[Mailman-Developers] premature closure? (was Weirdness with 2.1)

[Richard Barrett]

At 16:03 06/02/2003, Bryan Fullerton wrote:

>On Thursday, February 6, 2003, at 10:45 AM, Bill Bradford wrote:
>
>>2.1 works fine for me, *except*
>>
>>When I go to http://mysite/mailman/admindb/<listname>, it asks for the
>>list password.  I give it, get the "held for moderation/approval" queue
>>of messages, pick my choices on each, and hit submit.
>>
>>I then get the password auth page again, have to do it over, and hit
>>submit *again*, then it "takes" and processes the messages per my
>>admin choices.
>
>Are you running Mailman 2.1 and 2.0 lists in the same URI space? If so,
>you're hitting bug #664466, which is fixed in CVS (and the bug closed,
>so you won't see it if you browse open bugs in SourceForge).
>
>http://sourceforge.net/tracker/index.php?func=detail&aid=664466&group_id=103&atid=100103
>
>Thanks,
>
>Bryan


Am I alone in thinking that closing a bug report should only be done when 
the numbered release is made that incorporates the fix.

Closing the bug when the fix goes into the CVS version only "hides" the 
existence of fixes for known bugs in the current production (numbered) 
release. It means that in trying to consider what might be applicable to 
the latest numbered release effectively all bug fixes, both open and closed 
have to checked.

I would have thought that pending would be a better category for bugs whose 
fixes are in the CVS but not in the numbered release, with 
re-categorisation as closed when the fix appears in a numbered release.