I have a suggestion that could improve the admin and end user experience. It is related to SPAM filtering, and in particular, using the DSPAM software. Normal DSPAM works in line with the MTA just before the mail is delivered to a users maildir/mailbox. If the user wants DSPAM to learn or unlearn an email as being SPAM, the user must forward that email back to an email address that pipes the email back into DSPAM with the appropriate flags. This is really unnecessary and cumbersome. It also requires the MTA to keep spam and false positive email aliases for each user.
My suggestion is this: Is it possible to create two special binc imap folders called 'LEARN_SPAM" and "UNLEARN_SPAM" ?. Email still enters the inbox filtered by DSPAM, which is inline with the MTA. If the user decides a particular email is SPAM he simply moves the email into the "LEARN_SPAM" folder, which triggers binc to pipe the email into the dspam binary with the proper flags, ie. "|/usr/local/bin/dspam --user [userid] --class=spam --source=error" or in the case of a false positive, the user copies the email into the "UNLEARN_SPAM" folder, ie. | /usr/local/bin/dspam --user [userid] --class=innocent \ --source=error --deliver=innocent This has the effect of removing the other aliases from the MTA and the user can simply drag and drop email into a specific folder to retrain DSPAM. The work flow from an admin and an end user is much more elegant, IMHO. I'm not sure how much work this would be to code or if it breaks any IMAP RFC's. Your comments would be welcome, even if you think this idea is way out in left field. Thanks, /RS -- ---------------------------------------------------------------------- Roger Sistla rsistla at rogers dot com ----------------------------------------------------------------------
