[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
duplicate suppression and splitting followups to go with parent
I'd like to collect some thoughts on duplicate suppression and on
splitting followups to go with the parents. Maybe I should define the
latter. It refers to the following situation: you receive a mail, you
manually move this mail to a different folder, now somebody sends a
followup. In such a situation, I prefer the followup to automatically
go in the same folder where I move the original.
I'm using Cyrus with plus addressing which gives me a few more
options.
* The Cyrus deliver program groks duplicate suppression and writes a
deliver.db which contains seen message ids as well as the groups they
were written to. Wonder whether it is possible to make use of that?
* I have written a patch to nnmail.el which augments the normal Gnus
duplicate suppression mechanism by storing the group along with the
message id in ~/.nnmail-cache. There is a function to be put in
nnmail-split-fancy which looks there for all message ids in the
References header and returns the corresponding group, if any.
There has been some previous discussion on adapting that to nnimap
in some way.
One issue is that my patch stores stuff in ~/.nnmail-cache which is
on the client side, but in an IMAP context you want to store stuff
on the server.
* I think some people modify the message ids of outgoing messages to
contain the group name they posted from. The split mechanism can
then look in the References header.
This methods fails when messages are moved manually.
* I could modify the From address such that replies will automatically
go to the right group because of plus addressing. (Plus addressing
means that mails sent to jrl+foo@frob.org get put in the INBOX.foo
folder by the Cyrus deliver program.)
This also suffers from the moving-manually thing.
Have I overlooked something? From the above it seems that
automatically storing the group name and the message id for all
messages which are moved or copied or otherwise entered into a group.
But maybe the end can be achieved using some other means I haven't
thought of?
kai
--
Life is hard and then you die.