Re: Procmail for site wide usage

2005-07-22 Thread jdow
There generally is no specific procmail log file. It is generally in one
of the mail log files in /var/log/wherever.

{^_^}
- Original Message - 
From: Thomas Arend [EMAIL PROTECTED]





Re: Procmail for site wide usage

2005-07-22 Thread jdow
You are developing a severe stutter.
{o.o}
- Original Message - 
From: Thomas Arend [EMAIL PROTECTED]
To: users@spamassassin.apache.org
Sent: 2005 July, 21, Thursday 20:40
Subject: Re: Procmail for site wide usage





Re: URIDNSBL and subdomains

2005-07-22 Thread Charles Sprickman

On Thu, 21 Jul 2005, Loren Wilton wrote:


Sounds like an surbl problem if spamsite.com isn't listed.


That's just an example I made up... :)

The leading subdomains are supposed to be trimmed off, since they are 
usually identifying strings for a given spam target rather than an 
actual part of the target name.


OK, so that's supposed to happen.  Is there any way to have the entire 
host checked?  I've seen a good volume of junk where the domain is clean, 
but if I do a manual lookup on the entire hostname in the spam it is 
indeed listed.


Thanks,

Charles

There are a few cases where things go 
to three levels rather than just two, but they are exceptions.


   Loren


- Original Message -
From: Charles Sprickman [EMAIL PROTECTED]
To: users@spamassassin.apache.org
Sent: Thursday, July 21, 2005 7:28 PM
Subject: URIDNSBL and subdomains



Hello,

I've been watching some of the misses that have passed through
spamassassin (3.0.4) lately and they are pretty clean; no DNS BL hits,
etc.

One thing I did notice is that many of them have a fairly contorted URL
for the spamvertized products, ie:

kjekliennxiffiennnkenc.spamsite.com

This doesn't trigger any URIDNSBL hits, but if I punch the entire URI into
the surbl.org checker it does hit.  It seems as if the SA check is looking
only at the domain part and not the subdomain.

Is this expected?  Is there a switch to flip to get the whole hostname
checked?

Thanks,

Charles

___
Charles Sprickman
NetEng/SysAdmin
Bway.net - New York's Best Internet - www.bway.net
[EMAIL PROTECTED] - 212.655.9344





Re: Procmail for site wide usage

2005-07-22 Thread jdow
Never mind - Earthlink had an email stick in its craw or else Fetchmail
did not like it at all.
{^_^}
- Original Message - 
From: jdow [EMAIL PROTECTED]
To: users@spamassassin.apache.org
Sent: 2005 July, 21, Thursday 23:16
Subject: Re: Procmail for site wide usage


 You are developing a severe stutter.
 {o.o}
 - Original Message - 
 From: Thomas Arend [EMAIL PROTECTED]
 To: users@spamassassin.apache.org
 Sent: 2005 July, 21, Thursday 20:40
 Subject: Re: Procmail for site wide usage
 
 



Re: URIDNSBL and subdomains

2005-07-22 Thread Loren Wilton
 OK, so that's supposed to happen.  Is there any way to have the entire
 host checked?  I've seen a good volume of junk where the domain is clean,
 but if I do a manual lookup on the entire hostname in the spam it is
 indeed listed.

I *suspect* what is happening here is that the domain isn't in surbl when
the mail comes through, but it is 15 minutes later when you check.

The alternative is that there is a difference in the two-level/three-level
split decision between SA 3.0.4 and surbl.  Neither works with the full
domain; that was a decision from the start at surbl.  The split is based on
the tail of the domain name.  It is (I think) a little more than the TLD
that goes into the decision, but that is the basic concept.

For instance, dribble.spammy.com will only check spammy.com.
dribble.spammy.co.uk will check spammy.co.uk.  The idea is to find the
real host name denuded of any directory names.

Sometimes this will deliberately miss a spamsite.  For instance,
myspamsite.yahoo.com will not be listed, because yahoo.com isn't a pure spam
host.  Instead, a request will be sent to Yahoo to get the site removed.

Jeff will likely be along sooner or later and can give considerably more
detail on how all of this really works.

Loren



Re: Procmail for site wide usage

2005-07-22 Thread Thomas Arend
Am Freitag, 22. Juli 2005 08:15 schrieb jdow:
 There generally is no specific procmail log file. It is generally in one
 of the mail log files in /var/log/wherever.

Yes. But you can create user user specific lof file with 
LOGFILE=$HOME/.procmail.log

Thomas

-- 
icq:133073900
http://www.t-arend.de


pgpNKHB3Ygxln.pgp
Description: PGP signature


Bayes poisoning ?

2005-07-22 Thread Ramprasad A Padmanabhan
Hi
  We are using Spamassassin + Postfix + Mailscanner on our SMTP servers.
Of late I have noticed that a lot of ham mails are getting a high BAYES
score.

I have overriden bayes with lower scores in order to avoid false
postives ( and possibly mail loss ) 

How do I de-poison the bayes database, and are there any ways to avoid
bayes poisoning ? 


Thanks
Ram




--
Netcore Solutions Pvt. Ltd.
Website:  http://www.netcore.co.in
Spamtraps: http://cleanmail.netcore.co.in/directory.html
--


Re: Bayes poisoning ?

2005-07-22 Thread Loren Wilton
The best thing to do is probably throw the current database away and start
over.  As you seem to have several users, you should have bayes working
again within a very few hours, or less.

You should delete the current database, reset the scores to normal (and
increase the bayes_99 score to something around 4 if you aren't using
3.0.4), and then manually train Bayes on a few hundred known ham and spam
before letting autolearning take over.

The other thing you should do is decrease bayes autolearn ham threshold to 0
or even -.1 or so.  By default it is too high, and will far too often lead
to bayes poisioning if the state of the database isn't watched carefully.
You may also want to take the bayes autolearn spam threshold up to a higher
value than it has by default; although this usually isn't required.

Loren



Re: URIDNSBL and subdomains

2005-07-22 Thread List Mail User
...
On Thu, 21 Jul 2005, Loren Wilton wrote:

 Sounds like an surbl problem if spamsite.com isn't listed.

That's just an example I made up... :)
...

Bad choice of example:  spamsite. com is an actual spamsite.
The domain example.com is reserved for exactly this type of usage
and should generally be used when speaking of generic domains;  Also,
the notation domain.tld is well understood to be an example of a
generic domain also.

In general listing the real site causing problems is good for
everyone else on the list - some will block it, others will take more
extreme action:  Letting everybody know is only rarely a bad thing.


Paul Shupak
[EMAIL PROTECTED]


Detecting ISO Encoded Subjects

2005-07-22 Thread Joseph D. Wagner
A lot of my spam lately has had an ISO encoded subject line, like:
=?iso-8859-1?B?T2ZmaWNlIHNvZnR3YXJlIC0gNzUlIE9GRg==?=

Since none of my friends ever use ISO encoded subject lines, I wanted to create 
a rule to flags those messages.  However, everything I've found indicates that 
Spamassassin tests the decoded subject line, not the raw subject line.

I've tried:
  header SUBJECT_ISO_ENCODED Subject =~ /=?iso-8859/i
but that doesn't seem to work.

Is there any way I can run a rule on the raw header to detect whether or not 
it's encoded?  If so, could you please show me how?

TIA.

Joseph D. Wagner


Re: Detecting ISO Encoded Subjects

2005-07-22 Thread Duane Hill

On Friday, July 22, 2005 at 12:32:34 PM, [EMAIL PROTECTED] confabulated:

 A lot of my spam lately has had an ISO encoded subject line, like:
 =?iso-8859-1?B?T2ZmaWNlIHNvZnR3YXJlIC0gNzUlIE9GRg==?=

 Since none of my friends ever use ISO encoded subject lines, I
 wanted to create a rule to flags those messages.  However,
 everything I've found indicates that Spamassassin tests the decoded
 subject line, not the raw subject line.

 I've tried:
   header SUBJECT_ISO_ENCODED Subject =~ /=?iso-8859/i
 but that doesn't seem to work.

  From a previous message to the list I dug up:

  header RULE_NAME HeaderName:raw ~=//. (3.0.0 or higher) Checks header without
  any QP or base64 decoding. Best for looking for illegal unencoded
  characters in...

 Is there any way I can run a rule on the raw header to detect
 whether or not it's encoded?  If so, could you please show me how?

 TIA.

 Joseph D. Wagner


---
 Duane Hill  |  Network Operations  - YourNetPlus.Com, Inc.
 | E-mail Administrator - [EMAIL PROTECTED]
---
This message is made of 100% recycled electrons.



Re: Procmail for site wide usage

2005-07-22 Thread Chris Barnes
Mark Williams [EMAIL PROTECTED] wrote:
 (Q) Given that this RH machine runs only POP3 (management will not
 allow anything else)

This is really the key - from a SA standpoint, the best you can do is 
mark the message as spam and let the MUA (Outlook) deal with putting 
things into the proper folders on the user's machine (in the .pst file).

I don't know OL well enough, but I suspect that there is likely a 
registry hack you can do or a rule you can create that the users can 
import that will look at the headers and put the message into the proper 
folders.

-- 

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
Chris Barnes   AOL IM: CNBarnes
[EMAIL PROTECTED]Yahoo IM: chrisnbarnes





behavior with amavisd-new-20030616-p10

2005-07-22 Thread Jonathan Gonzalez

Hi,

i would like to know if there's some special feature that needs to be 
enabled/disabled in order to work 100% compliant with amavis.


I'm my actual installation (Exim 4.50 + amavisd-new-20030616-p10 + 
ClamAV 0.86 + SpamAssassin 3.0 on a Debian Sarge platform) the MTA 
passes the email to amavis that checks for antivirus correctly but about 
the spam checks and it's behavior i'm not sure that all works fine.


I can see in the amavis logs that the spam checks are being made but no 
headers are added to the body of the email before being delivered to the 
user Mailbox.


This X-Headers are for me quite important because further tasks are 
based on those headers.


In example, i can see that a email is classified above the spam limit 
(11 points) and the message is delivered anyway. Seems that the 
interaction between MTA and amavis is not good, or between SA and amavis 
is not good maybe.


Any ideas will be welcomed.
Thanks in advance,

jonathan


Re: behavior with amavisd-new-20030616-p10

2005-07-22 Thread Matt Kettler
Jonathan Gonzalez wrote:
 Hi,
 
 i would like to know if there's some special feature that needs to be
 enabled/disabled in order to work 100% compliant with amavis.
 
 I'm my actual installation (Exim 4.50 + amavisd-new-20030616-p10 +
 ClamAV 0.86 + SpamAssassin 3.0 on a Debian Sarge platform) the MTA
 passes the email to amavis that checks for antivirus correctly but about
 the spam checks and it's behavior i'm not sure that all works fine.
 
 I can see in the amavis logs that the spam checks are being made but no
 headers are added to the body of the email before being delivered to the
 user Mailbox.
 
 This X-Headers are for me quite important because further tasks are
 based on those headers.

Amavis does it's own message tagging, and does not keep SA's markups. It
consults spamassassin merely as a scanner.

I'm not an amavis expert, so I can't tell you how, but I can tell you that it's
amavis you need to reconfigure if you want these added, as only amavis will
generate them.


Re: behavior with amavisd-new-20030616-p10

2005-07-22 Thread Matt Kettler
Matt Kettler wrote:
 Amavis does it's own message tagging, and does not keep SA's markups. It
 consults spamassassin merely as a scanner.
 
 I'm not an amavis expert, so I can't tell you how, but I can tell you that 
 it's
 amavis you need to reconfigure if you want these added, as only amavis will
 generate them.

As a follow-up to myself, I found it in the amavis FAQ:

http://www.ijs.si/software/amavisd/#faq-spam


Amavis ONLY adds the x-spam-* headers when the score is above the tag level

Note that tag level isn't when amavis considers a message to be spam, that's
tag2 level (brilliantly clear, no?)






Re: behavior with amavisd-new-20030616-p10 - SOLVED!

2005-07-22 Thread Jonathan Gonzalez

:-) Funny!

I have been reading, quite quick :) and have found the error. Was a 
error of mine. I didn't populate a variable with my domains, because 
such variable indicates amavis to put or not put the X headers.


The variable is @local_domains_acl = qw( .example.com .example2.com );

Once populated X-headers started to be written.

Thanks for your help :)
BR,

jonathan

Matt Kettler wrote:

Matt Kettler wrote:


Amavis does it's own message tagging, and does not keep SA's markups. It
consults spamassassin merely as a scanner.

I'm not an amavis expert, so I can't tell you how, but I can tell you that it's
amavis you need to reconfigure if you want these added, as only amavis will
generate them.



As a follow-up to myself, I found it in the amavis FAQ:

http://www.ijs.si/software/amavisd/#faq-spam


Amavis ONLY adds the x-spam-* headers when the score is above the tag level

Note that tag level isn't when amavis considers a message to be spam, that's
tag2 level (brilliantly clear, no?)







Re: Procmail for site wide usage

2005-07-22 Thread Kelson

Chris Barnes wrote:
This is really the key - from a SA standpoint, the best you can do is 
mark the message as spam and let the MUA (Outlook) deal with putting 
things into the proper folders on the user's machine (in the .pst file).


I don't know OL well enough, but I suspect that there is likely a 
registry hack you can do or a rule you can create that the users can 
import that will look at the headers and put the message into the proper 
folders.


No need for registry hacks, depending on how you flag the message. both 
Outlook and Outlook Express will filter on words in the subject, so a 
subject tag will work easily (Tools-Message Rules).  I'm not familiar 
enough to know whether you can filter on an arbitrary header, though.


--
Kelson Vibber
SpeedGate Communications www.speed.net


Verify using SARE updates

2005-07-22 Thread Curtis Vaughan
I used an update script from SARE (www.rulesemporium.com) and  
subscribed to the following lists:


TRIPWIRE, EVILNUMBERS, SARE_RANDOM, SARE_FRAUD, SARE_ADULT, ANTIDRUG

Now the update seems to work just fine. And I find the rules in /etc/ 
spamassassin. But how can be sure they are working? I mean I haven't  
seen a real increase in positive identifications. And some things  
that I think would be caught by some of these rules (although I  
haven't looked through them) get through undetected.


Or perhaps there are some other rules I should consider implementing  
and some of these that I should consider removing?


Thanks for any input.

Curtis


www.rulesemporium.com unreachable

2005-07-22 Thread wolfgang
Hi,
http connections to www.rulesemporium.com are timing out here. Maybe someone 
in charge is reading this and can fix it ...

regards,

wolfgang



Re: www.rulesemporium.com unreachable

2005-07-22 Thread Ed Kasky

At 12:58 PM Friday, 7/22/2005, wolfgang wrote -=

Hi,
http connections to www.rulesemporium.com are timing out here. Maybe someone
in charge is reading this and can fix it ...


This seems to happen once in a a while...

I wonder if it's ISP related?

Ed Kasky
~
Randomly Generated Quote (454 of 480):
Warranty and guarantee clauses are voided by payment of the invoice.



Re: www.rulesemporium.com unreachable

2005-07-22 Thread Chris Thielen

wolfgang wrote:


Hi,
http connections to www.rulesemporium.com are timing out here. Maybe someone 
in charge is reading this and can fix it ...


regards,

wolfgang


 

There seems to be an issue with the filesystem on that box.   Email sent 
to the appropriate people, but it looks like a manual power cycle will 
be necessary.


signature.asc
Description: OpenPGP digital signature