>Aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it
>deosn't mttaer in waht oredr the ltteers in a wrod
>are, the olny iprmoetnt tihng is taht the frist and
>lsat ltteer be at the rghit pclae. The rset can be a
>total mses and you can sitll raed it wouthit porbelm.
>Tihs is bcuseae the huam
Is anyone else seeing this?
Since I installed SA 2.55 recently and subscribed to this list, I'm
getting the digest indexes delivered properly, but the emails are not
including any of the actual messages sent to the list.
I use PINE for all my email, in case that helps.
Regards,
CPAN into Perl
5.8.0. Note that I couldn't use Sun's default Perl 5.6.1 installation
since at least one of the modules listed in the INSTALL file insisted that
it would not install with anything less than Perl 5.8.0.
I know there's an issue using Perl 5.8.0 which is mentioned in the doc
s though, the headers appear as they were received.
Nope:
[craig@balam craig]$ dig example.com mx
; <<>> DiG 8.3 <<>> example.com mx
;; res options: init recurs defnam dnsrch
;; got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4
;; flags: qr rd ra
Putting on my Commercial Software hat, Deersoft is entering final beta
on our SpamAssassin for Exchange product. We're looking for beta
testers. If anyone's interested, please contact me quickly to get in.
Current version of the product is targeted at Exchange 2000; 5.5 is in
the pipeline b
TOTALSPAMNONSPAM
32142280 934 SMTPD_IN_RCVD
So there's about twice as much spam as nonspam sent through that
mail server. The way scores are set is related to the relative
frequency of spam vs nonspam triggering a particular rule, but
the optimizat
y, July 18, 2002, at 01:10 PM, Hess, Mtodd, /mth wrote:
> I was all exited about this until I noticed it was just for
> Outlook. I was
> hoping for SA that would run on a Windows SMTP server...
>
>
> Todd
> [EMAIL PROTECTED]
>
>
>
> -Original Message-
>
Hmm, does this count as spam?
Deersoft announced today the release of their first
SpamAssassin-derived product for windows:
http://biz.yahoo.com/bw/020718/180175_1.html
Basically, an outlook plugin with an SA backend. Now Windows
users can share the benefits of SA.
For those of you who are
Smarter spammers who are trying to avoid Postini etc. will also
be smarter at avoiding our filters.
C
On Thursday, July 18, 2002, at 09:56 AM, Martin Nile wrote:
> However it seems that a higher percentage of SPAM that slips in beneath
> spamassassin came in via the secondary MX.
-
I think this is a side effect of the way Postini (and some
others) do mail filtering for enterprises -- basically they sell
a mail forwarding service, which you create as your low-score MX
record, they scrub mail, then forward to your "real" mail
server. The sales pitch says "If our service g
l/bin/perl
use Mail::SpamAssassin
my $test = Mail::SpamAssassin->new();
my $username = 'test';
$score = $test->load_scoreonly_sql ($username);
print "Score: ". $score;
-------
Thanks,
Regards,
Crai
I vote for "liability for personal injury or death" -- how
likely is a spammer to stick that in their messages?
C
On Sunday, July 14, 2002, at 11:47 AM, Suzanne Britton wrote:
> In the many months I've been using SpamAssassin, I've only seen
> one false
> positive. I just checked it against
Whitelist entries can either be on separate lines, as Mike
describes, or you can put multiple address patterns on one line,
separated by whitespace. No quotes nor commas.
C
On Sunday, July 14, 2002, at 01:30 AM, Mike Burger wrote:
> On Sat, 13 Jul 2002, Chris Ogles wrote:
>
>> Can someone h
Jim,
please forward the spam to [EMAIL PROTECTED] -- they'll hopefully
be able to deal with it.
Thanks,
C
On Saturday, July 13, 2002, at 06:46 AM, Jim Meyering wrote:
> FYI,
>
> I was dismayed to find obvious spam in my inbox today.
> This came through with such a high score because amazon.
On Friday, July 12, 2002, at 02:05 AM, Matt Sergeant wrote:
> Andre Bonhote wrote:
>> Hi SA-folks!
>> This just dropped in this morning and got -0.3 hits. There might be
>> something wrong somewhere.
>
>
>> X-Spam-Status: No, hits=-0.3 required=4.0
>> tests=DEAR_SOMEBODY,US_DOLLARS_2 version=2
I've long been of the philosophy that spamassassin should focus
on exactly one thing: identification and tagging of spam. Leave
the delivery to something which does delivery well. Leave the
SMTP to something which does SMTP well. etc, etc, etc. I
haven't chopped out the delivery stuff alto
That's probably illegal in a lot of jurisdictions.
C
On Thursday, July 11, 2002, at 04:13 PM, Jeremy Zawodny wrote:
> On Thu, Jul 11, 2002 at 01:34:19PM -0600, Mike Diehl wrote:
>>
>> The other day, I got this little jewel in my e-mail. So I
>> decided to have a
>> little fun. I sent a repl
Is the concensus that Razor2 is stable enough for us to release
it? I'm somewhat wary of releasing a SA version with Razor2
support and having to answer daily emails where the answer is
"upgrade to version 2.x -- prior versions are buggy". In the 2
weeks I was gone, have things settled down
On Monday, July 8, 2002, at 08:03 AM, Justin Mason wrote:
>
> Patrice Fournier said:
>
>> I was wondering if there was someplace where we could have a new rule
>> tested automatically against both spam and non-spam corpus to see how
>> effective it is. Is this something everyone must make on th
Hello folks,
I'm back and well rested after 2 weeks on tropical islands.
Highly recommend both Petit St Vincent
(http://www.psvresort.com) and Martinique to anyone wanting to
get away for a while.
Turns out that about a day after I left, one of my colleagues
broke my server, with the result
So the origin of the problem here is that Justin is Irish, and I
grew up in the UK so my s/z o/ou etc. spelling plus my grammar
is some weird half-assed mishmash of brit and american.
C
On Saturday, June 22, 2002, at 08:57 PM, Danita Zanre wrote:
> Only if you're an American - and by the sp
I noticed this problem at the time, but ignored it, cos I
figured anyone who actually bothered reading the changelog would
likely know what the deal was. But that was probably a bad
assumption since our userbase has started growing so huge now.
We should probably file a bugzilla ticket again
Ok folks,
This is probably the last time I'll check list mail before I head out tonight.
Have fun while I'm gone, keep up the good work, don't let the bedbugs bite, etc.
See you all in 2 weeks.
C
---
Sponsored by:
ThinkGeek at http://www.Th
It also depends on the people. Some people definitely prefer
easy-to-intall-and-configure worriless time-limited shareware to
have-to-be-a-unix-sysadmin free (as in beer) software. Which is why we think
DeerSoft can thrive by putting a lot of work into the free (beer and speech)
SpamAssassin pro
Matthew Cline wrote:
MC> http://slashdot.org/article.pl?sid=02/06/19/1335210&mode=thread&tid=111
MC> http://story.news.yahoo.com/news?tmpl=story&u=/cn/20020619/tc_cn/937300&printer=1
MC>
MC> CmdrTaco says that SA catches over a hundred spam a day for him; pretty good
MC> publicity. The Razor for
Ok, enough bugfixes and scorefixes to 2.30 are in, and it looks to be in good
enough shape that I've packaged it up and released it. This is really a service
release to get something a little better than 2.30 out, since I'm heading off
tonight for my 2 week carribean honeymoon. First week, no te
Derrick 'dman' Hudson wrote:
DdH> On Tue, Jun 18, 2002 at 05:11:54PM +0200, Tobias von Koch wrote:
DdH> | On Tue, 18 Jun 2002 09:30:06 -0500, Derrick 'dman' Hudson wrote:
DdH> |
DdH> | D> | Hmmm...would it be safe to assume that any attachment type of
DdH> | D> text/| should be scanned while anyt
I'm not so sure outlook won't display all kinds of weird stuff if it says it has
a particular MIME type (and is maybe even encoded), if:
a) there are no other parts
b) the thing actually decodes to text, even if the MIME type is wrong
c) the thing actually decodes to X (where X is MSWord, RTF, HT
The original link is toast, but it looks to be archived at
http://www.eatmycomix.com/02.02.13.html
C
Bob Proulx wrote:
BP> > I started using SA, and now I worry why inbox is so empty. I guess I really
BP> > only get spam. :)
BP>
BP> A few months ago a list reader posted a pointer to a cartoo
[EMAIL PROTECTED] wrote:
> Does anyone know a way of whitelisting mail from certain email servers? We
> are trying to do this because originally we whitelisted our domain name, now
> we are getting spams through with a forged source address from our domain
> name. So we were thinking we could whi
I've checked in to the HEAD cvs branch a manual application of Theo's patch, so
if you get the latest CVS along with the new Razor2, try it out and let me know
if it's not working.
C
Theo Van Dinter wrote:
TVD> On Mon, Jun 17, 2002 at 11:39:07AM -0700, Craig R Hughes w
Actually, you would gain a tiny bit -- if a spamc call happens at the moment
when spamd is dead and not yet re-listening on the port, then you'll have a
message or two not filtered (depending on load and timing). If spamd re-reads
its config file without stopping listening on the port, then thoug
Plus Vipul seems to be pretty smart, so I'm betting that Razor will only get
better and better with time.
C
Theo Van Dinter wrote:
TVD> On Mon, Jun 17, 2002 at 09:32:12AM -0700, Bart Schaefer wrote:
TVD> > > Not sure why so much effort is given to Razor.
TVD> >
TVD> > I'm using neither Razor no
Did you restart spamd?
C
Christophe Zwecker wrote:
CZ> Hi,
CZ>
CZ> I upgraded from 2.20 to 2.30, but below header shows version 2.20, is
CZ> that normal or did something go wrong in my upgrade ?
CZ>
CZ> X-Spam-Status: No, hits=-0.7 required=8.0 tests=FROM_AND_TO_SAME,AWL
CZ> version=2.20
CZ>
C
Miles Fidelman wrote:
MF> On Sat, 15 Jun 2002, Simon Lyall wrote:
MF>
MF> Does the latest release of spamassassin support Razor v2 now?
Not yet -- there are some errors which crop up when invoking Razor from spamd --
we're working with the Razor guys to resolve the problem. We have a patch read
Great work Theo. Thanks! We'll probably want to make a 2.31 release with some
updated scores and Razor 2 support asap.
C
Theo Van Dinter wrote:
TVD> On Sat, Jun 15, 2002 at 04:39:25PM -0400, Theo Van Dinter wrote:
TVD> > I've sent in a patch to add support, but there were some bugs in the
TVD
It was probably me when I broght all the 2_3_0 line changes forward onto the
trunk.
C
Daniel Quinlan wrote:
DQ> Bart Schaefer <[EMAIL PROTECTED]> writes:
DQ>
DQ> > Hmm, I just did a "cvs up" (on the head, not the branch) and:
DQ> >
DQ> > score: FORGED_RCVD_TRAIL 1.000 -> absent
DQ> > score: MSG
spammy indicators though. I would suggest we manually reset this to ~1.5
MM> score INTL_EXEC_GUILD-0.039
Ditto. Only 16 instances in the corpus though here.
MM> score LINES_OF_YELLING -0.036
Hmmm, rule need fixin?
MM> score MONEY_BACK
Daniel Quinlan wrote:
DQ> >>> score X_NOT_PRESENT -1.920
DQ> >>
DQ> >> This one is on my hitlist as well. Didn't work out very well.
DQ>
DQ> Craig R Hughes <[EMAIL PROTECTED]> writes:
DQ>
DQ> > But it actually turns o
Daniel Quinlan wrote:
DQ> Craig R Hughes <[EMAIL PROTECTED]> writes:
DQ>
DQ> >> score: BUGZILLA_BUG -2.000 -> 0.921
DQ>
DQ> > Moved to the right section of the scores file, and score reverted to -2.0
DQ>
DQ> But why is it positive? Doesn't it mean th
Daniel Quinlan wrote:
DQ> Michael Moncur <[EMAIL PROTECTED]> writes:
DQ>
DQ> > And a few slightly questionable scores:
DQ> >
DQ> > - This was 0.87 before. Less and less useful?
DQ> > score FROM_AND_TO_SAME -2.071
DQ>
DQ> I think this one should go. It's a common way to send email t
Bart Schaefer wrote:
BS> These look suspicious:
BS>
BS> score: ASCII_FORM_ENTRY 0.036 -> -1.660
Changed back to 0.5 -- as mentioned in previous message, this is triggering on
the sourceforge-appended footers on mailing list mails.
BS> score: BUGZILLA_BUG -2.000 -> 0.921
Moved to the right sect
It's a missing use line, not a missing file. Add the following line to
spamassassin.raw, among the other use lines at the top:
use Pod::Usage;
I've fixed this on the 2_3_0 CVS branch, so it'll be in 2.31 which might issue
sometime next week.
C
Skip Montanaro wrote:
SM
Rule added.
C
Skip Montanaro wrote:
SM> rawbody OBFUSCATING_JAVASCRIPT /charCodeAt|fromCharCode/
SM> describe OBFUSCATING_JAVASCRIPT JavaScript which tries to hide the message
___
Don't miss the 2002 Sprint PCS Appl
Razor2 support will be added to the CVS branch very soon, but will probably not
be stable for a few days.
C
Matthew Cline wrote:
MC> On Saturday 15 June 2002 12:47 am, Jim Scott wrote:
MC> > Ok I have just installed on a new machine SpamAssassin 2.30 and Razor 2.x
MC> > and SpamAssassin is repo
it's cool that I'm listed on the "Credits" page;
MC> thanks! :-) You should probably take down the "FAQs" link until you actually
MC> do have a FAQ up. And I'm curious if all of the DeerSoft support is going to
MC> be handled just by Craig and Gordon, or i
Ok, I think I've fixed the MANIFEST problems now, so if you already grabbed 2.30
and it failed for you, try again now and you should be OK.
C
___
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las V
.sh
KO> debian-rc-spamd.conf README.spamd-vpopmail solaris-rc-script.sh
KO>
KO> I'm trying to build this on Solaris 7 box running perl 5.6.1-multi-thread.
KO>
KO> I can see spamc code in 2.20, but not in this one. I've tried the tgz and
KO> the zip files with no s
Matthew Cline wrote:
MC> On Friday 14 June 2002 05:49 pm, Craig R Hughes wrote:
MC>
MC> > Viking: spam song
MC> > Dentist: Tooth hurty
MC>
MC> Spam hurts your teeth?
Think about it.
C
___
Don&#x
Matthew Cline wrote:
MC> On Friday 14 June 2002 04:46 pm, Craig R Hughes wrote:
MC>
MC> > Ok, 2.30 is now official.
MC>
MC> > SpamAssassin 2.30 (Viking dentist l'orange) Released!
MC> > ---
Viking:
> Was there supposed to be a Makefile.PL is the tar.gz file?
BM>
BM> Brian
BM>
BM> - Original Message -----
BM> From: "Craig R Hughes" <[EMAIL PROTECTED]>
BM> To: <[EMAIL PROTECTED]>
BM> Sent: Friday, June 14, 2002 4:48 PM
BM> Subject: [SAtalk] [S
time their email
program opens, it is of grave concern to parents protecting their children's
email. Craig Hughes and Gordon Kruberg today announced the formation of
Deersoft, Inc. and the company's plans to develop and release anti-spam software
for consumers and the businesses
Ok, 2.30 is now official.
SpamAssassin 2.30 (Viking dentist l'orange) Released!
---
Here's the super-brief summary of the major changes from 2.20:
* Rule Fixes/Updates/Additions
* Bugfixes/Typofixes
* Speed improvements
* New options
Not fully judged that way yet -- still about 5 minutes until 2.30 is released!
C
Duncan Findlay wrote:
DF> On Fri, Jun 14, 2002 at 06:11:41PM -0400, Miles Fidelman wrote:
DF> > On Sat, 15 Jun 2002, Simon Lyall wrote:
DF> >
DF> > Does the latest release of spamassassin support Razor v2 now?
DF>
2.30 coming Real Real Real Soon now! I'm just working through some last second
packaging issues.
C
Matt Sergeant wrote:
MS> Daniel Quinlan wrote:
MS> > Michael Moncur <[EMAIL PROTECTED]> writes:
MS> >
MS> >
MS> >>I think the problem is simple: We have a SUBJ_FULL_OF_8BITS rule for 8-bit
MS> >>
Sean Rima wrote:
SR> Actually RAM has always been a problem for me because it is not that
SR> cheapo here, it pees me off when people say that they can get a strip
SR> for $20 when I would be looking at 150+ :(
Plus 17.5% VAT no doubt ;)
C
_
Derrick 'dman' Hudson wrote:
DdH> OTOH, saying that DCC is I/O bound means that the limiting factor is
DdH> bandwidth. In this case it is network bandwidth. You can put a
DdH> super-powerful Athlon in that box, but DCC won't run any faster
DdH> because it mostly sits idle waiting for data to co
Actually, using -a with -S should increase the false _negative_ rate, not the
false positive rate, since it will lower the AWL score for a spammy sender to
around 5.
C
Duncan Findlay wrote:
DF> On Mon, Jun 10, 2002 at 02:34:54PM +0200, Michael Stauber wrote:
DF> > Hi Craig,
DF> &
Bart Schaefer wrote:
BS> Aside from the image attachments, the only identifying marks are extremely
BS> poor written English. If they weren't so freaking huge, I'd read them for
BS> the amusement value alone. Last time one slipped through, it included an
BS> invitation to visit Mr. River's fact
Mail::Audit doesn't let you change the contents of a mail message once it's
created, which SA needs to be able to do easily. So we implemented our own
more-or-less compatible Mail::Audit replacement, called
Mail::SpamAssassin::NoMailAudit -- try using that class instead.
C
Jake Edge wrote:
JE>
OSX startup stuff happens by running scripts called
/Library/StartupItems/ProductName/ProductName -- sniff around in
/Library/StartupItems/ and you'll see what I mean.
If I get bored I might try making an OSX SpamAssassin package.
C
Ian Vännman wrote:
IV> Date: Tue, 11 Jun 2002 16:29:22 +020
Freeze for 2.30, call for nonspam.log submissions
ON>
ON> Craig,
ON>
ON> I rsync a mass-check and posted bug #421 that has the rule description
ON> in French corresponding to the 2.30 coming tomorrow.
ON>
ON> Olivier
ON>
ON>
ON>
_
Matthew Cline wrote:
MC> On Sunday 09 June 2002 11:09 pm, Craig R Hughes wrote:
MC>
MC> > you are a nonspam.log submitter, please check out that tag from CVS using:
MC>
MC> Why just the nonspam.log?
Cos I have plenty of spam to generate a very good spam.log -- all I need is
Ok, if anyone is about to submit a patch, it's time to do it. I'd like to get
2.30 out by wednesday, which is pretty aggressive, but things are pretty stable
right now. Please don't anyone check anything in (those of you with commit
privs). Instead, please attach any changes to a bugzilla ticke
Sean Rima wrote:
SR> But if I disable spamc in Exim (I used your example config) and leave
SR> dcc only I never get above a load average of 2.3 and dccd is flooding
SR> with remote servers as well, pointing out that I use spamd with -L.
Actually, *not* using -L should *decrease* your load -- in
Derrick 'dman' Hudson wrote:
DdH> On Sun, Jun 09, 2002 at 11:06:32PM +0100, Sean Rima wrote:
DdH>
DdH> | Yeah the spam[c/d] setup. My average is around 15 seconds, well it is an
DdH> | old p133 the slowest appears to be 93 seconds. I am a dialup user and
DdH> | when I go online off peak for the f
Sean Rima wrote:
SR> I thought it would not affect me too much but even with a -m 2 and a -s
SR> 61440 I reach a load average of over 15 which cripples my poor old mail
SR> box :)
Hmm, my understanding of "load" is "number of processes in the WAIT queue",
which with -m 2 can only be 3 for SA --
The -S flag to spamd should also help greatly in constrained hardware
situations.
C
Michael Stauber wrote:
MS> Hi Sean,
MS>
MS> > My ISP is also looking at SA but performance maybe a problem.
MS>
MS> Yes, it sure can be. I've set up SA for a couple of ISPs on their Cobalt RaQ
MS> servers. Those
I'm picturing some developer with a line like:
print "Approved for $5000...For $user only...";
sending out a test message sees that it reads:
Approved for ...For fred only...
And saying to himself "Doh! $5000 is a bad variable -- I should use '' there!"
-->
print 'Approved for $5000...For $
Olivier Nicole wrote:
ON> > The tricky part is maintaining a list of which rules will work only in CVS vs
ON> > which rules are OK for the nightly distro.
ON>
ON> I beleive regexpr are stand alone.
Not always -- for example when "rarebody" gets added as a regexp type.
ON> I beleive most eval a
Jeremy Zawodny wrote:
JZ> Does anyone with a large spam collection have good stats on the
JZ> average message size?
SizeCount
-
0-1k89
1k-10k 57170
10k-50k 17062
50k-100k 530
100k-200k 282
200k-500k 128
500k-10M47
That's a
I answered this question first time you asked it. use -s mail not mail.info
You set the facility, not the level of logging.
C
Steven Núñez wrote:
SN> To be exactly, under OpenBSD 3.1, I am starting spamd with the following
SN> command line:
SN>
SN> /usr/bin/spamd -d -a -c -s mail.info
rODbegbie wrote:
r> Craig R Hughes wrote:
r> > It's now
r> > been quite a while since our last release, and so I'm going to do the
r> > 2.30 release imminently. I will post a separate thread message with
r> > the usual freeze request plus requests for n
Olivier Nicole wrote:
ON> >The tricky part is that frequently between releases new rules get built which
ON> >depend on new library features -- so CVS rules won't necessarily work on
ON> >anything but CVS libs. Certainly at least EvalTests.pm would need to be in the
ON> >"rules update" distro
ON
Duncan Findlay wrote:
DF> Would it not be easier to simply release the non-eval rules. Sure,
DF> scores wouldn't be ideal. But hey, that's the price you pay :-)
Yes, more or less. Except also excluding other rules like, say, uri rules
before the 2.20 release, or "ok_languages", say.
C
__
Hanser, Kevin wrote:
HK> Hello,
HK> A message that was sent from on of our mailservers was caught in a
HK> spamassassin filter. (No, we don't send spam...) One of the items that was
HK> caugh was "Message-Id is not valid, according to RFC-2822". I need some
HK> help in deciphering the regex th
Jesse Sheidlower wrote:
JS> What do I do to inform Mail::SpamAssassin where my rules directory
JS> is for the local install?
new Mail::SpamAssassin({ rules_filename => '/path/to/rules' });
It's called "filename" but it wants just the path, not the actual filename.
C
_
The tricky part is that frequently between releases new rules get built which
depend on new library features -- so CVS rules won't necessarily work on
anything but CVS libs. Certainly at least EvalTests.pm would need to be in the
"rules update" distro, but probably also other stuff from time to t
Just a final summary of what got checked in after all this back and forth:
spamd -D [but not -d] will print SA debug messages to stderr
and log its own activities to syslog (as before)
spamd -D -d will log SA debug messages to syslog
[EMAIL PROTECTED] wrote:
> Craig R Hughes <[EMAIL PROTECTED]> writes:
>
> > BH> Just seems to me that if a program's logging, e.g., SA, is set to go to
> > BH> the screen, then, unless it offers some sort of config. or switch,
> > BH> there's
[EMAIL PROTECTED] wrote:
>
> [Slaughtered Subject line ALERT: Not sure what this belongs under,
> it showed up on my system sort of buggered up]
>
> Bryan Hoover <[EMAIL PROTECTED]> writes:
>
> [...]
>
> >> STDOUT. If you specify -D and -d together, then spamd should
> >> definitely install
> >
Daniel,
for sanity in installation for end users, can we do one of the following:
a) keep the single-unified language file for the distribution, and just stick
the component model files in a subdirectory under contrib, with the merge
program under tools, but where those files are in MANIFEST.SKI
perldoc IPC::Open2
C
Steve Thomas wrote:
ST> Anyone know how to open a file/command (open FOO, "|foo"), write to it and
ST> then read back from it? It seems like something that would be needed fairly
ST> often, although I can't say that I've run across the need for it in the
ST> couple of years
Bryan Hoover wrote:
BH> Ryan Hayle wrote:
BH>
BH> > * Progress of Senate anti-spam bill
BH> >
BH> > Over a year ago, Sen. Conrad Burns (R-MT) introduced S.630, the CAN
BH> > SPAM Act of 2001. This bill would require UCE to have a valid return
BH> > address to facilitate consumers' removal from sp
Bryan Hoover wrote:
BH> Craig R Hughes wrote:
BH>
BH> > SA when it was
BH> > first built was not expecting to be used in a detached-daemon
BH> > environment where
BH> > its stderr was disconnected -- so when we disconnect it in spamd, it's
BH> >
Frickin' manifest. Ok, I updated it so tonight's build should work.
C
Bart Schaefer wrote:
BS> On Wed, 5 Jun 2002, Craig R Hughes wrote:
BS>
BS> > Don't know what to tell you guys -- TextCat.pm is definitely in CVS.
BS> > Maybe there's some weird permi
Yes, I just noticed this myself. I think it's a bug in the patch from Bugzilla
#343 -- looks like a copy/paste bug -- how can you expect to read from a
descriptor which perl just told you it failed to open...
I'll delete that line.
C
Bart Schaefer wrote:
BS> During "make test" (which BTW I t
Bryan Hoover wrote:
BH> Craig R Hughes wrote:
BH>
BH> > No, you can't just redirect STDERR, if you're using the -d flag, spamd
BH> > will
BH> > detach from the launching tty, and disconnect from that tty's STDERR,
BH> > STDIN and
BH> > STD
No, it's worked out now, and fully operational in CVS.
C
[EMAIL PROTECTED] wrote:
> Craig R Hughes <[EMAIL PROTECTED]> writes:
>
> > There's a rule for that now in CVS. Any repeated word in all caps which is 3 or
> > more letter long will trigger.
>
&
Jan Chrillesen wrote:
JC> On Tue, 04 Jun 2002, Duncan Findlay <[EMAIL PROTECTED]> wrote:
JC>
JC> > Try doing a cvs update. Or if you downloaded of the web site, try
JC> > downloading again. (I.e. today's build rather than yesterday's)
JC>
JC> Same problem here. Both in the cvs snapshot downloaded
Bryan Hoover wrote:
BH> Craig R Hughes wrote:
BH> > spamd is the thing which is detaching itself from the terminal and
BH> > therefore losing SA debug messages;
BH>
BH> But these messages are dropped when SA is run from .procmailrc aren't
BH> they? That is, aren&#
usefully go -- in the
case of spamd, that's syslog, through the logmsg() function defined in spamd
C
Bryan Hoover wrote:
BH> Craig R Hughes wrote:
BH>
BH> > $SIG{__WARN__} = sub { logmsg($_[0]); }
BH> >
BH>
BH> I assume you would add the above to spamd? Would it make
M. Brownsworth wrote:
MB> Ah, sorry. My procmail is doing system-wide filtering and invokes
MB> spamc. From /usr/local/etc/procmailrc:
MB>
MB> :0fw
MB> | spamc
MB>
MB> :0:
MB> * ^X-Spam-Status: Yes
MB> /var/log/spam
Ok, so why not just add a rule at the top of your procmail recipe which looks
Olivier Nicole wrote:
ON> > I don't see how the Subject contains anything remotely like a unique
ON> > ID code. Is this right?
ON>
ON> I did not check at the rule (too lazy.buzy) but maybe the unique ID is
ON> characterised by some string at the end of Subject, after a lot of
ON> spaces.
Yes, 7
Harry Putnam wrote:
HP> Probably should go back to default, and see what I observe. The
HP> first time around a message with FREE FREE in the subject and
HP> several times in the body came thru, it may have been a fluke.
There's a rule for that now in CVS. Any repeated word in all caps which
Bart Schaefer wrote:
BS> Hrm. This was reported once before, and I thought the description had
BS> been updated then. Sorry, I should have checked.
Ok, I just checked in a better description.
C
___
Don't miss the 2002 Sprint PCS A
Harry Putnam wrote:
HP> I've also had to edit a number of other rules score values so
HP> probably just got off on the wrong foot.
You're just not going to do a better job of optimizing scores than the GA will,
though you might be seeing some "unusual" email that's not paralleling the
corpus wel
Harry Putnam wrote:
HP> Duncan Findlay <[EMAIL PROTECTED]> writes:
HP>
HP> [...]
HP>
HP> >> Going to the debug output, how can I find the output that pertains to
HP> >> that message? No date no msgid.
HP> >> I don't see why its being overlooked either. Its not a restart
HP> >> problem though.
Lars Hansson wrote:
LH> On Wednesday 05 June 2002 00:20, Tony L. Svanstrom wrote:
LH> > Looks like IT-news in swedish qualifies as part porn whenever something in
LH> > the news ends (end in swedish == slut). =D
LH>
LH> Heh, I suspect the number 6 (sex in swedish) will have the same effect ;)
I
Bryan Hoover wrote:
BH> Harry Putnam wrote:
BH>
BH> > So I think Duncan has hit it, in that -d (daemonize) silences the
BH> > debug output.
BH>
BH> Indeed sir. I was woefully ignorant of the behavior.
BH>
BH> Does that mean too, that when spamd calls SA, SA does not log as usual?
BH> My thought
1 - 100 of 1019 matches
Mail list logo
