Re: new toy

On Mon, Aug 27, 2001 at 11:09:51AM +0200, H. Merijn Brand wrote:
> As said, I did not deal with the network side of the database and until now
> I've always closed the bugs through the (too slow) web interface. I've
> discussed some of it with Richard on YAPC::Europe.
> 
> So, I've got no code like what you are discussing, but I'm sure I might wanna
> use it when it *is* available.
> 
> One point of attention I need to show: What if more than one of the bugadmins
> sends in a close mail with *different* comments and/or different close reasons
> (e.g me thinking to close the bug as duplicate with comments on why it is a
> duplicate of which bug, Richard sending a mail to close it as notabug, and
> Micheal closing it just as closed) consider what to do with the comments or
> what status should win (duplicate might be a win over closed and/or notabug,
> but also might fall to pray for circular references)
> 
> Just some thoughts.

I'm not sure what the current code does. There is a behaviour I think
is simple and reasonable:

Each action is a request to close a bug. Only open bugs can be closed -
it does not make sense to speak of closing an already closed bug. Therefore,
if you (merijn) send the first mail, you close the bug, and richard and
I both get an email back rejecting our change).

However in the current implementation this is interesting, as far as I'm
aware. Because if I wish to write a script to make a note on a bug,
and then close it, it must send two mails - a note, and a closure.
The note makes sense in any context, but the close might (perhaps) be
rejected if the bug is already closed.

I will think about what I want to do, and write a man page for it 
(hopefully today) before I write any more code, so people can
comment.

Richard - is there an email API for making a note on a bug and closing
at the same time? If not, let me know when the email code is not in flux
so I can start reading it...

Michael