>>>>> On Tue, 25 Mar 2003 10:37:53 +0100, Arthur Bergman <[EMAIL PROTECTED]> said:
>> The flow would look like this:
>>
>> 1. CPAN authors use a form on PAUSE...
>> 2. Email gets sent to RT and placed in the modules queue
>> 3. RT sends a copy of the request to [EMAIL PROTECTED]
>> 4. RT tracks any discussion that happens on [EMAIL PROTECTED]
>> 5. When it is approved/rejected/altered/etc, one of the modules@
>> admins goes into the RT system and marks it as such.
>>
>> 6. Status reports can be generated that show the status of the
>> queue.
>> (5 in queue, 3 under discussion, 2 new)
>>
>> We would be doubleplushappy to host this on the perl.org RT instance.
> I think this would be an excellent idea.
Yes! And I'd be doubleplushappy if I could deal with RT as a (mostly)
black box.
The one problem I still see in this scenario, would be the RT ticket
number. Somewhere between (3) and (5), an admin can click on the URL
that comes with the mail/note generated by (2). By clicking on that
URL they approve the module. I'm happy that we have that
one-click-approval thingy working. But PAUSE should then be able to
tell RT about the successful approval. To do that, it seems, that
PAUSE would need the ticket number assigned by RT.
Did I express myself clearly? I hope there's a way for PAUSE to find
out the ticket number. One possible solution would run like that
(point 2 and 4a are different from above):
1. CPAN authors use a form on PAUSE...
2. A HTTP request is sent to RT and placed in the modules queue.
PAUSE records the ticket number it gets from RT.
3. RT sends a copy of the request to [EMAIL PROTECTED]
4. RT tracks any discussion that happens on [EMAIL PROTECTED]
4a IF an admin decides to approve by clicking on the approval
button, PAUSE informs RT that the ticket can be closed.
5. When it is approved/rejected/altered/etc, one of the modules@
admins goes into the RT system and marks it as such.
Actually, (5) would then quite often be done automatically. Is this
doable? And if so, please let's start coding.
--
andreas
