Re: [SAtalk] 'Opt'

2003-09-30 Thread Matt Kettler
At 03:03 PM 9/30/03 -0700, Robert Leonard III wrote: I hadn't thought of that! That sounds like a bug to me then.. surely my company isn't the only one that has the letters opt in their name?! Actualy the "OPT_HEADER" rule isn't just looking for "opt" in the domain name. It's looking for things l

Re: [SAtalk] Default behavior of check

2003-09-30 Thread Matt Kettler
At 03:40 PM 9/30/03 -0700, Steven Manross wrote: Where check would be performed with "local_only" tests first... and then if it is tagged using locals only, you don't bother with RBL and/or external tests, because it's already been tagged, and you don't need to add the cost of the external tests i

Re: [SAtalk] Bayes (again)

2003-09-30 Thread Matt Kettler
At 09:30 PM 9/30/03 -0400, Jack Gostl wrote: to the end of my user_prefs, and I still get lots of high Bayes scores. Is that normal? I more than a little surprised at some of the messages it flags as spam. Have you used check_bayes to take a look at what tokens are in your bayes db? you shouldn't

Re: [SAtalk] Auto-learning problems

2003-09-30 Thread Matt Kettler
At 12:04 PM 9/30/03 +0200, Euro Cocolo wrote: ** Sep 30 11:51:19 mail1 spamd[21760]: logmsg: Still running as root: user not specified with -u, not found, or set to root. Fall back to nobody. Sep 30 11:51:19 mail1 spamd[21760]: Still running as root: user not specified with -u, not found

Re: [SAtalk] Forged mail pretending to be from MS

2003-09-30 Thread Matt Kettler
At 10:03 AM 9/30/03 -0400, Jeff Koch wrote: Can someone explain what triggers 'FORGED_MUA_OUTLOOK'. We're using 2.55 and have seen some cases where for reasons we cannot explain this is getting triggered and with the default scoring of 3.5 legit mails are getting spam filtered. it's triggered b

Re: [SAtalk] No Xspam score?

2003-09-30 Thread Matt Kettler
At 01:16 PM 9/30/03 +0200, Jan Erik Skogsholm wrote: Hi! I have reinstalled the sw. Exim and spamd is running. What can posssible be wrong. Almost anything. I'd first try checking your configfiles for syntax errors: spamassassin --lint Then I'd try out spamc via the command line: spamc < samp

Re: [SAtalk] Some spams are slipping through without analysis?

2003-09-30 Thread Matt Kettler
At 05:53 PM 9/29/03 -0700, [EMAIL PROTECTED] wrote: Yes, I'm using spamc/spamd. AFAIK, the message is small so I don't think it's hitting the maximum scan limit. I'm running RH 7.3 with Sendmail 8.11.6-27 along with SA 2.60 and Razor 2.36. I'm calling SA from /etc/procmailrc with: Are you sure i

Re: [SAtalk] No tests run at all??

2003-09-29 Thread Matt Kettler
At 02:09 PM 9/29/03 +0300, Odhiambo Washington wrote: Sometimes I see an e-mail and looking at the full headers I see: X-Spam-Checker-Version: SpamAssassin 2.60 (1.212-2003-09-23-exp) on ns2.wananchi.com X-Spam-Status: No, hits=0.0 required=5.0 tests=none autolearn=no version=2.60 This is

Re: [SAtalk] Some spams are slipping through without analysis?

2003-09-28 Thread Matt Kettler
At 01:16 PM 9/28/03 -0700, Henry Kwan wrote: I've noticed that after upgrading to 2.60, a few spams are slipping through without being analyzed. Not a lot, probably only one in a few hundred, but it's a bit of a head scratcher. Anyone have any ideas? Do you use spamc/spamd? if so, is the message

Re: [SAtalk] Rule array?

2003-09-27 Thread Matt Kettler
At 06:05 PM 9/27/03 -0400, Mike Kuentz (2) wrote: Any one know if, without (or even with) writing an eval test, is there a way to make something similar to an array for tests that match? I have about 100 tests for h1d1ng w0rds w1th numb3r5 and l3tt3rs. I'd like to have something that would count

Re: [SAtalk] Did MSGID_GOOD_EXCHANGE get dropped in 2.60?

2003-09-27 Thread Matt Kettler
At 04:29 PM 9/27/03 -0400, Mike Kuentz (2) wrote: I can't seem to find it, but it's up on the rules page at spamassassin.org. Thanks, Mike Yes, it's dead.. it looks like the "tests" page is currently being generated from the 2.5x CVS head. In general never count on the "tests" page of the websit

Re: [SAtalk] Recent trend in spaming tactics

2003-09-27 Thread Matt Kettler
At 01:16 PM 9/27/03 +0200, Carlo Wood wrote: I noted this too. It doesn't seem like a good idea therefore that spamassassin comes with a pre-built whitelist; it makes it too easy for abusers to know which addresses to fake. Well, bear in mind however that the default whitelist entries are NOT mere

Re: [SAtalk] Recent trend in spaming tactics

2003-09-26 Thread Matt Kettler
At 04:51 PM 9/26/03 -0700, Chad Simmons wrote: I've noticed after installing Spamassassin, I'm still getting a large number of messages passed through without matching any of the rules. Many spammers are using paypal.com to get their emails "whitelisted". Also I've noticed that they will use in

Re: [SAtalk] Mail isn't spam, but subject line getting rewritten

2003-09-26 Thread Matt Kettler
At 07:09 PM 9/26/03 +0200, Spam wrote: For the past couple of months I have had some email subject lines being rewritten as "SPAM" even though the spam score is much lower than my threshold (currently set for myself at 4.1). It happens sporadically and I can't seem to see a pattern. I hav

Re: [SAtalk] Why is this microsoft_executable not detected?

2003-09-26 Thread Matt Kettler
At 03:18 PM 9/26/03 -0700, Ron Snyder wrote: I'm running SA 2.60-rc1, and have noticed that spamassassin is not catching microsoft executable attachments when they arrive as an enclosed bounce message. Shouldn't SA be detecting the microsoft_executable stuff since the test is done on rawbody? The t

Re: [SAtalk] 2.60 install problems

2003-09-26 Thread Matt Kettler
At 08:55 AM 9/26/03 -0400, Jack Gostl wrote: I'm trying to install 2.60 out of CPAN. I get the usual taint errors for AIX, but after a "force" install, I try to run spamd and get: Insecure directory in $ENV{PATH} while running with -T switch at /usr/opt/perl5/lib/5.6.0/Cwd.pm line 85. I

Re: [SAtalk] Auto-learn failed

2003-09-26 Thread Matt Kettler
At 03:33 AM 9/26/03 -0700, JP Kelly wrote: Why is SA looking in /dev/null/ ??? SA looks in the home directory for the user it runs as. In the case of spamd, if no user is ever specified, and it is started as root, it will fall back to the user "nobody" for security. Try passing the -u parameter

Re: [SAtalk] Disable relays.osirusoft.com checks

2003-09-26 Thread Matt Kettler
At 11:42 AM 9/26/03 +0200, Jonathan G - Mailing List wrote: Hi folks, i would like to know how can i disable the relays.osirusoft.com rbl checks form my spamassassin (v2.55 from CPAN) my config file (i provide it because i'm a newbie in SA) is in http://www.surestorm.com/data/SpamAssassin_loca

Re: [SAtalk] local.cf ignored by SA 2.6

2003-09-25 Thread Matt Kettler
At 04:26 PM 9/25/03 -0400, Dominic Germain wrote: I just set up a new server with SA 2.60. It's almost the same setup I already done on my "real" server. There is 2 differences: I use SA 2.60 instead of 2.55 and I use Perl 5.6.0 instead of 5.8.0... I can't figure why SA 2.6 ignore my /etc/mai

Re: [SAtalk] X-Spam-Status and content scoring disagree?

2003-09-25 Thread Matt Kettler
At 04:59 PM 9/25/03 -0400, Sean McCrohan wrote: This is my first week with SpamAssassin, so I may be missing something obvious, but the message I've attached sections of below puzzled me. While the body report scores it at 7.1 points, the X-Spam-Status in the header only gives it 1.8 - this is the

Re: [SAtalk] autolearning too much ?

2003-09-25 Thread Matt Kettler
At 11:16 AM 9/25/03 +1200, Simon Byrnand wrote: X-Spam-Checker-Version: SpamAssassin 2.60 (1.212-2003-09-23-exp) on xxx.xxx.xxx.xxx X-Spam-Level: X-Spam-Status: No, hits=-2.4 required=5.0 tests=BAYES_00,FORGED_HOTMAIL_RCVD2 autolearn=ham version=2.60 Notice that even though it matched FOR

Re: [SAtalk] Low score for HTML Message

2003-09-25 Thread Matt Kettler
At 03:36 PM 9/25/03 +0200, Céline REDON wrote: Hi, Could you please tell me if you facing the same: I am running SA 2.55 (Razor but no Bayes , no DCC) and I can see that quite all spam which are entirely written in HTML as an image are not tagged... Is it possible to recognize those spam as wel

Re: [SAtalk] Bayesian auto-learning? (and a test idea)

2003-09-25 Thread Matt Kettler
At 11:23 PM 9/24/03 -0500, Jeremy M. Dolan wrote: There's no further information on the auto-learning in the sa-learn or spamassassin man pages, or on the web page. What gives? man Mail::SpamAssassin::Conf should describe the autolearning by describing the options that control it. Also: are you

Re: [SAtalk] Capital letters onli in the test score?

2003-09-25 Thread Matt Kettler
At 01:18 PM 9/25/03 +0200, Jan Erik Skogsholm wrote: Hello! I have putthis string into may file /etc/mail/spamassassin/local.cf and the test work only for captal letters. Lower case lettesr do not work. Which string? In general rule names are always capitalized, is that what you are refering to

Re: [SAtalk] sa-learning

2003-09-24 Thread Matt Kettler
At 10:30 AM 9/24/03 -0400, Jon Fraley wrote: I just want to clarify this fact. I have seen contradictory statements posted. I am going to have users forward received spam and ham to respective accounts. Don't, unless they can forward the entire original message, complete with original headers, a

Re: [SAtalk] which works better for a spamassassin gateway amavisd-new or mailscanner?

2003-09-23 Thread Matt Kettler
At 08:26 PM 9/23/03 -0500, Chris Tatro wrote: Which has better integration (meaning easier to setup works well with) spamassassin? I've never tried amavisd, but I use MailScanner regularly. The only "drawback" of MailScanner is it's double-queue mechanism. This is nice, because your same old MTA

Re: [SAtalk] What gets auto learnt?

2003-09-23 Thread Matt Kettler
At 05:33 PM 9/23/03 +0200, Jim Knuth wrote: its wrong. Not working in V 2.55. But auto_learn_threshold_nonspam auto_learn_threshold_spam works fine. Why?? I'm sorry, I was quoting the 2.60 manpage, since no version was specified and it's what I had handy. The man Mail::SpamAssassin::Conf for 2.55

Re: [SAtalk] What gets auto learnt?

2003-09-23 Thread Matt Kettler
At 08:29 PM 9/23/03 +0700, Robert Nicholson wrote: So, just exactly where is the documentation that defines the criteria for what gets auto learnt? In short, the email is autolearned based on computing it's score as if bayes were disabled, and then comparing that score to a threshold you set. Th

Re: [SAtalk] Cannot change scores

2003-09-22 Thread Matt Kettler
At 11:15 PM 9/22/03 -0400, Jay R. Yablon wrote: Whenever I try to change the scores in the rules of configuration file, with, e.g., score DIET 7.0 score WEALTH 8.0 etc., Spam assassin stop schevking for that rile, i.e., it has the effect of setting it to zero. Using v 1.2.1 Help? Um, 1.2.1?

Re: [SAtalk] Philosophical SA questions

2003-09-22 Thread Matt Kettler
At 11:43 PM 9/22/03 +, Darren Madams wrote: Note in advance, I'm assuming sa 2.55 since this is a recent install. Philosophical question #1: Am I expecting too much to be disappointed with so many false negatives? I'm [obviously] nowhere near the numbers you guys are quoting. A lot of

Re: [SAtalk] upping BAYES_80 score

2003-09-22 Thread Matt Kettler
At 03:34 PM 9/22/03 -0700, Jonathan Nichols wrote: Ugh. How do I change the score again? I put: score BAYES_80 2.00 2.00 2.00 in local.cf - but it's not working. I'm trying to catch the stuff that's slipping through with only 4.4, etc. Thanks! try: score BAYES_80 0 0 2.00 2.00 Note there's _f

Re: [SAtalk] Could not create INET socket:

2003-09-22 Thread Matt Kettler
At 07:30 AM 9/22/03 +0100, TUNC ERESEN wrote: Hello, how could I fix this error ? [root /root]# spamd -D debug: Score set 0 chosen. debug: running in taint mode? no Could not create INET socket: Address already in use IO::Socket::INET: Address already in use Best Regards, Sounds like spamd is alr

Re: [SAtalk] Blacklist from

2003-09-22 Thread Matt Kettler
At 12:30 PM 9/22/03 +0200, Jim Knuth wrote: Can I add an entry " blacklist_to [EMAIL PROTECTED]" ?? in my local.cf? That there is no support for blacklist_to currently, however if my foggy memory serves me right, it was being added to 2.60 somewhere along the line. Most of the time if you want to

Re: [SAtalk] sa-learn

2003-09-22 Thread Matt Kettler
At 11:50 AM 9/22/03 +0200, =?ISO-8859-2?Q?Tom=E1=B9_Macek?= wrote: a] I have virtual user accounts and all my users have their accounts in the database. Their home directories/maildirs are somewhere in the /var. I'd like to know, if will make all the users via the webmail interface able to use th

Re: [SAtalk] All these Microsoft patch spams

2003-09-22 Thread Matt Kettler
At 06:35 AM 9/22/03 -0400, landy wrote: do you recommend a virus scanner for linux On the free-software side, look into clamAV. F-prot makes a very decent linux scanner on the commercial side and their small-business version covers most mailsever type usage at a reasonable price last time I chec

Re: [SAtalk] SpamAssassin filters seem too weak out of the box...

2003-09-22 Thread Matt Kettler
At 11:36 PM 9/21/03 -0700, Mike Klein wrote: Basically email consisted of an all caps subject "INCREASE YOUR PEN*S SIZE NOW!!!" and several lines in the body with same text and a url to go to. BTW, I didn't make the above typo in my email...I spelled the organ part correctly. Why is the rating s

Re: [SAtalk] All these Microsoft patch spams

2003-09-21 Thread Matt Kettler
At 12:34 PM 9/20/03 -0500, Philip Mak wrote: Can SpamAssassin do something about these messages? There's been a big flood of them recently, and they're all slipping through the filter. Well they're viruses from the sven worm, not spams, so your best bet would be to install a virus scanner and let

Re: [SAtalk] why it pass SA

2003-09-21 Thread Matt Kettler
At 06:34 PM 9/21/03 -0400, landy wrote: any idea why this pass thru SA Well, first, that message is not spam, it's a virus. It's the result of a newer mail worm/virus called the swen worm. SpamAssassin sometimes tags viruses, but that's not it's job in life. It's job is to identify spam and take

Re: [SAtalk] i would like to know how to score

2003-09-21 Thread Matt Kettler
At 07:17 PM 9/21/03 -0400, landy wrote: how does this sum 4.0, Not to be rude, but this is so easy you really should be able to answer it yourself. The scores of the rules are added together. So, looking at 50_scores.cf and yanking out the score lines we get: score FREE_TRIAL 0.0 0.0 0.0 0.017 s

Re: [SAtalk] Blacklist from

2003-09-21 Thread Matt Kettler
At 09:07 PM 9/21/03 +0200, Jim Knuth wrote: Hallo SA-List, how can I add an entry in the blacklist. I mean a faked to-address (this is not mine) The sender added like so? --snip spamassassin [EMAIL PROTECTED] --snap Thank you for help. That command adds it to the AWL system, and is NOT a static b

Re: [SAtalk] Editing scores

2003-09-12 Thread Matt Kettler
At 01:07 PM 9/11/2003 -0400, Alicia Forsythe wrote: I just cought an FP where the SUBJ_HAS_SPACES rule gave 3.5 points. My local.cf has no reference to this rule, but I did find it in 50_scores.cf: SUBJ_HAS_SPACES 2.425 2.026 1.101 2.329 I think you're looking in the wrong place. That message was

Re: [SAtalk] whitelist_from_rcvd

2003-09-12 Thread Matt Kettler
At 12:28 PM 9/9/2003 -0600, Alan Fullmer wrote: Am I doing something wrong? This is the statement I have in my list: whitelist_from_rcvd [EMAIL PROTECTED] xnote.com but it isn't working. do i have the syntax right? where [EMAIL PROTECTED] is

Re: [SAtalk] RCVD_IN_OSIRUSOFT_COM still in SpamAssassin version 2.60-rc3??

2003-09-12 Thread Matt Kettler
At 08:40 AM 9/11/2003 -0600, Ken Gordon wrote: I just noticed the following in a spamassassin report. I thought 2.60 (which I think I am running) didn't use this test. Am I wrong? Should I be zeroing it in local.cf? RCVD_IN_OSIRUSOFT_COM (0.9 points) RBL: Received via a relay in relays.osirus

Re: [SAtalk] What the F****??

2003-09-11 Thread Matt Kettler
At 10:36 PM 9/11/03 -0500, Larry Rosenman wrote: a SPAM message being learned as HAM? X-Spam-Flag: YES X-Spam-Checker-Version: SpamAssassin 2.60-rc4 (1.203-2003-08-29-exp) on lerami.lerctr.org X-Spam-Level: * X-Spam-Status: Yes, hits=5.5 required=5.0 tests=BAYES_99,HTML_MESSAGE

Re: [SAtalk] user_prefs

2003-09-11 Thread Matt Kettler
At 08:52 PM 9/11/03 -0600, Alan Fullmer wrote: So I have 2 files: /home/dick/.spamassassin/user_prefs and /home/jane/.spamassassin/user_prefs It does not even touch those user_prefs files (checked it with the -D option) Well, you've never told spamc to use jane or dick's userid.. so it's going

Re: [SAtalk] spamd performance

2003-09-11 Thread Matt Kettler
At 09:42 PM 9/11/03 -0400, Paul Farber wrote: New machine... dual 1.2Ghz AMD CPU's, 1 Gb RAM, Ultra-SCSI-3 (160Mbps) drive: 1) exactly what comands do you launch spamd with? 2) what exact command are you executing to process your messages in this test and time them? If spamc isn't part of the ans

Re: [SAtalk] user_prefs

2003-09-11 Thread Matt Kettler
At 08:03 PM 9/11/03 -0600, Alan Fullmer wrote: my options i have ./spamd -d -c No.. what entries are you trying to put into the user_prefs itself that aren't working? If it's a rule, well then read the manpage.. rules are ignored in user_prefs when using spamd for security reasons. If it's so

Re: [SAtalk] user_prefs

2003-09-11 Thread Matt Kettler
At 04:33 PM 9/11/03 -0600, Alan Fullmer wrote: I am having difficulty, and was wondering if anyone can shed some light on the subject. I can't get spam assassin to read any user_prefs file. it works with the local.cf just fine. What settings are you trying to put in user_prefs? How do you launch S

Re: [SAtalk] spamd performance

2003-09-10 Thread Matt Kettler
At 06:21 PM 9/10/03 -0400, Paul Farber wrote: Whats causing the 3X difference in msg scanning? Are you using _ANY_ network checks? DNS blacklists, razor2, dcc, pyzor, etc all have wildly varying times because they are heavily dependant on the load at the remote server and the load of every inter

Re: [SAtalk] bayes test is absent in some messages

2003-09-10 Thread Matt Kettler
At 12:52 PM 9/10/03 +, Michael V. Sokolov wrote: Why? The much of the "undecided" mid-range bayes has a zero score, thus never shows up. For example, if the bayes engine calls it an even 50/50 chance of being spam/nonspam, there's no points assigned. And that makes sense since bayes is more-

Re: [SAtalk] No BAYES_* ?

2003-09-10 Thread Matt Kettler
At 03:55 PM 9/10/03 +0200, Carlo Wood wrote: The attached mail got through spam-assassin without a problem... Why? Where is the BAYES_* test? Does this mean that the bayes engine thinks this is less than 10% chance to be spam? That would be ridiculous! How can I test on which keywords it is basin

Re: [SAtalk] [RD] Subject rule question

2003-09-09 Thread Matt Kettler
At 03:04 PM 9/9/2003 -0400, Mike Kuentz (2) wrote: Does SA translate that before hand and that's why it's not hitting? Yes, QP and base64 are decoded. They are decoded so that they can't avoid SA's normal subject rules so easily. body rules also get html tags stripped and end-of-lines removed pr

Re: [SAtalk] Not sure if my Bayesian filter is adding to the score ...

2003-09-09 Thread Matt Kettler
At 10:57 AM 9/9/2003 -0400, James Herschel wrote: Is it not taking the Bayes into consideration because I'm below 200 emails in the DB right now, or am I missing something here? Correct. SA has only been trained with 132 spam messages so far, and will not activate bayes until the database has been

Re: [SAtalk] RBL used??

2003-09-09 Thread Matt Kettler
At 06:00 PM 9/9/2003 +0200, Céline REDON wrote: Here is my config : SA Proxy Daemon 2.55 , Postfix I am currently testing SA with some users. I did not yet activate Razor /Dcc/Pyzor. But I am wondering if the tests are really all activated? and if the RBL are used ?? How can I be sure of it ??? F

Re: [SAtalk] sa-learn

2003-09-09 Thread Matt Kettler
At 10:54 AM 9/9/2003 -0500, Chuck Baker wrote: Why am I getting this since I upgraded to 2.60 and what can I do to remedy it? From the looks of it, the upgrade was somewhat abortive and didn't correctly upgrade.Those messages indicate SA code didn't get correctly installed and the old code is run

Re: [SAtalk] Barracuda Spam Appliance Review & Question

2003-09-09 Thread Matt Kettler
At 11:34 AM 9/9/2003 -0400, K Old wrote: Also, they tech guy from the company was able to connect to port 22, but how? We got a connection refused. For ssh to be able to validate a key it has to at least allow a connection. Does anyone have any idea how they achieve this kind of security? This ki

Re: [SAtalk] grrr - Re: stuff (fwd)

2003-09-08 Thread Matt Kettler
At 04:47 PM 9/8/03 -0700, Jonathan Nichols wrote: I've been getting a *lot* of these. Not always from "mail.ihyphen.com" but they *always* have PowerMail as the X-Mailer (legit mailer, IIRC) and they put "jnichols@(spamdomain).com" - these are the *only* types that get through at all. Any ideas on

Re: [SAtalk] Rbl checks

2003-09-08 Thread Matt Kettler
At 08:28 PM 9/8/03 -0400, Scott Kopel wrote: It's clear that my SA is now performing rbl checks and I apologize for being slow, but I still don't see how to configure my local.cf to get SA to perform checks at specific rbl lists eg spamcop.net I checked the suggested urls, but none of them shows

Re: [SAtalk] writing rules

2003-09-07 Thread Matt Kettler
At 09:05 AM 9/7/03 -0400, [EMAIL PROTECTED] wrote: 1 How do I write my own rules? is there a tutorial someplace? Yes there is a tutorial someplace, or a least a fairly decent howto on the subect. http://mywebpages.comcast.net/mkettler/sa/SA-rules-howto.txt 2 Spam changes with time, which means t

Re: [SAtalk] Performance optimization for bigger setups

2003-09-05 Thread Matt Kettler
At 04:56 PM 9/5/2003 +0200, Jochen Tuchbreiter wrote: Is there a chance that I may significantly increase performance by omitting some rules that contain costly regexps? Is there an easy way to find out how much time spamassassin spends on each regexp? Do you think that changing spamd/spamc so that

Re: [SAtalk] Error creating user_prefs

2003-09-05 Thread Matt Kettler
At 09:36 AM 9/5/2003 -0500, Segree, Gareth wrote: I am getting the following error from the spamd daemon when running from exiscan Creating default_prefs [//.spamassassin/user_prefs] Cannot write to //.spamassassin/user_prefs: No such file or directory Where can I change the location that spamass

Re: [SAtalk] Auto-whitelist explanation?

2003-09-04 Thread Matt Kettler
At 01:21 PM 9/4/2003 -0400, Pat Traynor wrote: I think I have a general understanding of how the auto-whitelist process works, and please correct me if I'm wrong. As you get "acceptable" emails from addresses (and/or ip addresses?), that address improves its "score". Then, if that address sends a

Re: [SAtalk] Mailscanner- spaassasin messages in the email main body

2003-09-04 Thread Matt Kettler
When using mailscanner, all of the message modifications are done by mailscanner itself and SA is only used to generate a score. Ask on the mailscanner list and someone there should be able to help you if nobody here knows the answer. I use mailscanner, but I'm running an older version so I'm n

Re: [SAtalk] Can base64 encoded messages learnable by sa-learn?

2003-09-04 Thread Matt Kettler
At 10:59 AM 9/4/2003 -0500, Thomas Cameron wrote: All - I have a client with a Red hat 9 + SA 2.55 + spamass-milter server in front of a Lotus Notes server. The client has hired a Notes developer to give their users an extra button in Notes to forward false negatives to a spam account the Linux s

Re: [SAtalk] Need to change 'required hits' but no local.cf file

2003-09-04 Thread Matt Kettler
At 09:41 PM 9/3/2003 -0400, Jason Lieurance wrote: I have a freebsd 4.7 mail server with qmail, courier-imap, and spamassassin 2.56 installed from the ports collection. The default of 'required_hits 5' I need to change to 7 or so. The trouble is I don't have a local.cf file and I have 100+ virtua

Re: [SAtalk] why so low

2003-09-04 Thread Matt Kettler
At 06:31 AM 9/4/2003 -0400, landy wrote: i have been getting many of these and even after doing sa-learn the score is super low, these emails are really pissing me of Are you *SURE* those aren't a GENUINE email related to your real ebay account? The fact that the message matches GENUINE_EBAY_RCVD

Re: [SAtalk] osirusoft still working?

2003-09-04 Thread Matt Kettler
At 12:35 PM 9/4/2003 -0400, Jim Porter wrote: I have added the following lines to my etc/mail/spamassassin/local.cf file. >From what I understood, this would disable rbl checking of osirusoft.com, but I am still seeing lines like this in my log file. Do you use spamd? if so, did you restart it?

Re: [SAtalk] Very Slow when using SpamAssassin

2003-09-03 Thread Matt Kettler
At 04:38 PM 9/3/03 -0700, Jesse Butler wrote: Now it filters spam effectively, but there seems to be a pause between each mail. The pause is perhaps 30 seconds long, taking forever to download and process 300+ mails This sounds like a DNS blacklist timeout.. Try a quick test running an email throu

Re: [SAtalk] Why tagged as spam when HITS < REQUIRED ??

2003-09-03 Thread Matt Kettler
At 09:47 AM 9/4/03 +1200, Steve Brorens wrote: On my experimantal 'gateway' configuration (Postfix/Amavisd-new/SA) I see a few instances where msgs get classified as spam tho the report shows the hits DO NOT exceed the required: ---snip form the report--- Content ana

Re: [SAtalk] comment out osirusoft ?

2003-09-03 Thread Matt Kettler
At 05:05 PM 9/3/03 -0700, J. Fowler wrote: Rather than defining separate cf files, could you simply comment out referrences to osirusoft in the 20_head_tests.cf? Thanks, Yes, The only reason they suggest a separate .cf file is that editing files in /usr/share/spamassassin is usually a VERY bad

Re: [SAtalk] FormMail messages getting caught

2003-09-03 Thread Matt Kettler
At 03:29 PM 9/3/2003 -0400, Pat Traynor wrote: When I have FormMail generate messages on my server, they're getting caught by SA. Here's what I'm getting: This is the unfortunate result of Formmail having been widely abused as an open relay by spammers. Your site appears to be using formmail 1.

Re: [SAtalk] Related to spamassasin paramters

2003-09-03 Thread Matt Kettler
At 10:03 AM 9/3/2003 -0700, swapna ghosh wrote: server [manager1 mail]$ sa-learn --spam --mbox spammails Failed to create default user preference file /home/sites/site61/users/manager1 .spamassassin/user_prefs Learned from 224 messages. _

Re: [SAtalk] Bayes filter and autolearning

2003-09-03 Thread Matt Kettler
At 07:52 PM 9/3/2003 +0200, Dave Kliczbor wrote: | X-Spam-Status: No, hits=4.1 required=5.0 | tests=BAYES_90,NO_REAL_NAME | autolearn=ham version=2.55 auto_learn_threshold_nonspam is set to 1.9. Now I am wondering why on earth SpamAssassin learns that message as _ham_? As far as I can s

Re: [SAtalk] osirusoft gone mad?

2003-09-03 Thread Matt Kettler
At 08:00 PM 9/3/2003 +0200, Ralf G. R. Bergs wrote: > SA reads any config files in /etc/mail/spamassassin automatically. Sorry, wasn't aware of this, it's not in the manpages (or is it?!). Yes, it is in the manpage for spamassassin. It's not entirely clear, but basically it will find the first "

Re: [SAtalk] Re: Must scores be listed in 50_score.cf?

2003-09-03 Thread Matt Kettler
At 09:19 AM 9/3/2003 -0400, Debbie D wrote: Thanks for that.. does the local.cf get over written also then when upgrading??? No, that's the reason why you want to do all customizations in your local.cf.. because it stays. Section 1.2 of the rules howto discusses this. http://mywebpages.comcast.n

Re: [SAtalk] spamassasing consuming lot of memory

2003-09-03 Thread Matt Kettler
At 03:21 PM 9/3/2003 +0530, Keerthi wrote: Hello All, I'm using spamassassin-2.55 with qmail. The spamassassin process is consuming lot of memory. Is there any bug related to memory leak in spamassassin. Not that I'm aware of, however the bayes engine does use a very large amount of memory. You

Re: [SAtalk] SA being called twice?

2003-09-03 Thread Matt Kettler
At 01:48 PM 9/3/2003 +0100, Angel Gabriel wrote: Is it possible for me to check if SA is being called twice? I call it in my procmailrc but, I think maybe postfix calls it too, because it takes an age for my email to be scanned Send a GTUBE email. If SA is being called twice and you're using the d

Re: [SAtalk] sa-learn

2003-09-03 Thread Matt Kettler
At 06:24 AM 9/3/2003 -0600, Karl Larsen wrote: Hearing people using "sa-learn" makes me curious. I do not have an application by that name. What is sa-learn? sa-learn is part of spamassassin 2.50 and newer. If you've got 2.50 or newer, it should be on your system. sa-learn allows you to t

Re: [SAtalk] osirusoft gone mad?

2003-09-03 Thread Matt Kettler
At 07:11 PM 9/3/2003 +0200, Ralf G. R. Bergs wrote: >Here are the lines to disable the rules. Add these in a file called >/etc/mail/spamassassin/no-osiru.cf. Where does SA know that it is supposed to read this file? I guess you need to either tell SA to read the file or to include the configurat

Re: [SAtalk] Question about how to make *ALL* mail go through SA.

2003-09-03 Thread Matt Kettler
At 12:07 PM 9/3/2003 -0400, Steven W. Orr wrote: Is there something I can do to make mail addressed to a list address go through SA? I suspect you'll need to integrate SA at the MTA layer instead of procmail to make this happen. Procmail is called as the mail is delivered to a local user, but tha

Re: [SAtalk] Bug: Failed to create default user preference file //.spamassassin/user_prefs

2003-09-02 Thread Matt Kettler
At 01:15 AM 9/1/2003 +0200, Carlo Wood wrote: > SpamAssassin uses ~/ to determine the home directory of the current user. > And yes, that is standardized and should be implemented by your OS. But that makes not much sense in the case of sa-learn. As is obvious from the "//.spamassassin/user_prefs",

Re: [SAtalk] garbage

2003-09-01 Thread Matt Kettler
At 12:16 PM 9/1/2003 -0500, David Dyer-Bennet wrote: I've been seeing more and more spam using sequences of garbage character strings; I'm guessing to get around Bayesian filtering? Also the obfuscating comments seem to contain garbage strings, but the existing rule finds those pretty well. Is anyb

Re: [SAtalk] plz help? question regarding to from and expressions.

2003-09-01 Thread Matt Kettler
At 02:33 AM 9/1/03 -0600, Alan Fullmer wrote: I guess essentially what I am trying to do is check the reply to: because I am getting a lot of spoofs there, and it whitelists it. has anyone found a way around this? Use whitelist_from_rcvd, not whitelist_from -

Re: [SAtalk] plz help? question regarding to from and expressions.

2003-09-01 Thread Matt Kettler
At 09:05 PM 8/31/03 -0600, Alan Fullmer wrote: I'm trying to learn how to write certain expressions. I've tried several ways to do this, but have come to the conclusion i am a noob at this. I am trying to write a rule that checks to see if the Return-Path: =equals= the To: if Return-Path == T

Re: [SAtalk] Bug: Failed to create default user preference file //.spamassassin/user_prefs

2003-08-31 Thread Matt Kettler
At 01:50 PM 8/31/2003 +0200, Carlo Wood wrote: What does it use to determine the users home directory? SpamAssassin uses ~/ to determine the home directory of the current user. And yes, that is standardized and should be implemented by your OS. -

Re: [SAtalk] upgrade spamassassin from 2.44 to 2.55 on redhat 9

2003-08-31 Thread Matt Kettler
At 02:47 PM 8/31/03 +0200, Andreas Mueller wrote: But NOTHING is filtered anymore? Why? First thing I'd suggest is running spamassassin --lint. This will tell you if spamassassin is unhappy with your user_prefs or local.cf and is spitting it out as a result. There are a few options that were pre

Re: [SAtalk] hardware recomendations

2003-08-30 Thread Matt Kettler
At 01:57 PM 8/30/03 +, [EMAIL PROTECTED] wrote: Until it died of hardware failure, I had a P150 with qmail, spamc/spamd, and a lightly loaded Apache handling about 6K messages/day. CPU load was always in the .2-.3 range, except after an ISP outage where load was at 50 to 60 for over an hour. We

Re: [SAtalk] New test proposal

2003-08-30 Thread Matt Kettler
At 05:31 PM 8/30/03 +0200, Kai Schaetzl wrote: why? Can you name a single nonspam mail sender who doesn't have a Date: header in the message when he delivers it to the MTA? No message-id is somewhat common for mass mailings (spam or nonspam) but I've never seen a nonspam that was also missing a

Re: [SAtalk] -d --lint scores ?

2003-08-30 Thread Matt Kettler
At 12:31 PM 8/30/03 +0200, Kai Schaetzl wrote: I'm talking of a specific message, the one you use for "spamassassin -D --lint" when I specify no message to process. I don't know how it looks like, so I don't know if it should score as ham or spam. But there apparently must be a message you use for

Re: [SAtalk] Bayes does work manually but not with spamc

2003-08-30 Thread Matt Kettler
Your problem is that when you execute spamc from exim, you are telling it to use a home directory of /tmp. This isn't likely the same home directory you have when you execute spamassassin manually. The bayes db is probably being yanked from /tmp/.spamassassin as a result, since ~/ is now /tmp/.

Re: [SAtalk] -d --lint scores ?

2003-08-30 Thread Matt Kettler
At 12:11 AM 8/30/2003 +0200, Kai Schaetzl wrote: I noted that I get quite different scores for spamassassin -D --lint on different machines depending on Bayes. On one system I get BAYES_10 (a well trained system with lots of mail in the db), on another system I get BAYES_70 (with only a few hundred

Re: [SAtalk] Turning on Bayes

2003-08-29 Thread Matt Kettler
At 03:13 PM 8/29/2003 -0500, David Dyer-Bennet wrote: I notice that there are different scores for tests depending on "how" spamassassin is running, and that one of the differences is whether Bayes filtering is turned on. But I've explicitly set use_bayes, and yet I never see a Bayes score turn u

Re: [SAtalk] hardware recomendations

2003-08-29 Thread Matt Kettler
At 03:15 PM 8/29/2003 -0400, Jon Fraley wrote: It will probably handle 3k to 5k messages daily. Does anyone have a recommendations on hardware. I have an old Pentium 166 that I thought about using. A p-166 strikes me as awfully light on CPU power for a mailserver running SA on 3-5k messages per

Re: [SAtalk] Min Score

2003-08-29 Thread Matt Kettler
At 12:12 PM 8/29/2003 -0600, Lucas Albers wrote: I just installed 2.60-rc3. I noticed that the minimum score I am getting on messages is 0.0. Is their a new default minimum score sa will assign email? The lowest score it will give for example, is 0.0? Ideas? There's never been any sort of minimum

Re: [SAtalk] Changing Bayes scoring

2003-08-29 Thread Matt Kettler
At 12:42 PM 8/29/2003 -0400, [EMAIL PROTECTED] wrote: and in order to increase the scores from certain bayes confidence levels, make edits, such as: body BAYES_70 eval:check_bayes('0.70', '1.80') body BAYES_80 eval:check_bayes('0.80', '1.90') body BAYES_9

Re: [SAtalk] defining the min/max score per test

2003-08-29 Thread Matt Kettler
At 09:30 AM 8/29/2003 +0100, Mark C wrote: For example in the above, if it finds anything in the body that indicats it came from the MSN Communities, it gives it a score of 0.0 (which if I am correct doesn't add anything to the total spam test score), whereas on the test marked: Q_FOR_SELLER, if th

Re: [SAtalk] Change points in AWL

2003-08-29 Thread Matt Kettler
At 02:12 AM 8/29/03 +0200, Jim Knuth wrote: In other words, the command-line spamassassin --add-to-whitelist [EMAIL PROTECTED] add to whitelist? *shame on me* No, spamassassin [EMAIL PROTECTED] Works that way... Spamassassin --add-to-whitelist expects an actual email message, not an email

Re: [SAtalk] Change points in AWL

2003-08-29 Thread Matt Kettler
At 12:25 AM 8/29/2003 +0200, Jim Knuth wrote: how can I change the points in awl.db? I have an wrong entry in this. (9.8 (9.8/1) -- [EMAIL PROTECTED]|ip=66.151) Look in the spamassassin manpage at the following command-line parameters. All are meant to help you manipulate the AWL database

Re: [SAtalk] Reporting *to* razor via SA

2003-08-28 Thread Matt Kettler
At 11:15 AM 8/28/2003 -0500, David Dyer-Bennet wrote: I seem to be looking incoming spam up in razor okay (I sometimes see hits on those rules in the reports). But when I send a spam into spamassassin --report, I always get an error that it was "unable" to report to Razor. In debug mode it's stil

Re: [SAtalk] filter identifying vaild word as masked

2003-08-27 Thread Matt Kettler
At 11:37 AM 8/27/2003 -0500, Robin Witkop-Staub wrote: body RAVEN_MaskedWordsF /\b(?:excIusive|GiangBiang|sIut|ganigbainged|duides|hairdciore|ExcIude|pIz)/ i describe RAVEN_MaskedWordsFmasked spam word(s) scoreRAVEN_MaskedWordsF10.0 1) I bet you'll find that the "plz" isn't an L, i

<    2   3   4   5   6   7   8   9   10   11   >