Re: [Evolution] spam filtering via evo with an alternate mailbox?
- From: Not Zed <notzed ximian com>
- To: Dan Stromberg <strombrg dcs nac uci edu>
- Cc: evolution lists ximian com
- Subject: Re: [Evolution] spam filtering via evo with an alternate mailbox?
- Date: Tue, 25 May 2004 10:01:07 +0800
you probably have to just do ctrl-a ctrl-y to kick it off. if you've got a big inbox, it could take a while to run.
filtering will only be applied automatically, when told to, for recent messages. its kind of server dependent what recent means (and kind of actually badly specified too), but if you're using the one client on a given server it should work fairly reliably.
On Mon, 2004-05-24 at 16:59 -0700, Dan Stromberg wrote:
First off, let me say, my incoming mailbox, which I've switched to
~/mail/mybox, is huge. I'm trying to set a good example for my
endusers, by not leaving this large quantity of mail in my maildrop,
/var/mail/strombrg.
We used to have antispam via spamassassin on our mailserver, but we had
to turn it off, because it was loading the poor old machine up so much
that it had ypserv problems. Why ypserv is so flakey is a story for
another night.
Anyway, I set up a filter which "Pipes message to shell command"
spamassassin -e (with a hard path), and refiles the message if there's a
nonzero exit status to my "evospam" folder.
It's not working. I'm bummed. I get HUGE quantities of spam.
Anyway, if I save one of these new spam messages into /tmp/msg, and then
run spamassassin -e < /tmp/msg (with a hard path again), I get an exit
status of 5, which would seem to be what evo needs to do its thing, no?
I also clicked "tools/settings/receiving options/apply filters to INBOX
on this server".
Do I have to leave my mail in my maildrop(/var/mail/strombrg) for this
to work properly?
I'm using an sslified uwimapd, if that makes any difference.
I'm eager to switch to dovecot, possibly with maildir, but I have to
find the time somewhere.
TIA.
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]