Angel Gabriel wrote:
I want to setup a machine, with a sole purpose, of scanning email. All
email to be delivered to my domain, will be accepted by this machine,
and then, after scanning delivered to my mail server.
How can I achieve such a setup?
There are more ways to do this than I can count
Robert Leonard III wrote:
I am wondering, what, in your opinions, is the best choice for building SA..
I am using it for a Site-Wide proxy, to pass on to my MS Exchange server.
No local mail accounts, simply pass-through.
Qmail?
Postfix?
Sendmail?
I'm running all three in various places, and
In most cases I can think of, spamc doesn't get the message until after
the connection is already finished. A milter is the only way I can
think of you can get a result from SA before Sendmail's finished
receiving the mail, and all the milters I'm aware of are capable of just
the sort of behav
Steven Manross wrote:
I'm seeing a few/lot of spam that has a reverse lookup name that is like the
originating IP.
i.e. If it were 192.168.52.45 that was the originating IP, the reverse
lookup might be 192-168-52-45-clients.domain.com
This is a *very* common naming convention for ISPs that ass
There's a thread currently going on about this exact same subject. Check
out the "Open Source is Naughty" thread.
Andrea Riela wrote:
Hi folks,
How could I filter this type of spam:
Penis
P.enis
Pe.nis
Pen.is
Peni.s
In the html body?
Thanks for your advices
Andrea
--
Gerhard Hofmann wrote:
I thought I could furtherly improve configuration by adding something
like "whitelist_subject" or "blacklist_subject".
It seems to me that'd be pretty redundant... There's a mechanism in
place already to score e-mails that have specific tags in the subject.
For examp
Chris Santerre wrote:
I'm not sure if this is a Spam or legit! Anyone ever got one of these from
the list?
A bounce message from someone whose susbcribed e-mail address is no
longer valid? Sure, I see them all the time.
This message is intended only for the use of the person(s) listed above
Fred Marton wrote:
Great! Thanks. Only now, when I try to run spamd, I get:
Could not create INET socket: Cannot assign requested address
IO::Socket::INET: Cannot assign requested address
Either it's already running, or something else is running on the port
spamd wants.
This message is i
Tobin wrote:
Hello,
I was wondering if anyone could help me fix a broken rule. Im getting a
error
"Failed to compile body spamassassin tests, skipping:
(syntax error at /ect/mail/spamassassin/local.cf, rule Porn, line 1,
near "/)
Debugging a custom rule would be a lot easier if you showed us
Darryl Snover wrote:
I've done the patch, and also re-installed everything, checking my
steps carefully, watching for errors. I can now run the 'spamassassin
--lint' without errors, and things like 'spamassassin -V' return as
expected. However, still no success on actually running an email or
Darryl Snover wrote:
I've just re-made spamassassin, and now, running the spamassassin
--lint results in the error:
razor2 check skipped: No such file or directory Inescure dependency in
connect while running with -T switch at
/System/Library/Perl/5.8.1/darwin-thread-multi-2level/IO/Socket.pm
Then it's probably functioning correctly. "autolearn=no" means *that*
message wasn't autolearned.
Bill Polhemus wrote:
How can I know for sure if Auto-Learn is functioning correctly?
It does seem to be working for SOME messages but the headers in many
of them say:
X-Spam-Status: … autolearn=n
Patrick Beard wrote:
After a few more tests it seems that it is nothing to do with pdf
attachments.
Emails with a total attachment of less than 200k come through fine. Emails
greater than 200k are corrupted.
The fact that spamassassin is not logging anything in debug would suggest
its having real
Johann Spies wrote:
+*1ADBMt-00024Y-00*obt1AWUAynY*
X-Spam-Checker-Version: SpamAssassin 2.60 (1.212-2003-09-23-exp) on
maties1.sun.ac.za
X-Spam-DCC: :
X-Spam-Status: No, hits=6.8 required=8.0 tests=BAYES_90,NO_REAL_NAME,
RCVD_IN_BL_SPAMCOP_NET,RCVD_IN_SORBS autolearn=no version=2.60
Tony White wrote:
The subject says it all. I know the defaults on this, and have not
changed them. I'm fairly certain that SA is autolearning spam --
because of the growth of the bayes_seen and bayes_toks databases. But
I'm not sure about ham at all. Currently, I'm not sending outgoing
mai
AltGrendel wrote:
I agree with that. I see addresses at my client that haven't existed in
5+ years. No one should kid themselves that spam lists are cleaned or
that they learn from being blocked at the firewall.
According to my most recent spam statistics, two out of the top three
spam recipie
Peter Rosa wrote:
Hello list's friends.
I'm sorry for writting same message third time, but please be patient with
me - I have about 200 spam messages a day so I REALLY need to block them.
You'd probably have more luck un a procmail list, since that's where
your problem really lies. That sai
You're going to get a *lot* of meaningless hits on some of these -- any
mail from level3.net, for example. They'd score a hit on your rules, and
they're a pretty decent-sized ISP with a *lot* of legit servers
(including a few of mine). You'd also get automatic hits for sending
mails from serve
Unless I'm reading this regex incorrectly:
/^https?\:\/\/rd\.yahoo\.com\/(?:[0-9]{4,}|partner\b|dir\b)/i
it's pretty specific about looking for "http[s]://rd.yahoo.com".
Colin A. Bartlett wrote:
Can anyone hazard a guess as to why a message with an image and several
links in this fashion did no
Hugh Caley wrote:
I'm trying to build spamassassin 2.60 as a non-root user on my ISP's
mail server. RedHat 7.3, perl version 5.004_04. I'm not sure where
to go from here. I'm getting the following errors:
That's a really ancient version of Perl. I'm not sure you'll be able to
make it work.
Guessing from the incomplete stuff you posted, it appears you're running
SA through another package -- amavisd, maybe. Check there -- whatever
package you're using probably has a lower spam threshold set than the
one you've set in SpamAssassin.
---
Kris Deugau wrote:
I found that even this wasn't lightweight enough on the filter server
I'm adminning, so I wrote a microprogram to use the load average
instead. Ideally, this should be hacked into sendmail itself in some
way; I'm not going there myself. I *might* poke it into procmail if I
ge
Mike Schrauder wrote:
I get:
debug: tokenize: header tokens for X-MimeOLE = "Produced By Microsoft Exchange
V6.0.5762.3"
Can anyone tell me if this is screwing up the reclassification
of these msgs as spam? Do I have to strip out all headers that were
added after the spam got past the postfix bo
Thomas Kinghorn wrote:
Hi List.
Just a quick question.
Do all custom rules have to be added to the /etc/mail/spamassassin/local.cf
file OR
can a seperate .cf file be placed in the spamassassin directory?
Either will probably work, but it depends a bit on how you call
SpamAssassin. Generally
[EMAIL PROTECTED] wrote:
Is it possible to send the mail elaborated from spamd to another
external server.
I use smtproutes in order to select the relay with
AntiVirus and/or AntiSpamming.
domain1.com:localhost:2525#antivirus
domain2.com:localhost:783 #antispamming
===
Vittorio Muth wrote:
I've read in the FAQ, that spamassassin does not drop mails. OK.
Is there anyone who got it, to configure spamassassin like that:?
As you read in the FAQ, SpamAssassin doesn't drop e-mails. Warnings
that deleting mail SpamAssassin marks as spam aside, it's impossible to
sa
The permissions most likely aren't OK. Are you running the
"check_whitelist" command as root? If not, you probably have no access
to the /root directory.
Jim Knuth wrote:
Hallo SA-List,
I wont use the tool "check_whitelist".
When I use the command: ./check_whitelist | grep [EMAIL PROTECTED
Bill Polhemus wrote:
I realize this is a Perl problem, but can anyone help me diagnose
this? I know next to NOTHING about Perl.
The following is an error message I get when attempting to run SA from
an email alias under Sendmail:
- Transcript of session follows -
Insecure
Bill Polhemus wrote:
This one got through. Can we figure out how and why?
With just the body of the message and no headers, probably not. That
really not enough to work with.
---
This SF.net email sponsored by: Enterprise Linux Forum Conf
Welcome to SourceForge... Happens all the time with their lists.
Colin A. Bartlett wrote:
Moderately Off Topic:
Does anyone else have trouble getting their messages on the list in a timely fashion? It seems when I reply to the list it takes 15 to 45 minutes for my message to appear. By the time
Yevgeniy Miretskiy wrote:
On [10/17/03 09:54], Colin A. Bartlett wrote:
Diego Puppin Sent: Friday, October 17, 2003 8:55 AM
is it possible to add the spamassassin "hits" score to the spam mails,
so that I can sort my spam box and find the "spammest" emails?
Yep, it's in the Docs at
Everything in /etc/mail/spamassassin, and all the files from wherever
your Bayes stuff is stored. I'll admit I'm a bit fuzzy on how to pull
off the AWL transfer -- maybe someone else could pipe in on that.
This is, of course, assuming you've installed SA on the new box, and
weren't planning on
Assuming you've got all the Perl modules installed on the new box, and
the same version of the Berkeley DB libs, a direct copy should work fine.
Mike Carlson wrote:
After a hugely successful test of spamassassin, I now need to move it to a bigger badder production box. Can I just copy file from
How are you calling SpamAssassin? If you're running it via some sort of
helper app like Amavis or MIMEdefang, it may be stripping the headers
normally added by SpamAssassin.
Michael Balamuth wrote:
Further info. Upon reviewing the Mail::SpamAssassin::Conf documentation, I
noticed that one hea
Arlo Gilbert wrote:
Spamassassin 2.6, perl 5.8 on a redhat 9 machine
called from perl i get the following debug messages even though it
appears to read those two settings as sa does abide by them.
debug: Failed to parse line in SpamAssassin configuration, skipping:
report_header 1
debug: Faile
I got one of these, too, last night, from a machine at a Korean ISP, and
I've been trying to figure what I did.
Somehow I feel less special now. :)
Chris Santerre wrote:
Did anyone else get a nasty email this morning? I did! This weekend ROCKED for my SA config. Jennifer, if you were here I'd k
You probably have two versions installed on that system. Remove the old
one.
Petre Bandac wrote:
[EMAIL PROTECTED]:~/kits/Mail-SpamAssassin-2.60# spamassassin -V
SpamAssassin version 2.43
if I use spamd/spamc, I get in the headers sa-version=2.60, but if I user
spamassassin I get version=2.43
Jeff Lasman wrote:
If I should, then I'll probably uses messages sent to this list, as I
know there's no spam hidden in them; I can't guarantee that about any
of my other folders. Is that a reasonable assumption?
Should I try to put through the same number of messages? Or doesn't it
matter?
Brad wrote:
body MOON /moon/i
score MOON5.0
I'm not positive, but I think the "describe" line is also necessary.
This message is intended only for the use of the person(s) listed above as the
intended recipient(s), and may contain information that is PRIVILEGED and
C
Mike Van Pelt wrote:
It's still having spells of rejecting connections as
"it's busy" when CPU is less than 50%.
(This is on a Solaris box.)
Are you running MIMEDefang with the multpilexor? If so, have you tried
raising the maximum number of allowed connections?
---
[EMAIL PROTECTED] wrote:
Under Solaris 8, applying the Razor2.patch to fix the taint problem Dies
horribly:
Are you using Solaris's "patch," or the GNU one? If you're not using
GNU's, you might give it a shot.
smime.p7s
Description: S/MIME Cryptographic Signature
This message is intended
[EMAIL PROTECTED] wrote:
Is it possible to write in log the IP of the source server and not
127.0.0.1?
I use qmail->qmail-scanner ->spamd (spamassassin)
Thanks
Sep 8 07:03:47 email spamd[7832]: connection from localhost
[127.0.0.1] at port 40938
Sep 8 07:03:47 email spamd[18364]: chec
Daniel Quinlan wrote:
In theory, it's safe. In practice, it's slow and ... ?
In practice, in my case, it's not slow at all -- But NetApps seem to
have a pretty decent NFS implementation. Access to a lot of stuff
(though I haven't clocked Bayes specifically) is often faster than local
disk
Dan Wilder wrote:
On Tue, Oct 07, 2003 at 11:54:07AM -0700, Patrick Morris wrote:
I thought i read somewhere that somebody was sharing bayes over NFS
with a netapp filer with no > probs.
I'm doing it -- no problems to report.
With things writing to the same db from d
> I thought i read somewhere that somebody was sharing bayes over NFS
with a netapp filer with no > probs.
I'm doing it -- no problems to report.
---
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
Gwen Morse wrote:
How do you run SpamAssassin. I'll bet you a dollar
SA's not what's saying it's possible spam. You're probably running
it through a helper app that's interpreting SA's output incorrectly.
I'm running it through spamc, as my initial message
indicated.
Please Paypal that do
Yusuf Goolamabbas wrote:
Hi, I want to upgrade from SA 2.53 to 2.60. It's very likely that some
rules may have been added/deleted. I have modified some scores in my
local.cf
What happens if rules corresponding to the modified scores no longer
exist in SA 2.60 ?
Nothing will happen... You'll ju
Gwen Morse wrote:
2) below is the result of my test of
sample-nonspam.txt. I find it confusing because
although it scores 0.0/5.0 (Which I assume reads a 0
total score when 5 is the default for it to be spam),
the message *says* it's been identified as possible
incoming spam. Is that normal for _a
Scott Rothgaber wrote:
debug: executable for pyzor was found at /usr/local/bin/pyzor
debug: Pyzor is available: /usr/local/bin/pyzor
debug: entering helper-app run mode
debug: Pyzor: got response: /usr/local/bin/pyzor: not found
s2:[/etc/mail/spamassassin]
> ls -l /usr/local/bin/pyzor
-rwx--x--x
Looks like you're probably scanning the message twice.
Jay Pfaffman wrote:
X-Spam-Checker-Version: SpamAssassin 2.60 (1.212-2003-09-23-exp) on
learn.coe.utk.edu
X-Spam-Level:
X-Spam-Status: No, hits=4.0 required=5.0 tests=CLICK_BELOW,FRONTPAGE,
HTML_MESSAGE,SAVE_ON_INSURANCE autolearn=no
Robert Nicholson wrote:
when running with qmail on a plex box with virtual users where will
spamassassin look for the "users" spamassassin configuration? ie. what
normally is ~/.spamassassin
It won't. If the user has no home directory, it won't use user prefs
(since there's no place to put the
Sure:
sa-learn --spam --dir /var/spool/mailbox/224/sa_spam/in/*orourke.ca
Dan Tappin wrote:
I have created spam and ham accounts to utilize sa-learn. I am using
Tenon's Post.Office on OS X as a mail server.
I was running sa-learn like this:
sa-learn --spam --dir /var/spool/mailbox/224/sa_spam/
Oops... drop the --dir parameter (assuming your copy of SA is updated
-- the --dir and --file parameters are deprecated).
Patrick Morris wrote:
Sure:
sa-learn --spam --dir /var/spool/mailbox/224/sa_spam/in/*orourke.ca
Dan Tappin wrote:
I have created spam and ham accounts to utilize sa
The 8.0 value you're seeing for the score is a rounded value, and is
probably slightly below 8.
John Schneider wrote:
I have been working too much lately, so my eyes might be a little
crossed But, the following message seems to have met the required value for spam, but was not flagged:
[s
Check out MIMEDefang, amavisd, spamass-milter, or one of the other
packages that can tie SpamAssassin to Sendmail.
Vivek Kumar wrote:
Hi there,
I am using sendmail and wat ot install spamassassin. The way sendmail is
configured is all the incoming mails comes to sendmail(on Linux box) and
then
Check out the README and install files, or look at the list archives for
messages about Razor. you'll need to patch it.
Scott Rothgaber wrote:
[2.60 on FreeBSD 5.0]
Watching the mail log go by and I noticed this...
razor2 check skipped: Bad file descriptor Insecure dependency in open
while r
Justin Mason wrote:
Rather disappointed in how many things are broken in 2.6 , like
installation, working with mimedefang, bayes, and who knows what
next.
What's wrong with MIMEDefang?
I'll admit I'm a bit curious about this myself. We run SA on four
servers with MIMEDefang here and didn
That's why reading the README and INSTALL files is always a good idea --
you'll probably run into other problems (use Razor?) if you don't.
Danny Aldham wrote:
I have recently upgraded spamassassin from 2.55 to 2.6 . It looks
like a new method of creating the BAYES database is used, that
require
You might try "force install Digest::SHA1." It looks like *parts* of it
are up to date, but not all of it.
John Schneider wrote:
Can I do that in CPAN or do I have to do it outside to correct the problem?
When I do a "install Digest::MD5" and "install Digest::SHA1" in CPAN
currently it tells
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 an e-mail from "out there",
Usually, this error means something connected to port 25 of your server,
then disconnected before issuing any commands.
Do you have a system monitoring the SMTP port on that server? I get
these once a minute from my monitoring system, that's just checking to
make sure the daemon's alive.
Matt
Since you didn't post the e-mail, what rules got hit (either time), or
how you've got SA set up (network tests enabled? DCC? Razor?), it's
impossible to say why you got two different scores -- but it's not too
uncommon for the same message tested at different times to get different
scores, fo
Robert Nicholson wrote:
> RE:
>
> X-Spam-Status:
> Yes, hits=10.7 required=0.6 tests=DATE_IN_PAST_03_06,
> DATE_SPAMWARE_Y2K,FORGED_MUA_OUTLOOK,FORGED_OUTLOOK_HTML,HTML_60_70,
> HTML_FONTCOLOR_BLUE,HTML_FONTCOLOR_GREEN,HTML_FONTCOLOR_RED,
> HTML_MESSAGE,MIME_HTML_ONLY,MIME_HTML_ONLY_MULTI,MISSING_
A "milter" is a filter program that works exclusively with Sendmail. A
quick Google search found it here:
http://savannah.nongnu.org/projects/spamass-milt
Milters don't work with Qmail.
Tomáš Macek wrote:
Can you tell me, where can I get the "spamass-milter" program?? Is it a script? Does it w
Mark London wrote:
Does spamassassin simply total up the scores of all the tests which
are found to be valid for a message, or does it increase the score
based on whether a combination of tests exist?
Sometimes. SpamAssassin has both additive rules (each hit is worth a
set number of points)
1. Did you enable rules in user prefs in the system sa config? The
default settings don't allow rules to be defined in user prefs (See man
Perl::SpamAssassin::Conf).
2. The Bayes rules that score 0 (most of the mid-range of Bayes checks)
don't show up in the report, but that doesn't mean Bayes ch
Adam Lasnik wrote:
1) I have mails over a threshold of 5 thrown into a "Spam" IMAP folder
by SA automatically. That works fine. But I wanted to use Outlook's
"in header" rule to move items scoring 2 or 3 or 4 stars into a
"Possibly spam" folder.
Good luck using Outlook's filters on an IMAP fol
Nope -- it's got nothing to do with SpamAssassin at all, so no amount of
whitelisting it in SA will help you. You need to whitelist it in the
software that's actually marking it as spam.
Jennifer Wheeler wrote:
what am i doing wrong here? I am trying to unblacklist an address
getting tagged b
The site.config.m4 file looks OK -- are you sure you're using a copy of
Sendmail built with those options? Maybe you've got more than one copy
floating around on that system.
McWhirter,Julia wrote:
Having problems integrating spamassassin/mimedefang and sendmail on our
external SMTP server.
Us
SA reads any config files in /etc/mail/spamassassin automatically.
Ralf G. R. Bergs wrote:
Martin Radford schrieb:
[...]
4) I am new to spamassassin. What parameter should I use in my
configuration to cause spamassassin to stop using osirusoft open relay?
See the news article at http://news.s
That has got to be the single funniest piece of spam I've seen in a long
time. :)
jherschel wrote:
-Original Message-
*From:* Tyler Madison [mailto:[EMAIL PROTECTED]
*Sent:* Wednesday, August 27, 2003 2:05 PM
*To:* [EMAIL PROTECTED]
*Subject:* Best Child Pornography Site
Welcome to th
It's no longer an option in recent versions. It used to be the score at
which a message would automatically be reported as spam to DCC, Razor,
et al. For various reasons (all of them good ones, IMHO), it was removed.
Louis LeBlanc wrote
Most of this I understand, and I know some of this has c
It depends how you have SA set up, and what calls it. SpamAssassin will
not ever delete mails -- it just tags them.
Daniel Kaliel wrote:
I have been trying to find documentation on how to do this, but have
not found any. I spoke with the partnership of my company on this and
how it is safer
Maybe the big image with a Viagra ad in the middle of it has something
to do with it. :)
Jim wrote:
I've been getting these once in a while, they just seem to have random
words in them.. no sales pitch that I can see. And they are always
different sets of words.
I made a few tiny edits to obfus
What are the permissions on the file (and the directories it's in)?
Maybe SA can't read it.
Luis Muñiz wrote:
Hi there!
I've just setup (not system-wide[1]) Mail-SpamAssassin-2.55. My
problem is that every time a mail is received, procmail' log shows
the following:
Can't locate HTML
You can use the whitelist_to, more_spam_to, and all_spam_to
configuration options, but it won't cause SpamAssassin not to process
the mail, just not to tag it as spam (to varying degrees). Another
downside is that, if a mail comes in to a whitelisted user and a
non-whitelisted user, it'll most
This option no longer exists. I used to use it myself, but after it was
gone I realized it was a lot more efficient to stop them before they get
to SA (in my case, by having Sendmail reject mail to invalid accounts).
Walter S. Arnold wrote:
This is quite different from blacklist_from as I unde
Alex J. Avriette wrote:
In /etc/mail/spamassassin/local.cf I have the following rules:
[snip]
X-Spam-Status: No, hits=3.5 required=4.0
tests=AMIKA_SPAM,HTML_70_80,MIME_HTML_ONLY,RCVD_IN_SBL
version=2.55
Is there a problem with my rules, above? Or is it that sa isn't
reading rules from local.cf?
See "bayes_file_mode" in the Mail::SpamAssassin::Conf man page.
Andreas Vogt wrote:
Is it possible to set acces rights for global bayes files in local.cf
(e.g. 660 or 666)?
---
This SF.Net email sponsored by: Free pre-built ASP.NET sit
Scoring changes between versions because -- well, spam does. New rules
or added and old rules are retested before every release.
Stephen Boals wrote:
Just upgraded to 2.55, and in testing found that I am getting
different scores between versions. Has the scoring changed for
rules? Can I get
Unfortunately, as far as I know if you use SQL prefs, you're stuck with
site-wide Bayes. I don't think individual user Bayes databases are
possible when SA runs under a single account.
A single site-wide database works, though.
Ricardo Kleemann wrote:
I know SA can handle user prefs via SQL,
Trying to seed the Bayes databases with other people's spam probably
isn't a good idea -- the whole idea is that it learns what *your* ham
and spam look like. Filling it up with mail from other sources is
probably just going to lead to innacurate results.
Ricardo Kleemann wrote:
I've trained
I don't do postfix or procmail, so I can't really help there, but you
definitely should *not* use SpamAssassin 2.44. It's outdated, and
you're sure to be disappointed with the results.
SpamAssassin is like a virus scanner: it needs to be kept updated to be
effective.
Mark Brosius wrote:
I ha
I think the point is that your rule is probably doing exactly what you
told it to do: assigning any message that has the letters "sex" in it 50
points. I'll bet if you save the encoded messages out to text files and
do a "grep sex [filename]" on them, you'll find they do have those
letters in
How are you passing mail to SpamAssassin?
Greg Ennis wrote:
I have set up two Red Hat 8.0 systems using sendmail 8.12.9. I am using one system as a gateway for internet mail, and the other as a network hub that functions as a POP3 server for several PCs. The gateway mail server has local mail a
It depends how your reverse DNS is handled. Do you do it in-house? Does
your ISP do it for you?
We really aren't who you need to ask if you don't know. :)
Indii wrote:
My email server is a debian box while my dns server is an MS 2000 server.
Which would i need to setup the reverse dns on and h
I'd immediately disable a rule like this. Many sites either have no
website (such as the domain I use at home), or don't allow ICMP through
to their webservers (like the domain I use at work). You'd probably get
FPs on mail from me no matter where I sent it from.
Michael Clark wrote:
Would a
That's normal when you're using MIMEDefang... you won't see
SpamAssassin headers unless the message is deemed to be spam. Your
headers at least tell you MIMEDefang is running, so at this point it
looks like you've done everything right.
Guillermo Delmastro wrote:
Hi
It seems that spamassassi
MIMEDefang can also do this if the mimedfang-filter is incorrectly
configured... you might want to tell us how you call SpamAssassin.
Chris Trudeau wrote:
This is likely NOT SA.
It is likely either amavis or other external program that is calling SA.
There is a SA config parameter where it will
It means the "V" must be preceeded by a "word boundary" (such as a
space, line break, etc.).
Robin Witkop-Staub wrote:
What does /b in front of this filter mean?
body RAVEN_VIAGRA_OBSCURED /\bV.i.a.g.r.a/i
describe RAVEN_VIAGRA_OBSCURED contains advertising Viagra by any name
score RAVEN_VIAGRA
You can't define rules in user prefs unless you also define the
allow_user_rules variable. See the Mail::SpamAssassin::Conf man page
for reasons why this my not be a good idea.
Daniel Carrera wrote:
Hello,
Below I include a spam email I got. I have a series of rules that should
have been ac
Yes -- forwarding alters the mail headers, which will cause problems
with learning.
Russell Premont wrote:
Our mail server controls mail for multiple domains. What I wanted to do is
setup an address that users could forward spam to, so I could then run
sa-learn on the mail folder. My question is
That's probably your problem right there... Procmail usually only kicks
(Bin as a message is being delivered to its mailbox. There are ways around
(Bthat, but chances are that's why you're getting what you're getting.
(B
(BYou've got a few options, but the easiest would probably be to use a
(B
You're misinterpreting the man page:
"Another thing to be aware of, is that typically you should aim to train
with at least 1000 messages of spam, and 1000 ham messages, if
possible. More is better, but anything over about 5000 messages does
not improve accuracy significantly in our tests."
T
Out-of-the-box, no. In any case, you probably wouldn't want to try
running milters on Sendmail 8.11.x, as they tend to be somewhat less
than stable.
Patrick Murphy wrote:
I am running Solaris 8 with Sun's sendmail 8.11.6 and am in the process
of choosing a milter to use. Does anyone know if Su
It's unneccesary. MIMEDefang doesn't call spamassassin -- it uses SA's
Perl modules internally.
Mark Emerle wrote:
Hi-
I have a related question. Is there a way to have mimedefang to call
spamc instead of spamassassin? I have yet to find documentation stating
that.
smime.p7s
Description
You can't call SA directly from Sendmail. You'll need to either use a
milter such as MIMEDefang (which I use here, there are others), or
something like procmail.
Leo Huang wrote:
Hello,
I want to call spamc in the Sendmail, so the messages won't be re-scanned if
it is delivered to multiple us
The header will only appear if the message is near the low or high end
of Bayes scoring. Your average message, which will have no Bayes score,
also won't have a Bayes header.
Chris Halverson wrote:
I've run into this issue too. I have a user who fed (via sa-learn) a
few thousand messages to SA
I think custom rules may be a problem. I don't believe (though I could
be wrong) SA allows rules to be defined in user prefs.
Kai Schaetzl wrote:
Matt Thoene wrote on Wed, 25 Jun 2003 12:56:01 -0700:
I have been running SA for a few local users for quite a while. It's
worked so well that I
If there are no weird characters in that line that didn't translate into
the e-mail, you're not running the version you think you are. You may
want to check and make sure you don't have an older copy of SpamAssassin
installed in a different place from 2.55.
Jim Ford wrote:
On Fri, Jun 20, 200
1 - 100 of 120 matches
Mail list logo
