On 2007-08-06 08:16:08 -0500, Ron Johnson wrote: > Your original question was woefully lacking in detail.
I didn't want to enter into technical details as this would lead to nowhere, IMHO. > > I've reported the problem. The bug was tagged as wontfix > > because the maintainer of the package doesn't care that things can > > break in other packages and prefers to return with an exit status of 0 > > even if the program failed (which is against what an exit status is > > meant for). > > Maybe you filed the bug against the wrong package? > What package did you file it against? apt? It was the right package: apt-listchanges. Note: the bug first came from of a bug in apt/python-apt, but the problem is that apt-listchanges did ignore the error. Hence my bug report against apt-listchanges "apt-listchanges lets the package installation without user confirmation in case of error". > Or maybe you now need to file a bug saying that "it" returns 0 even > in a failure condition. I suggested that as a simple workaround. The maintainer doesn't want to change it because some users (running unstable) don't know how to disable a package (it may be needed to disable apt-listchanges temporarily to fix things). That's quite surprising for users who run unstable... Of course, this may need more work from the user, but with less risk to break the system. But there could be better solutions such as using a wrapper, as I suggested first in my bug report: When apt-listchanges cannot run (e.g. because of bug 435674), the user misses all the changelog information (which may contain critical information, such as some NEWS announcements). So (in particular because breakage due to python occurs too frequently), it is important that apt-listchanges runs in a wrapper with few dependencies (e.g. only the libc) so that it can warn the user if the changelog/news information cannot be displayed and let the user abort the installation. -- Vincent Lefèvre <[EMAIL PROTECTED]> - Web: <http://www.vinc17.org/> 100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/> Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
