[SAtalk] Enable localized rule descriptions

Hi, just a quick question: How do I enable localized rule descriptions? There's a lot of german rule descriptions in the stock SA distribution, but they're not used on my (german) mail server setup. Do I miss a point? --ck --- The SF.Net em

[SAtalk] Bizarre spam

Hi, I received this here just some minutes ago. It went to a role account and through a ticketing system so there's no usable headers (it wasn't scanned by SA either), but the content speaks for itself... Looks like pure bayes poison. Did anyone else receive this and can tell me if it's correc

Re: [SAtalk] SA - flag some, delete some

Jim Viton wrote: I have an interest in flagging email as spam at a certain threshold and deleting at another threshold. Has anyone done this? I am using sendmail, spamass-milter, and spamd - latest and greatest. I suppose I could set up another milter calling a a different spamd daemon with a di

Re: [SAtalk] "no such user" directories

[EMAIL PROTECTED] wrote: Hello, I've been running SA for about a year on this particular server, and I've never noticed it until today, but when using the -a flag with -c, SA is trying to create user preference files for users that don't exist. I'm using qmail, vpopmail, and SA, and I've got a cat

Re: [SAtalk] Obfuscation by Punctuation

Greg Webster wrote: Here's what I've recently done: rawbody GWW_PUNCT /([a-z][:punct:]+[a-z])|( [A-Z][:punct:]+[a-z])/i score GWW_PUNCT 2.0 It's not perfect, but it does the job. As well, I've noticed a lot of these include the domain 'doctor45.com', so I've been giving a good high score for th

Re: [SAtalk] Obfuscation by Punctuation

Fred wrote: That's a fake header name with your e-mail address encoded with base64. I guessed it is some spam devilry. Actually, I don't care if harvesters pick up this address, it's also under SA monitoring :-) Add the following rule to your local.cf and you will never see those again ;) Isn't h

Re: [SAtalk] Obfuscation by Punctuation

Gary Funck wrote: A pattern like the following: /([a-z][;][a-z]+.*){5}/i might get some traction. This has to be run after the HTML is stripped. That exact message got through here, too. Actually, it was using the whitelist_from trick to get a whopping -93.6 points, but OTOH, bayes_60 and l

[SAtalk] False Negatives increasing in number?

Hey, has anyone else seen false negatives increase in number over the last weeks? Although I think I have a pretty well-fed bayes database, I still get about one to three false negatives a day. This wouldn't normally interest me, but those mails are pretty dumb spam that _should_ be caught. Th

[SAtalk] CPU Usage while sa-learning?

Hi, I have been initializing our Bayes databases after a system rebuild and 2.60 upgrade, and have noticed that feeding it very large mbox files causes some undesired behavior - namely an extremely high load of between 9 and 12. This load is rising even further during the course of sa-learn's exa

RE: [SAtalk] Multi-user SpamAssassin setup on vpopmail

Hi, > Take a look at the new vpopmail integration in SA 2.20 first before > resorting to SQL. See the README.spamd-vpopmail in the spamd dir of the > 2.20 distribution for details. It gives support for virtual > vpopmail users. > I wrote the patch and use it daily and works great. great, thank

[SAtalk] Multi-user SpamAssassin setup on vpopmail

Hey there, since yesterday, I'm a happy SpamAssassin user (thanks to Sascha Schumann, who might even read this list, too) on my VPopMail setup. I'm now investigating the possibility of setting SA as the default spam blocking tool for all our domain customers. Unfortunately, vpopmail uses virtual