Upgraded to SA 2.61, deleted the bayes_* files and relearnt from my spam
and ham collections.
sa-learn --dump now reports 1513 spams and 195 hams so looks like it is
happy again.
Thanks Theo :)
--
Regards,
+-+-+
| Peter Kiem
remove
Greetings to everyone on the list!
I recently installed SpamAssassin 2.60 onto my new server, however, it
seems it's ignoring the whitelist_from and blacklist_from directives in
/etc/mail/local.cf. Another server we have, running SA 2.55, has these
directives working just fine. I tried
Anybody who's installed A 2.61
Can you tell me what the score for the following email was
From: Advance in Pay [EMAIL PROTECTED]
Subject: Need cash in the bank ASAP?
Date: December 12, 2003 4:28:24 PM CST
To: Robert David Nicholson [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
Jennifer, I have been testing your chickenpox.cf rules, and I noticed that
there is plenty of other punctuation marks besides the period . that are
also being used between the letters of works to obfuscate them. So I
modified you rules a bit from, for example:
/\s[a-zA-Z]{9}\.[a-zA-Z]{1}[
On Fri, 12 Dec 2003, Raquel Rice wrote:
I've tried several times to run spamc from a site-wide procmailrc.
It just hangs. This is what I have in the /etc/procmailrc:
:0fw
* 256000
| spamc
and I've made certain to have started spamd.
You don't say how you
Hi.
I've tried to run Spamassassin (spamc/spamd) with vpopmail on a per-user
basis by adding the following line to the .qmail-user file:
| spamc | /var/vmail/bin/vdelivermail '' bounce-no-mailbox
but unfortunately qmail refused to deliver any messages complaining in the
log:
qmail:
Sorry if this has been discussed over and over
again, but I've just started to run SA and I'm not happy with the way it's
mangling the tagged mail. I want it to change the subject but leave the
body of the message alone. I've changed report_safe from 1 to 0 and html
mail is still being
On a related note...I presume that when the bayesian classification runs
during normal operation, SA headers have not yet been added to the messages.
Does that imply one needs to be careful to feed non-filtered messages
(i.e., without the spam assassin filters already added) when training
At Sat Dec 13 14:02:53 2003, J. S. Greenfield wrote:
On a related note...I presume that when the bayesian classification runs
during normal operation, SA headers have not yet been added to the messages.
Does that imply one needs to be careful to feed non-filtered messages
(i.e., without
At Sat Dec 13 13:56:47 2003, Frank M. Cook wrote:
Sorry if this has been discussed over and over again, but I've just =
started to run SA and I'm not happy with the way it's mangling the =
tagged mail. I want it to change the subject but leave the body of the =
message alone. I've changed
Hi Bill
-Original Message-
From: [EMAIL PROTECTED]
[mailto:spamassassin-
[EMAIL PROTECTED] On Behalf Of Bill Landry
Sent: Saturday, December 13, 2003 7:04 AM
To: [EMAIL PROTECTED]
Subject: [SAtalk] chickenpox.cf
Jennifer, I have been testing your chickenpox.cf rules, and I
Attached is an example.
It would appear that I was wrong about double reply-to's. that's just
something that OSX's Mail.app is doing.
From [EMAIL PROTECTED] Fri Dec 12 16:28:13 2003
Received: (qmail 24655 invoked from network); 12 Dec 2003 22:28:12 -
Received: from d.dailye-mail.com
Robert Nicholson wrote:
Attached is an example.
It would appear that I was wrong about double reply-to's. that's just
something that OSX's Mail.app is doing.
This is what I got before teaching it to Bayes:
debug: is spam? score=5.621 required=5
At 08:56 AM 12/13/03 -0500, Frank M. Cook wrote:
Sorry if this has been discussed over and over again, but I've just
started to run SA and I'm not happy with the way it's mangling the tagged
mail. I want it to change the subject but leave the body of the message
alone. I've changed
So I wanted uninstall 2.60 which is installed in a version specific
directory on my setup.
bash$ ls
SpamAssassin SpamAssassin.pm
bash$ pwd
/home/robert/lib/perl5/site_perl/5.8.0/Mail
However when I do this with the Makefile used to build it I get
bash$ make uninstall
no packlist file found:
Hi,
Is this the correct place for SA to put installed when installed
locally and not system wide?
/home/robert/lib/perl5/site_perl/5.8.0/Mail
ie. does it belong in a 5.8.0 specific directory?
---
This SF.net email is sponsored by: IBM
Simon Byrnand wrote:
Just a bit of a me too post, I checked my last two days email including
Ham and Spam and checked the hitrate of DCC and RAZOR2 and here were the
results:
Ham: 0 DCC hits, 1 RAZOR2 hit out of 203 Ham messages.
Spam: 174 DCC hits, 57 RAZOR2 hits out of 242 Spam messages.
At 01:35 PM 12/12/2003, [EMAIL PROTECTED] wrote:
How can I see a spamassassin report of an email ? I'd like to do something
like pipe the email through a command.
pipe it through spamassassin -t
Note: -t mode FORCES sa to generate a report, and it WILL contain text
declaring it to be spam, even
Robert Nicholson wrote:
I noticed the scores you reported were with 2.60.
I don't understand why your scores as so much higher than mine.
Do you have any custom rule modifications?
I will take the message below and run it thru 2.61 now.
Why do you get more hits?
Your debug output
Robert Nicholson wrote:
Is the mail or checksum sent to a DCC server?
Checksum.
Could you try this one. It's also low scoring in 2.61
Here's the output I got:
Received: from localhost [127.0.0.1] by coll.pair.com
with SpamAssassin (2.60 1.212-2003-09-23-exp);
Sat, 13 Dec
Today I have dumped my bayes db and calculate some statistics.
742753 - total number of words in it
515654 - total number of words which have been seen only once
80485 - ... twice
35325 - ... 3 times
This statistics shows that most of the db us not used, just eating my hard drive (44
MB
Hi!
I suggested this once before, and did not see any response.
Many rules that I see suggested on this list all have the characteristic
of being a good test against e-mail that contain a large number of
occurences (a high 'count') of a particular 'trick' or 'obfuscation'.
BUT these rules have to
I would guess this is normal. Think of things like Message-Id's, vs. common
words like the which will appear very many times.
-Original Message-
From: Alexander Litvinov [mailto:[EMAIL PROTECTED]
Sent: Saturday, December 13, 2003 10:10 AM
To: [EMAIL PROTECTED]
Subject: [SAtalk]
Alexander Litvinov wrote:
Today I have dumped my bayes db and calculate some statistics.
742753 - total number of words in it
515654 - total number of words which have been seen only once
80485 - ... twice
35325 - ... 3 times
This statistics shows that most of the db us not used,
At 02:45 PM 12/13/2003, Bryan Hoover wrote:
742753 - total number of words in it
515654 - total number of words which have been seen only once
80485 - ... twice
35325 - ... 3 times
This statistics shows that most of the db us not used, just eating my
hard drive (44 MB total size). Is
Matt Kettler wrote:
Well, 5mb * (743/100) = 37.15mb... that's pretty close to 44mb at an
estimate. Doesn't seem large at all given the specs..
Heh.. True that. My brain is *not* a arithmatic calculator :).
Plus, Alexander may have been just counting up the total Bayes directory
size, which
Charles Gregory wrote:
Hi!
I suggested this once before, and did not see any response.
Many rules that I see suggested on this list all have the characteristic
of being a good test against e-mail that contain a large number of
occurences (a high 'count') of a particular 'trick' or
You could use procmail to call spamc, eg. .qmail-user files look like this:
| preline procmail -t -p ./the_users_name_here/Maildir/procmailrc
and the user's procmailrc:
LOGFILE=/var/log/procmail
VERBOSE=ON
### Spam Assassin
:0fw
* 256000
| /usr/bin/spamc -u the_users_name_here -f
:0
*
On Sat, 13 Dec 2003, Bryan Hoover wrote:
But if we were able to check the COUNT of how many times a particular
rule was matched, we could easily distinguish runaway use of obfuscation.
It is an interesting idea. It is analysis of the analysis, or meta
analysis.
Not really. It is counting
-Original Message-
From: Charles Gregory
Sent: Saturday, December 13, 2003 12:56 PM
[...]
On Sat, 13 Dec 2003, Bryan Hoover wrote:
But if we were able to check the COUNT of how many times a particular
rule was matched, we could easily distinguish runaway use of
obfuscation.
So I'm executing
my $status = $spamtest-learn ($mail);
and the status in the debugger shows all the right things suggesting
it's been learnt
however the nham count in sa-learn --dump isn't changed.
---
This SF.net email is sponsored by:
- Original Message -
From: Greg Cirino - Cirelle Enterprises [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Friday, December 12, 2003 2:33 PM
Subject: Re: [SAtalk] Virginia Busts Spammers
No noticeable decrease in spam here...
did they get the right guys?
g
Usually, 2-3 dozen
On 12 Dec 2003, [EMAIL PROTECTED] moaned:
On Thu, 11 Dec 2003 09:10:29 -0500, Adam Denenberg [EMAIL PROTECTED]
posted to spamassassin-talk:
What i want to start is a Bayes Corpus Project. I would like to be
able to allow people to submit confirmed ham and/or spam to a large
bayes
On Fri, 12 Dec 2003, Matthew van Eerde said:
I'm just curious. I've NEVER seen Yahoo! tagged bulk email
with SA tags.
Odds are the open relay that the spammers sent the mail
through was running SA.
Yes, you'd think anyone using SA would check for open relay,
but it does
happen..
Teergruber?
OK, I had to ask Google about this term of art:
http://www.iks-jena.de/mitarb/lutz/usenet/teergrube.en.html
From: Nix
Sent: Saturday, December 13, 2003 2:49 PM
[...]
It makes a lot more sense to teergrube the buggers and /dev/null the
results, I'd say.
(Anyone got a
The way this interface is defined
$status = $f-learn ($mail, $id, $isspam, $forget)
suggests that both $isspam and $forget are significant however
it will only ever do one or the other not both.
---
This SF.net email is sponsored by: IBM
At 06:53 PM 12/13/2003, Robert Nicholson wrote:
$status = $f-learn ($mail, $id, $isspam, $forget)
suggests that both $isspam and $forget are significant however
it will only ever do one or the other not both.
Well, forget inherently means you want to remove an entry instead of
learn.. at that
On 12/13/03 7:05 AM, "Vivek Khera" [EMAIL PROTECTED] wrote:
"JV" == Jonathan Vanasco [EMAIL PROTECTED] writes:
JV AOL IM
JV Maybe its an ichat only thing -- because everyone i know running iChat
JV gets 10+ AOL IM spams a day. I'm averaging 15.
zero for me,
After adding a bit of debugging information to Received.pm I was able to
track down the problem I was having with SpamAssassin and the -notfirsthop
rules. The following line is the code in spamass-milter that send the
pseudo header to Spamassassin:
assassin-output((string)Received: from +macro_s+
40 matches
Mail list logo