Aleksander Kamenik wrote:
> Michael Monnerie wrote:
> 
>> Sounds good, but a config option could be of value for some
>> environments, i.e. some people must receive even duplicate messages
>> (for automated e-mails).
> 
> What automated messages with different content do NOT differ by message-id?

Decent mailers generate unique message-ids no matter what. Dbmail
already generates a message-id if it's missing from the original message
since message-ids are needed for server-side threading
(thread=references) which we don't do yet.

> Anyway, I would say it's more of a user setting than system setting. And
> in dbmail's case a sieve thing and I don't think there's something like
> that implemented in sieve.

My current implementation checks resent-message-id/message-id values per
mailbox, rather than globally like cyrus does. So if you receive the
same message twice in different mailboxes you will still see it both times.

Enabling it on a per-user or even per-mailbox basis is conceivable. But
lets first do a simple and effective implementation that can be used in
2.2.6+ and see how users like it, and how they would like to see it
improved.

> 
> Also, is checking the message-id header secure enough? Maybe the cyrus
> community has seen some problems with this? I'm just naturally paranoid
> about ignoring messages based on a header.

Good point.
There are some know problems here with broken mailers (outlook and
groupwise)

http://www.irbs.net/internet/info-cyrus/0705/0184.html

so quite obviously, simply checking by message-id is not 100% reliable
for all cases. That's why it will be disabled by default in 2.2.6.

> It would be cool to find out which messages this affects. In the level 3
> trace maybe "message with ID blah is a duplicate"?

That message is currently at TRACE_DEBUG, but I'll make sure it's issued
at TRACE_INFO


-- 
  ________________________________________________________________
  Paul Stevens                                      paul at nfg.nl
  NET FACILITIES GROUP                     GPG/PGP: 1024D/11F8CD31
  The Netherlands________________________________http://www.nfg.nl
_______________________________________________
DBmail mailing list
[email protected]
https://mailman.fastxs.nl/mailman/listinfo/dbmail

Reply via email to