Okay... SpamAssassin 3.0 is out with its much-hyped plugin system. So, now it would be nice to have plugin that asks from DSPAM, if an E-Mail is spam.
http://www.nuclearelephant.com/projects/dspam/ What I exactly want is this: 1) Switch off that Bayesian filter of SpamAssassin, because it is implemented in slow interpreted language called Perl. 2) Use DSPAM as Bayesian-like filter, because it is implemented in lightning-fast compiled language called C. 3) If DSPAM can give percentage-based probability of spamminess, use it to give different scores. 4) Use Berkeley DB-4 as storage-method in DSPAM, first. Other databases may be added later. 5) When using spam-learning of SpamAssassin, make DSPAM to learn that spam (or ham), too. In addition, let SpamAssassin to give all spam to Razor2, Pyzor and DCC, too. You may find this helpful when doing that plugin. http://thread.gmane.org/gmane.mail.spam.spamassassin.general/44126 Any takers? * * * Of course, I know there is a little bit overhead, when DSPAM-process is started and fork()'ed, but hopefully it does not matter too much. It seems there is a library called "libdspam". Maybe you could use its functions instead of starting DSPAM-processes. But remember: libdspam is licensed under GNU GPL 2. And SpamAssasin 3.0 is licensed under Apache Software License, Version 2. According to FSF that license is incompatible with GNU GPL: http://www.fsf.org/licenses/license-list.html#GPLIncompatibleLicenses Maybe author of DSPAM could change license of that library to GNU LGPL. -- Juhapekka "naula" Tolvanen * http colon slash slash iki dot fi slash juhtolv "halpojen hoitojen maailma uljas haluaa taistosi latistaa, mielesi lipeävedellä valkaistuun ruotuunsa, joka on hautausmaa" CMX
