At Wed Jan 28 09:39:06 2004, jean-christophe valiere wrote:
> I agree with what you said, the probleme is that in the mail I attached the word
> have some swaped letters like:
> Hi,
>
> Genierc and Sepur Viarga (Caiils) available online!
> Most trusted online source!
>
>
>
On Wed, Jan 28 2004 (10:39:06 +0100), jean-christophe valiere wrote:
> Cilais or (Spuer Vagira)
> takes affect right away & lasts 24-36 huors!
> FOR SUEPR VAIRGA TOCUH HERE
>
> I wanted to know if somebody has already received such mail and if it was
> tagged as spam.
I get qui
On Tue, Jan 27, 2004 at 11:00:07AM -0500, Matt Kettler wrote:
> At 12:22 PM 1/27/04 +0100, jean-christophe valiere wrote:
>
> >Hi,
> >
> >I've got a little problem with the mail that is attached.
> >When I try spamassassin -t -D rulesrun=255 < mail.txt it is
> >nor c
At 12:22 PM 1/27/04 +0100, jean-christophe valiere wrote:
Hi,
I've got a little problem with the mail that is attached.
When I try spamassassin -t -D rulesrun=255 < mail.txt it is
nor considered as spam.
So I do spamassassin -r -D rulesrun=255 < mail.txt an
At 12:14 PM 1/20/2004, Kenneth Andresen wrote:
Will SA-learn filter all mails for everybody using the same rules, or
how can it work with different rule set for each user/mail account?
by default bayes databases and rulesets are specific to the user that
executes SA (note: that's execution, which
>
> You could use formail/procmail,
>
> formail -s procmail sa_learn.rc < mbox | sa-learn
>
> where sa_learn.rc might appear as follows:
>
>
> LOGFILE=$HOME/sa_learn.log # While debugging
> VERBOSE=yes# """"
> LOGABSTRACT=yes# """"
> SENDMAIL=
> From: Barton L. Phillips
> Sent: Saturday, January 17, 2004 9:28 AM
>
> I am using Mozilla and when I delete a message it is marked:
> Status: RO
> X-Status: D
>
> When I run sa-learn the deleted messages are learned. I can "compact
> this folder" but I sometimes forget. Is there a way to ha
Are you using the --mbox flag with sa-learn?
Brad
On Fri, 16 Jan 2004, Marcus Frischherz wrote:
> Hi!
>
> I use SA 2.61, exim 4.24 and WU imapd, I use IMAP4 and keep my mails on the
> server. The folders seem to be in the mbx format. When I run sa-learn on one of
> them it, sa-learn is busy f
In the latest versions of Guin there is a tools directory that should have EXPORTSPAM
http://www.beginfinite.com/html/gwen_manual_2/index.htm
create a shared folder that users put SPAM into, then extract this and feed your
database ... check out the above link to read up on it
>>> BJM <[EMAI
On Monday 05 January 2004 1:14 pm, Matt Kettler wrote:
[ ... ]
> Perhaps you should try to see if cat is modifying it. On some systems cat
> will wind up stripping certain non-printable characters. (Not that such
> charachters should be there.. however...)
Top tip! I've just tried that on my set
At 05:12 PM 1/4/04 -0800, Shane Wegner wrote:
My simplified testcase is as follows. I have a message in
mbox format saved as msg.3.
continuum:~$ sa-learn --debug-level --mbox --ham < msg.3
...
debug: bayes: Learned '[EMAIL PROTECTED]'
cat msg.3 | sa-learn --debug-level --mbox --ham
...
debug: baye
At 03:50 PM 12/29/03 -0800, Barton L. Phillips wrote:
Learned from xx message(s) (yy message(s) examined).
Sometimes the learned xx is less than the examined yy. My question is what
do these two numbers mean. I took a quick look at the code but it is quite
circuitous and I was hoping someone just
Jennifer Wheeler wrote:
Snipped
btw 42??? what did you mean by that. that was very creepy to see,
because i've tried to convince my brother from an early age, that the
number 42 *haunts* me and turns up *everywhere*! that'll either be a
very good year for me, or that's the year i'll buy t
> > btw 42??? what did you mean by that. that was very
> > creepy to see,
> > because i've tried to convince my brother from an early
> age, that the
> > number 42 *haunts* me and turns up *everywhere*! that'll
> either be a
> > very good year for me, or that's the year i'll buy the farm
> btw 42??? what did you mean by that. that was very
> creepy to see,
> because i've tried to convince my brother from an early age, that the
> number 42 *haunts* me and turns up *everywhere*! that'll either be a
> very good year for me, or that's the year i'll buy the farm per se!
http://
Hello there Rubin
>
> The ruleset name _was_ her idea 8^)
>
> I can see that my post could seem a little odd taken out of
> context, so let me clarify: Jenn's Backhair *ruleset* will
> help with the bogus html tags. I know nothing about Jenn's
> backhair. I must confess that I do, however, oc
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On
> Behalf Of Evan Platt
> Sent: Tuesday, December 23, 2003 6:26 PM
> To: Rubin Bennett
> Cc: SpamAssassin
> Subject: RE: [SAtalk] sa-learn from Exchange 2000
>
>
> --On T
The ruleset name _was_ her idea 8^)
I can see that my post could seem a little odd taken out of context, so
let me clarify:
Jenn's Backhair *ruleset* will help with the bogus html tags. I know
nothing about Jenn's backhair.
I must confess that I do, however, occasionally find myself pondering
(am
--On Tuesday, December 23, 2003 5:56 PM -0500 Rubin Bennett
<[EMAIL PROTECTED]> wrote:
> Jennifer's Backhair rules.
That sentence could be taken the wrong way... :)
Evan
---
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an
s anyone have any rules which are
> successfully blocking these messages?
>
> Tim Donahue
>
> > -Original Message-
> > From: Kang , Joseph S. [mailto:[EMAIL PROTECTED]
> > Sent: Tuesday, December 23, 2003 1:44 PM
> > To: Timothy Donahue; [EMAIL PROTECTED
AIL PROTECTED]
> Sent: Tuesday, December 23, 2003 1:44 PM
> To: Timothy Donahue; [EMAIL PROTECTED]
> Subject: RE: [SAtalk] sa-learn from Exchange 2000
>
>
> > -Original Message-
> > From: Timothy Donahue [mailto:[EMAIL PROTECTED]
> > Sent: Tuesday, December
> -Original Message-
> From: Timothy Donahue [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, December 23, 2003 12:36 PM
> To: [EMAIL PROTECTED]
> Subject: [SAtalk] sa-learn from Exchange 2000
>
> How can I process messages from our exchange server as spam?
> I have a whole bunch of messages
At 10:28 AM 12/23/2003, tobi wrote:
Use of uninitialized value in numeric gt (>) at
/usr/lib/perl5/5.6.1/i386-linux/DB_File.pm line 270.
Learned from 3 message(s) (45 message(s) examined).
Has anybody seen something like this before? I already tried to delete
the bayes data files but same result. O
At 10:06 PM 12/15/2003, [EMAIL PROTECTED] wrote:
I'm confused here. I have just installed Spamassassin on the server. I had
a folder with a few spams I'd recieved and wanted to feed them to sa-learn
so it would recognise these types of email as spam. But when I run
sa-learn --spam --mbox spamdump
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., wit
Set the -D and make sure it's reading the mbox right. I had a problem with
that. If it's running ok the spamassassin has already learned these mails.
It's smart enough not to learn the same or similar message more than once.
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECT
On Thu, 11 Dec 2003, Dan Tappin wrote:
[...]
> bayes_auto_expire 1
> bayes_learn_to_journal 0
>
> My maillog still shows 'permission denied' errors on the /var/root/.spamassassin
> folder. When I run sa-learn (as root) my bayes
> files are created in /var/root/.spamassassin/ but as user
On December 11, 2003 06:00 pm, Dan Tappin wrote:
> After setting-up spamassassin, RTFM and monitoring the list I am still very
> confused about the correct way to set-up and run spamassassin with bays
> (sa-learn).
>
> The basics:
>
> I am running Post.Office as a MTA on OS X client 10.2.8. Post.O
On Wed, 2003-12-10 at 23:38, David B Funk wrote:
> Just try completely removing those bayes_toks & bayes_seen files
> and do a 'sa-learn --rebuild'. It should take that bayes_journal file
> and use its data to create a new database.
Good to go, I think...
[EMAIL PROTECTED]:~/.spamassassin$ rm ba
On Wed, 10 Dec 2003, AthlonRob wrote:
> > Just for S&G, try doing a 'sa-learn --dump magic' and see if it
> > likes what it sees. If you cannot even --dump magic then it's
> > truly corrupted, no repair, just delete and start fresh.
>
> I got some funky output:
>
> [EMAIL PROTECTED]:~/.spamassassi
On Wed, 2003-12-10 at 21:41, David B Funk wrote:
> Hate to say it, but it looks like your database is hosed.
Well that's no fun. :-)
> Permissions are OK, it's looking at the correct files, locks are good,
> etc.
Ok...
> those 'File Exists' errors indicate that the DB_File library found
>
On Wed, 10 Dec 2003, AthlonRob wrote:
> On Wed, 2003-12-10 at 19:50, Adam Denenberg wrote:
> > in the same directory as the bayes DB files.
>
> Unfortunately, there are no .lock files in that directory.
>
> [EMAIL PROTECTED]:~/.spamassassin$ sa-learn --rebuild -DD
> debug: Final PATH set to: /usr
At 07:17 PM 12/10/03 -0800, AthlonRob wrote:
Cannot open bayes databases /var/amavis/.spamassassin/bayes_* R/O: tie failed:
...
Where might the lock files be?
Try /var/amavis/.spamassassin/bayes*.lock
---
This SF.net email is sponsored by: IB
On Wed, 2003-12-10 at 19:50, Adam Denenberg wrote:
> in the same directory as the bayes DB files.
Unfortunately, there are no .lock files in that directory.
[EMAIL PROTECTED]:~/.spamassassin$ sa-learn --rebuild -DD
debug: Score set 0 chosen.
debug: running in taint mode? yes
debug: Running in tai
in the same directory as the bayes DB files.
adam
- Original Message -
From: "AthlonRob" <[EMAIL PROTECTED]>
To: "Justin Mason" <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Wednesday, December 10, 2003 10:17 PM
Subject: Re: [SAtalk] sa-learn ..
On Wed, 2003-12-10 at 19:11, Justin Mason wrote:
> > [EMAIL PROTECTED]:~/.spamassassin$ sa-learn --rebuild
> > Cannot open bayes databases /var/amavis/.spamassassin/bayes_* R/O: tie failed:
> > Cannot open bayes databases /var/amavis/.spamassassin/bayes_* R/W: tie failed:
> > File exists
> > Cann
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
AthlonRob writes:
> Hello list-
>
> I've been using SA for a while, including the bayes stuff, training it
> with sa-learn every few weeks on the latest spam samples and ham I
> happen to have received.
>
> I just did... a root oops.
>
> I was runn
Good afternoon, Kris,
On Wed, 10 Dec 2003, Kris Deugau wrote:
> William Stearns wrote:
> > You should use _bounce_ or _redirect_, instead.
>
> Which, unfortunately, adds some new headers with most MUAs. :( Along
> with the extra set of Received: headers that go along with sending a
> message
William Stearns wrote:
> You should use _bounce_ or _redirect_, instead.
Which, unfortunately, adds some new headers with most MUAs. :( Along
with the extra set of Received: headers that go along with sending a
message (which you could probably work around).
The only way I've seen to get a mes
On Wednesday 10 December 2003 12:10 pm, Matt Kettler wrote:
> At 11:32 AM 12/10/2003, Larry Starr wrote:
> >My question regards scripts to ease processing of these mailboxes. Since
> > the messages are forwarded, from several different Email clients
> > (netscape, kmail, pine, AppleMail, etc), ext
Good afternoon, Larry,
On Wed, 10 Dec 2003, Larry Starr wrote:
> I currently have mimedefang (2.37) and spamassassin (2.60) running on a RH9
> mail gateway.
>
> Spamassassin is configured to block messages with a very high SA score and to
> tag and pass along everything else.
>
> I have two a
At 11:32 AM 12/10/2003, Larry Starr wrote:
My question regards scripts to ease processing of these mailboxes. Since the
messages are forwarded, from several different Email clients (netscape,
kmail, pine, AppleMail, etc), extracting the original message, for sa-learn
is proving to be non-trivial.
At Thu Dec 4 20:39:21 2003, Chris Thielen wrote:
>
> Vee Persaud said:
> > Another, hopefully not dumb, sa-learn question.
> >
> > I am quarantining any email that has a score of 8.5 to 15. Should I just
> > run sa-learn --spam on these messages ?
> >
>
> Sounds reasonable to me.
As long as yo
At Wed Dec 3 17:38:41 2003, Pedro Sam wrote:
>
> On December 03, 2003 12:07 pm, Tobin wrote:
> > I recently trained SA-Learn and its working very well. The question I
> > have is can I feed the 5 or 6 spams I get a day into SA-Learn without
> > have to give it 1000+ ham\spam? I see in the documen
Awesome, yes it works that way for me, too. I read the man page and
thought it had been deprecated
--On Saturday, December 06, 2003 10:38 AM +1000 Peter Kiem
<[EMAIL PROTECTED]> wrote:
So it defaults to folders already. So you would just need to do
sa-learn --spam /var/tmp/spam
and it s
OK, just tried it without the --dir and it still works.
further down in the man sa-learn you see this:
"Simply run this command once for each of your mail folders, and it will
'learn' from the mail therein."
So it defaults to folders already. So you would just need to do
sa-learn --spam /var/
> I have too many files to learn, so sa-learn craps out with "bad
> interpreter: Too many arguments". I am invoking 2.60 like this: sa-learn
> --spam -d /var/tmp/spam/*
Too many files in /var/tmp/spam which overflows the shell expansion
> I used to invoke 2.54 like this: sa-learn --spam --dir
Vee Persaud said:
> Another, hopefully not dumb, sa-learn question.
>
> I am quarantining any email that has a score of 8.5 to 15. Should I just
> run sa-learn --spam on these messages ?
>
Sounds reasonable to me.
--
Chris Thielen
Easily generate SpamAssassin rules to catch obfuscated spam phra
Vee Persaud said:
>
>> Please send the headers from both the first and the second messages.
>> Without more information all anyone can do is wildly guess.
>>
> BEFORE sa-learn
> X-Spam-Level: ***
> X-Spam-Checker-Version: SpamAssassin 2.60 (1.212-2003-09-23-exp) on
> rocco.ncr.pwgsc.gc.ca
> X
At 03:04 PM 12/3/2003, Vee Persaud wrote:
BEFORE sa-learn
From: "Vee Persaud" <[EMAIL PROTECTED]>
X-Spam-Status: No, hits=3.3 required=5.0 tests=AWL,HTML_20_30,HTML_MESSAGE
autolearn=no version=2.60
AFTER sa-learn
From: "Vee Persaud" <[EMAIL PROTECTED]>
X-Spam-Status: No, hits=3.0 requir
At 03:28 PM 12/3/2003, Vee Persaud wrote:
I was planning on just forwarding any suspect spam emails to an email
account on my SA server and running it through sa-learn. I guess that
this isn't the correct way to do it ???
No.. that is a *VERY* bad idea to do, because it will cause sa-learn to
l
Matt Kettler wrote:
> At 02:24 PM 12/3/2003, Vee Persaud wrote:
>> It got a score of 3.3. I ran sa-learn --spam on this message and
>> now I get a score of 3.0 if I resend the message. I'm a bit
>> confused. Did I do something wrong ???
>
> How did you "resend" the message?
>
> If you "resen
At 02:24 PM 12/3/2003, Vee Persaud wrote:
It got a score of 3.3. I ran sa-learn --spam on this message and now I
get a score of 3.0 if I resend the message. I'm a bit confused. Did I
do something wrong ???
How did you "resend" the message?
If you "resend" the message, merely by making a new
> Please send the headers from both the first and the second messages.
> Without more information all anyone can do is wildly guess.
>
BEFORE sa-learn
Delivered-To: [EMAIL PROTECTED]
Received: (qmail 18730 invoked by uid 197); 3 Dec 2003 18:59:54 -
Received: from [EMAIL PROTECTED] by rocco b
[EMAIL PROTECTED] wrote:
> I received an email on one of my systems that is not running SA and I
> decided to forward it to my SA system to see if it would get
> rejected. The subject was "Have amazing s\ex up to 20 times per day
> lugpsxw". They use the "/" in the body also to avoid spam detect
Vee Persaud said:
> I received an email on one of my systems that is not running SA and I
> decided to forward it to my SA system to see if it would get rejected.
> The subject was "Have amazing s\ex up to 20 times per day lugpsxw".
> They use the "/" in the body also to avoid spam detection.
--On Wednesday, December 03, 2003 2:24 PM -0500 Vee Persaud
<[EMAIL PROTECTED]> wrote:
> I received an email on one of my systems that is not running SA and I
> decided to forward it to my SA system to see if it would get rejected.
> The subject was "Have amazing s\ex up to 20 times per day lu
On December 03, 2003 12:07 pm, Tobin wrote:
> I recently trained SA-Learn and its working very well. The question I
> have is can I feed the 5 or 6 spams I get a day into SA-Learn without
> have to give it 1000+ ham\spam? I see in the documentation a reference
> to do something like this with the n
On Friday 28 November 2003 20:08, Erwan Le Moing wrote:
> my question is :
> if someone send an spam mail to [EMAIL PROTECTED], spamassassin will learn
> to consider spam to non-spam .
Set up a local dns server and use {spam,[EMAIL PROTECTED] or something like
that. Or use @127.0.0.1.
Of course t
On Fri, Nov 28, 2003 at 09:21:11PM +0100, Erwan Le Moing wrote:
> Matt Kettler wrote:
>
> >At 02:08 PM 11/28/2003, Erwan Le Moing wrote:
> >
> >>I've just configured Spamassassin on my postfix system and add
> >>autolearn :
> >>an address [EMAIL PROTECTED] where i send spam who pass through
> >>
At 02:08 PM 11/28/2003, Erwan Le Moing wrote:
I've just configured Spamassassin on my postfix system and add autolearn :
an address [EMAIL PROTECTED] where i send spam who pass through spamassassin
an address [EMAIL PROTECTED] where i send false positive mail
Just a warning:
Are you *sure* you are
Matt Kettler wrote:
At 02:08 PM 11/28/2003, Erwan Le Moing wrote:
I've just configured Spamassassin on my postfix system and add
autolearn :
an address [EMAIL PROTECTED] where i send spam who pass through
spamassassin
an address [EMAIL PROTECTED] where i send false positive mail
Just a warnin
-Original Message-
From: Robert Menschel [mailto:[EMAIL PROTECTED]
Sent: Thursday, November 27, 2003 7:27 PM
To: S. M. C. Butler
Cc: [EMAIL PROTECTED]
Subject: Re: [SAtalk] sa-learn question
Hello S.,
Thursday, November 27, 2003, 3:49:46 PM, you wrote:
SMCB> * sa-learn --show
Hello S.,
Thursday, November 27, 2003, 3:49:46 PM, you wrote:
SMCB> * sa-learn --showdots --spam ./Mail/spam_new
SMCB> and got
SMCB> .
SMCB> Learned from 0 message(s) (37 message(s) examined).
SMCB> Did this work? Seems like it didn't..
It looks to me l
At 03:49 PM 11/27/03 -0800, S. M. C. Butler wrote:
Ø sa-learn --showdots --spam ./Mail/spam_new
and got
.
Learned from 0 message(s) (37 message(s) examined).
Did this work? Seems like it didnt..
Yep, that means that SA has already learned all 37 messages i
At 04:47 PM 11/25/2003, Clive Dove wrote:
When I run spamassassin in kmail, I scoop uncaught mail into a folder
"zsa-learn-spam then when I run sa-learn, I use this command:
$ sa-learn --spam --showdots ~/Mail/zsa-learn-spam/cur/
It then processes all the messages in that directory, ignoring the o
On Tuesday 25 November 2003 14:31, Matt Kettler wrote:
> At 01:43 PM 11/25/2003, Dan Tappin wrote:
> >When I run sa-learn I only see the '1 message learned' output even thought
> >I forwarded multiple messages to that mailbox.
> >
> >Any ideas / comments?
>
> sa-learn --mbox
>
> unless you add the
At 01:43 PM 11/25/2003, Dan Tappin wrote:
When I run sa-learn I only see the '1 message learned' output even thought
I forwarded multiple messages to that mailbox.
Any ideas / comments?
sa-learn --mbox
unless you add the --mbox parameter, sa-learn expects a single rfc822
format message per file
At 10:58 21/11/2003 -0800, Ted Cabeen wrote:
I'm getting a crash and core-dump when I try to run
sa-learn -D --force-expire on my sizeable bayes database.
Here's the output I get, and the backtrace from the core file.
debug: Score set 0 chosen.
debug: running in taint mode? no
debug: using "/usr/l
aniel Sterling
> > Subject: Re: [SAtalk] sa-learn scripts for mbox files
> >
> >
> > At 06:40 AM 11/23/03 -0500, Pedro Sam wrote:
> > >On November 23, 2003 04:46 am, Daniel Sterling wrote:
> > > > sa-learn
> > > does not split messages apart itsel
At 06:40 AM 11/23/03 -0500, Pedro Sam wrote:
On November 23, 2003 04:46 am, Daniel Sterling wrote:
> sa-learn
does not split messages apart itself (at least in version 2.60).
sa-learn --mbox ?
I second pedro's motion.. sa-learn DOES accept mbox files and split them
appart itself, and has done so
On November 23, 2003 04:46 am, Daniel Sterling wrote:
> I wrote some small scripts to take an mbox file (like you would find in
> /var/spool/mail), split the messages apart, and send them to sa-learn.
> sa-learn
does not split messages apart itself (at least in version 2.60).
sa-learn --mbox ?
-
At 03:36 PM 11/20/03 +, Carlos Jorge Santos wrote:
The header X-Spam-Status shouldn't cause any trouble to sa-learn, right?
What about the other header ? Is there any problem removing all the
Received headers ? Something like this :
bayes_ignore_header Received
No.. X-Spam-Status is a standa
On Thu, 20 Nov 2003, Carlos Jorge Santos wrote:
> Thanks a lot for your answer.
>
> The problem now is that SpamAssassin (spamc to be more precise) is
> called by Qmail-Scanner, which in turn adds this headers to emails:
>
> Received: from [EMAIL PROTECTED] by mail.host-services.com by
> uid 101 w
>From the sa-learn man page:
"If the messages you are learning from have already been filtered
through SpamAssassin, the learner will compensate for this. In effect, it
learns what each message would look like if you had run `spamassassin -d'
over it in advance."
On Thu, 20 Nov 2003 12:39
Thanks a lot for your answer.
The problem now is that SpamAssassin (spamc to be more precise) is
called by Qmail-Scanner, which in turn adds this headers to emails:
Received: from [EMAIL PROTECTED] by mail.host-services.com by
uid 101 wi
(spamassassin: 2.60. Clear:RC:0:SA:1(11.3/8.0):.
Proce
At 12:39 PM 11/20/03 +, Carlos Jorge Santos wrote:
We're in the process of feeding sa-learn with spam and ham so we can use
The Bayes filters.
But a doubt has crossed our mind... All of our Spam has the SpamAssassin
header, feeding this mails to sa-learn will make the Bayes database thins
th
At Sat Nov 15 16:34:09 2003, Alan Munday wrote:
>
> Martin
>
> Thanks for the response. It lead to another question
>
> If I'm using Anomy (as I am) do I need to keep a copy of the raw
> original or undo the anomy defang'ing?
I don't know anything about Anomy. I'd say that you'll need to
dford
Sent: 15 November 2003 14:07
To: Alan Munday
Cc: [EMAIL PROTECTED]
Subject: Re: [SAtalk] SA-Learn. Clarification question.
At Fri Nov 14 23:03:42 2003, Alan Munday wrote:
>
> Having just read the FAQ's can I just check the requirements for the
> source SPAM/HAM mail for sa-lea
At Fri Nov 14 23:03:42 2003, Alan Munday wrote:
>
> Having just read the FAQ's can I just check the requirements for the
> source SPAM/HAM mail for sa-learn?
>
> Do the contents of the sources messages need to be false positives
> for the HAM file and false negatives for the SPAM file?
No. To w
On Fri, Nov 14, 2003 at 10:25:54AM -0500, Matt Kettler wrote:
> Yes. SA won't even try to use bayes scoring until 200 spam AND 200 ham
> messages are trained.
... not to mention that only learning spam means bayes will be skewed
to think everything is spam.
--
Randomly Generated Tagline:
Uh, so
At 05:49 AM 11/14/03 -0500, MIKE YRABEDRA wrote:
Is it necessary to have SA bayes learn 'ham' as well as 'spam'?
Yes. SA won't even try to use bayes scoring until 200 spam AND 200 ham
messages are trained.
---
This SF.Net email sponsored by: A
on 11/6/03 10:50 AM, Upwood, Jim at [EMAIL PROTECTED] wrote:
> We are set up the same way. This is what I do:
>
> Have a folder called "ham" in my outlook
> I periodically drag ham to that folder
> I then have an outlook rule that I only run on demand
> This rule is set to "redirect" all message
This sounds like my earlier problem trying to sa-learn "large" mailboxes.
Others here commented that it "must be a hardware problem," but my box is
fairly new with very decent components (built from scratch by me but with
"good" MB and memory, etc.) and nothing else I do, including using X, gives
My guess - if you examine the headers of your spam messages you will see
that some message are 'autolearned'. Message that generate a high enough
'spam value' get internally fed to 'sa-learn' and are marked with
'autolearn=spam' in the X-Spam-Status header. Once a message has been
trained via sa-le
> Learned from 304 message(s) (30419 message(s) examined)
> Learned from 253 message(s) (30430 message(s) examined)
> Learned from 301 message(s) (30452 message(s) examined)
I bet you are seeing some spam re-learning after it has fallen off of
the expiry backend.
so the line:
Learned from 253 mess
On Sat, Nov 08, 2003 at 11:18:46PM -0500, Jack Gostl wrote:
> I'm trying to find the token counts. In version 2.55, what is the
> equivalent of:
>
> sa-learn --dump magic
there is no direct equivilent, but you can run the check_bayes_db (I
can't even remember what the full name is) script i
ehalf Of Giles
Coochey
Sent: Thursday, November 06, 2003 5:21 AM
To: Ryan Parlee; [EMAIL PROTECTED]
Subject: RE: [SAtalk] sa-learn via email
As has been discussed previously, when forwarding an email within most MUAs
you lose a lot of the header information of the original email, and you will
also tea
On Thu, 2003-11-06 at 11:49, Evan Platt wrote:
> --On Wednesday, November 05, 2003 10:29 PM -0500 Bill Baker
> <[EMAIL PROTECTED]> wrote:
>
> > I received 3 identical spams from the same e-mail address that did not
> > get classified as spam. I tried to do a sa-learn --spam on the message,
> > an
A few days ago I posted that basic script I used to feed emails into
bayes via email, there is a small bug in that the first one will work
but ones afterwards won't because ripmime doesn't overwrite the files.
If you haven't already fixed it, just change the script so it either
clears the /tmp/
--On Wednesday, November 05, 2003 10:29 PM -0500 Bill Baker
<[EMAIL PROTECTED]> wrote:
> I received 3 identical spams from the same e-mail address that did not
> get classified as spam. I tried to do a sa-learn --spam on the message,
> and it said it learned from 1 message, but when I tried to
iles Coochey [mailto:[EMAIL PROTECTED]
Sent: Thursday, November 06, 2003 8:21 AM
To: Ryan Parlee; [EMAIL PROTECTED]
Subject: RE: [SAtalk] sa-learn via email
As has been discussed previously, when forwarding an email within most
MUAs you lose a lot of the header information of the original emai
As has been discussed previously, when forwarding an email within most
MUAs you lose a lot of the header information of the original email, and
you will also teach bayes that spam looks like the headers that are
produced by your MUA. This happens to be a pretty bad idea.
I've been wondering about
On Fri, 31 Oct 2003, Mike Hyde wrote:
> I am running into the following running sa-learn --dump:
>
> Use of uninitialized value in numeric lt (<) at
> /usr/lib/perl5/site_perl/5.8.1/Mail/SpamAssassin/BayesStore.pm line
> 1281.
> 0.000 0 0 0 non-token data: bayes db vers
At 11:02 29/10/2003 -0600, Bill Polhemus wrote:
I am
running SA 2.60 installed from the RPMs on Red Hat 9, on an AMD 2100+
based system with a half-gig of RAM.
This has now happened for the second time.
Before when it happened, about two weeks ago, I figured it was just a
coincidence. Now, Im
On Wed, Oct 29, 2003 at 11:58:35AM -0800, Justin Mason wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
>
> Bill Polhemus writes:
> > I am running SA 2.60 installed from the RPMs on Red Hat 9, on an AMD 2100+
> > based system with a half-gig of RAM.
>
> Could you post the output of "rp
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Bill Polhemus writes:
> I am running SA 2.60 installed from the RPMs on Red Hat 9, on an AMD 2100+
> based system with a half-gig of RAM.
Could you post the output of "rpm -qa"? And you're not using any
hand-compiled components, it's all RPMs, right
First of all, please don't post in HTML.
Bill Polhemus was manually quoted as having said:
> I am running SA 2.60 installed from the RPMs on Red Hat 9, on an AMD
> 2100+ based system with a half-gig of RAM.
Personally, I've avoided RH > 7.3 for server work- RH8 and RH9 have seen
any number of rea
At 10:35 AM 10/22/2003, Anne Ramey wrote:
when I sa-learn on a folder of spam it takes almost no time and says
"learned 1" no matter how many where in the file but when I do an sa-learn
of ham, it works fine. Ideas?
What command line did you use? and what format is the mail in?
If you're giving
On October 22, 2003 10:35 am, Anne Ramey wrote:
> when I sa-learn on a folder of spam it takes almost no time and says
> "learned 1" no matter how many where in the file but when I do an
> sa-learn of ham, it works fine. Ideas?
If the folder is of mbox format, you may want to specify the "--mbox"
1 - 100 of 227 matches
Mail list logo
