On Tue, Jan 02, 2001 at 05:10:12PM -0800, Joey Hess wrote: > I think that using a delivery mechanism that is explicitly *talking to* > someone, rather than making a comment in a file that is not part of a > conversation, tends to result less often in terse and uninformative > things like:
> * Unreproducable, probably fixed in 2.2, closes: #57026, #42726, #40768 > closes: #45848, #58367, #62990, #40870, #67296, #38897, #60099, #66769 Perhaps so. My assumption was that people who are bad at writing changelog entries are probably going to be bad at writing emails too, but maybe I just take my changelogs more seriously than some folks do. I would never create an entry like that. And I have gotten email replies to my bug reports that were no more informative than the above. Some people just aren't very communicative, and I'm not sure there's much we can do about it. (Except point and laugh and try to publicly shame them into better behavior.) > That's silly. Why should I record "It was user error, Closes: #nnnnn" in > my changelog? My program has not changed. Yes, sorry, I was thinking more of the case where the change happened in an earlier version, or where something else changed earlier to render the bug irrelevent. Mostly because I just had to deal with a whole lot of those cases recently, after adopting several packages. Your first example I would consider borderline (ignoring issues of bad style for the moment), because it says "probably fixed in 2.2", so it was probably a change in the package that fixed the bug, so it probably should go in the changelog. Bottom line, I would rather see us try to encourage people to write better changelog entries than try to forbid them from using changelog closers in certain cases. The former will help produce a better system overall, while the latter will simply prevent it from getting worse. cheers -- Chris Waters | Pneumonoultra- osis is too long [EMAIL PROTECTED] | microscopicsilico- to fit into a single or [EMAIL PROTECTED] | volcaniconi- standalone haiku
