On Wed, 2008-11-19 at 22:17 +0000, Daniel P. Berrange wrote: > Utterly useless. I already have a HUGE number of bug reports. The problem > is that 90% of them are essentially useless when first reported. It requires > several back/forth interactions between myself & the bug reporter to get > enough information to diagnose & resolve the problem. I've found that too, with Bugzilla. For system-config-printer which is also distributed by Ubuntu, I've found that the Launchpad bug reports usually contain most of the information I need to fix the problem: the package version number, and (usually) the Python traceback for example. This gets added to the bug report automatically using apport, without the user knowing how to do any of it. Of course there are plenty of Launchpad bug reports for system-config-printer that can't be solved that way. (I just ignore those.) I do think that having one-way crash reporting is useful to a certain degree, but implementing it in Bugzilla doesn't seem to be the right approach. A Bugzilla bug entry is quite a heavy-weight thing compared to just a entry in a database of crashes like kerneloops. The point is that software (never mind the user) is capable of reporting problems in enough details for fixes to be made in a lot of cases. In fact, I wrote the printing troubleshooter to do just that for the limited case of printing problems, and it helps *enormously* when someone attaches the troubleshoot.txt file from a printing problem. Tim. */
Attachment:
signature.asc
Description: This is a digitally signed message part
-- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
