httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <field...@kiwi.ICS.UCI.EDU>
Subject Re: [TEST] changes to bugdb.cgi
Date Mon, 02 Jun 1997 05:19:43 GMT
>>   To append more information to this report, send a mail message with
>>   To: APBugs@Apache.Org
>>   Subject: <whatever the auto-detect uses for this PR>
>    No-one (well, very few) will read it.  They'll do what they do now,
>    for us and other messages: reply.  Currently, such replies go to the
>    group member who touched the PR, and nowhere else unless the
>    submitter explicitly adds another address.  With the change, the
>    reply will get attached to the PR and everyone on the apbugs list
>    will get a copy.  Likewise, if one of the group members has a
>    thought about it, he can just reply to the updated message, and
>    *his* reply will go into the bugdb.  (This is all in theory; I'm
>    still testing it.)  Whether GNATS CCs the originator when it logs a
>    reply..  well, it probably should but I don't think it does.

It does every time the state changes, so I'd expect it to on replies
(assuming it allows replies without state changes).

My experience with mail systems (being a listmaster) is that people
reply without thinking and don't expect it to be saved, even on public
lists with a known hypermail archive.  The only time they think is when
they have to add the address themselves.  I'd rather force people to
think before posting rather than treat it like a public mailing list.

If the traffic gets too high on the apbugs list, the developers will
stop reading it.  Maybe we should have two lists -- one for the initial
bug report and another for replies.

Anyway, go ahead and make your changes.  If it turns out to be a problem,
we can always change them again.

>>>     o Changed "mumble-From-To: x-y" to "mumble-From->to: x->y" because
>>>       values that contain hyphens become hard to read
>>That may cause problems with the e-mail interface.
>    Why?  That's not standard GNATS; it's strictly something bugdb.cgi
>    is doing.  Those are informational heralds added to the case as
>    audit records.  In other words, they're content.

They are also keywords for the email interface to changing pr's.  It is
better to just leave them as-is unless the improvement is significant.


View raw message