[spambayes-dev] Imap and Spam Bayes

[Meyer, Tony]

> * When SpamBayes does run it seems to generate a lot of duplicate
>   messages. That is, when it filters messages in my INBOX (doing a
>   very good job of deciding!) it tends to generate two deleted
>   messages for everything, plus one undeleted message for the HAM
>   in the INBOX, or one undeleted message in the SPAM/Maybe-SPAM
>   folders.
[later]
> It appears that this might actually be an evolution problem.
> When I throw mozilla up against the mailbox it behaves as I
> would expect, when I quit evolution and restart it then it
> also shows the correct message counts. There is something
> screwy going on though, since spambayes is the only program
> I have seen that produces this behaviour in evolution. I will
> investigate further and see what I can find out.

That would be fantastic.  This is (I presume) the same as [
spambayes-Bugs-811759 ] Imap: odd message deletion behaviour:

<https://sourceforge.net/tracker/?func=detail&atid=498103&aid=811759&gro
up_id=61702>

Any light you could shed on it would be fantastic.

>  * It would be really nice if when deleting the messages it could
>    also mark them as read (makes evolution much happier)

Done, for current cvs.

>  * (This is a suggestion) Make the IMAP client do IDLE requests
>    where possible. Then it would not have to poll the IMAP server,
>    and can just be notified when new mail is placed into the
>    folders.

Yes, this would probably be the nicer way to do it.  I considered it
when writing imapfilter, but it seemed (and still does) like more work.
I doubt that I personally will be able to find the time to do this in
the near future, but if someone else wants to submit a patch, then I'd
be happy to test it out and commit it.  Either way, it's probably a good
idea to open a feature request (http://sf.net/projects/spambayes) for
this, or it'll get lost among the traffic.

=Tony Meyer