pinning when the > package no longer has the bug. > > Suppose you're running Debian unstable (or Debian testing) and you pin > a buggy package; then you shut the system down for some time; when you > boot it up again, the package could have been removed from unstable > (and testing); if the package never made it into a stable release, the > BTS web interface says that there is no maintainer and no record for > the package. Now, what would happen if apt-listbugs exited with > non-zero status? The cron.daily job would try to > run /usr/share/apt-listbugs/aptcleanup, which would exit with error, > and hence the pinnings in /etc/apt/preferences would not be properly n> cleaned up... > this is why we would have to adapt aptcleanup as well.. > As a consequence, I am under the impression that the behavior of > apt-listbugs on non-existent packages is *intended*, rather than a > *bug*. > > Ryan (maintainer of apt-listbugs), do you agree with my analysis? > not at all, I think this is a bug, and I will fix it. the counter argument of "well it would break another script in *our* package" isn't a very good one, since we can easily change that script to account for this and handle it correctly. I was holding off on replying since I am at work, but.. returning 0 (meaning "Success!!!!") and not giving an error message is *very* wrong behavior imo. anyways, I will propose a fix this bug when I get home. assuming I can convince you, I will then commit it. ;) oh, and s/maintainer/co-maintainer/, this is as much (if not more) your package than mine. :) Cheers, Ryan -- _________________________ Ryan Niebur ryanrya...@gmail.com
signature.asc
Description: Digital signature
