On Wed, 14 Feb 2001, Jürgen A. Erhard wrote: > I got a bug-closing notification on #59584. This bug was reported > against gltt-bin. But it was closed by a fetchmail upload! Yeah, sorry about that. I managed to screw up on that bug 2 or 3 times... Well, I restored the bug status after I noticed the problem. Sorry about the number of useless emails it must have generated :( > So... dinstall (which I *think* does this) check that the bug in a > Closes clause really belongs to the package being installed. Better add packages provided, as well. However, that will require some sort of indexing of the bugs per package to be retrieved by dinstall prior to its run... There's a way to do this without replicating the bug/package index: dinstall could email bugnumber-done-verify@d.b.o, including a new subheader with the packages the bug may apply to. d.b.o would then refuse to close the bug if it is assigned to a package not in the list. When refusing, it just needs to forward the mail back to the package uploader, who needs simply to forward it back to foo-done@b.d.o if the bug should indeed be closed. -- "One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie." -- The Silicon Valley Tarot Henrique Holschuh
Attachment:
pgpBw4cJI665n.pgp
Description: PGP signature