Sending the discussion to the discussion side of things instead of the bug ticket.
Stefano Lattarini wrote: > Karl Berry wrote: > > P.S. I took bug-coreutils out of cc simply because I didn't want to deal > > with "opening" a bug report (doesn't it require the bug number somewhere > > to append)? That was certainly the safe thing to do without having any other information about it. :-) > I don't think so; the debbugs software should be smart enough to > realize that you're answering to a massage that has already opened a > bug report, and it should thus avoid to open a new bug report. I > might be wrong on this, though, and I don't feel inclined to > experiment right now. Correct. With the exception that the debbugs robot needs to have had time to have received and processed the original message so that it can recognize the second message as a reply to it. This usually isn't a problem with replies that a human receives and responds to. But for example they can be sync'd up in time if two messages are in the hold queue waiting for approval in Mailman. If both are approved at the same time then it is a race condition as to which message Mailman will process first. Murphy says that case always generates a new bug ticket. In that case it is good to approve only the first message first, wait a few minutes or observe the passage of the original email, then approve the second one afterward. Then the debbugs robot can recognize the second as a reply to the first and do the right thing. That will avoid creating a second bug ticket. Bob
