On 2007-01-11 23:19:52 +0100, Michael Holzt wrote: > > Basically it sends each mail to spamd for each of the recipients using > > the "User:" header. Spamd then uses the config of the supplied user. > > Sounds rather inefficent, because the same tests are done multiple times > for the same mail. Shouldn't it be possible to have the mail scanned by > spamassassin once and then just use the list of matched rules as given > by spamassassing to have the score calculated for each user?
Theoretically no: Some rules (e.g., the Bayesian filters, whitelists,
etc.) depend on per-user configuration, so they may trigger for some
users but not for others.
Practically maybe: If your users cannot configure white- and blacklists
on the MX and can't update the database for the Bayesian filter
(autolearning doesn't work with spamd), you might as well scan the the
mail only once and compute only the score. It means that the
spamassassin plugin needs to parse the spamassassin configuration, though, so
you're duplicating some code from spamassassin, and the per-user config
files need to be readable for qpsmtpd.
I'd rather go into the direction of even less efficiency and more
flexibility and run spamc with appropriate privileges.
hp
--
_ | Peter J. Holzer | I know I'd be respectful of a pirate
|_|_) | Sysadmin WSR | with an emu on his shoulder.
| | | [EMAIL PROTECTED] |
__/ | http://www.hjp.at/ | -- Sam in "Freefall"
signature.asc
Description: Digital signature
