on
header checks and network tests.
SpamAssassin can ignore most HTML tricks like this, although I'm a bit
surprised this message didn't match one of the 'obscuring comment' rules.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Early mo
- presumably when the 'nospam' is stripped it
creates an invalid address, but real users can still reply. I get a few spam
to that address, but not nearly as many as other addresses that have been
posted to newsgroups.
--
Michael Moncur mgm at starlingtech.com http://www.starlingte
something wrong in someone's
corpus? It would seem like "ADV:" would only match spam.
Looking at the rules, neither requires the colon. Maybe a rule that checked
for ADV: or [ADV] would get a higher score?
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
&q
Ed Weinberg wrote:
> I am surprised that SA does not recognize spam sent by FormMail.pl.
The BUGGY_CGI rule catches that, it's been in SA for a long time. It must
work (in conjunction with other rules) because I haven't had a formmail spam
slip through in at least 6 months.
--
M
s are trying to work best for all
users.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Every crowd has a silver lining." --Phineas Taylor Barnum
---
This SF.NET email is sponsored by: FREE SSL Guide from Th
Razor function but Pyzor breaks.
I'm using -H /home/spam, and since /home/spam is the home for both Razor2
and Pyzor, that may be why mine works and yours doesn't.
I just double-checked, and checking and reporting work fine with Razor2,
Pyzor, and DCC.
--
Michael Mo
gered.
> And, yes, I was also confused by the wording in that man page.
Me too. I could still be wrong.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"The future is much like the present, only longer." --Dan Quisenberry
--
add points for anything non-English
ok_languages en fr -- add points for anything but English or French
ok_languages all-- turn off language checking
The rule UNDESIRED_LANGUAGE_BODY is used to add points to messages for this
one. Its score is 3.97 by default in 2.43.
--
Michael Mo
r-Version =~ /v /
describe MGM_DAILY_Mmgm: frequent spam header
score MGM_DAILY_M 3.0
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"2 is not equal to 3, not even for large values of 2." --Grabel's Law
-
sin test, but I think SA with Bayes and Razor would be far more
useful.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"All that is human must retrograde if it does not advance." --Edward Gibbon
---
This sf.
that matches five or more rules is marked spam, and it's really far more
sophisticated than that.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"I'm going to memorize your name and throw my head away." --Oscar Levant
e improved HTML percentage tests) catch them.
BTW, I think a threshold of 3.0 is a bad idea. But that's just me. I keep
mine at 7.0 and I've still had a false positive or two. (caused by RBLs)
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Time sneaks up
y message on the list! Trust me, list admins love this.
After using a Web spamtrap for a couple of months I found all sorts of
innocent mailing lists piling up in the 'spam' corpus. So I dropped that one
and now only use old addresses and never-used addresses (spammers love to
guess) as
ad-Topic - otherwise this is a
hole big enough for a spammer to drive a truck through...
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Everything you can imagine is real." --Pablo Picasso
---
This sf.net ema
> FWIW, there are no addresses at SpamAssassin.org worth mailing.
> Everyone lives elsewhere ;)
The "FAQ Owner" link on the FAQ page goes to [EMAIL PROTECTED] -
perhaps it shouldn't?
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"If the P
SPAM: * 0.0 -- BODY: T_HTML_P1_80_100
SPAM: * 0.0 -- BODY: T_HTML_TAG_EXISTS_CENTER
SPAM: * 0.0 -- BODY: T_HTML_SHOUTING1
SPAM: * 0.0 -- BODY: T_HTML_WIN_FOCUS
SPAM: * 0.0 -- BODY: T_HTML_CONSEC_IMGS04
SPAM: * 0.0 -- BODY: T_HTML_NUM_IMGS07
SPAM: * 2.7 -- Listed in DCC, see http:/
-Mailer-Version =~ /v /
describe MGM_DAILY_Mmgm: frequent spam header
score MGM_DAILY_M 3.0
uri MGM_HSM /hsmedia/i
describe MGM_HSMmgm: Hi-Speed Media: frequent spammer
score MGM_HSM 1.5
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com
SpamAssassin results
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Women who seek to be equal with men lack ambition." --Timothy Leary
---
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heave
;
Subject: E-mail to Groups? Download ZIM FastFeedback Now to Coordinate
E-mail Responses
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Fashion is something that goes in one year and out the other." --Unknown
core MGM_DAILY_M 3.0
body MGM_DISC /To discontinue the receipt of emails/i
describe MGM_DISC mgm: Discontinue receipt of emails
score MGM_DISC 2.5
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"It is impossible to defeat an ignoran
cc-tree/dcc.html DCC_CHECK 2.660
Listed in Pyzor, see http://pyzor.sf.net/ PYZOR_CHECK 2.500
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"All science is either physics or stamp collecting." --E. Rutherford
--
ed to remove it
from my test corpora? I know there were some recent changes in this regard
but don't know where things ended up.
(Hoping I don't need to run spamassassin -d over 5000 messages, it takes
forever.)
Thanks!
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.c
a major annoyance.
I'm aware that whitelist systems are the only way to truly dump 100% of
spam, but SA+Razor2 comes so close to eliminating my spam problem entirely
that it's not worth making everyone who wants to email me jump through
hoops.
--
Michael Moncur mgm at starlingtech.com ht
mail-spidering software used
by spammers won't see them. The method I use is to have email forms on my
Web sites, these send mail directly to me and are invulnerable to spammers.
Passwords might be a good compromise if you do lots of posting to USENET,
though.
--
Michael Moncur mgm at starlingt
hing scoring 5 to 10). The trouble with that is that SpamAssassin works
so well, I've only had about 5 false positives in the last two months, and
all of them were automated emails that wouldn't have made it through TMDA
anyway...
--
Michael Moncur mgm at starlingtech.com http://www.star
Kall us now to find out the sekrit to mass emial
marketting techniques that will maek you hundredds of dolars in the first
two moths!
If they do defeat the filter that way, though, we've achieved our goal of
making the spammers look (even more) like complete idiots.
--
Michael Moncur mgm at starling
I tried to get this to happen and it didn't set off that rule. Try posting a
bug to bugzilla.spamassassin.org and attaching the entire message (as an
attachment, not cut and paste)
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Confusion is always the m
am is tagged by one of them.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"I do not fear computers. I fear the lack of them." --Isaac Asimov
---
This sf.net email is sponsored by: See the NEW Palm
Tungsten T hand
an address and send mail to that forever.
(Just finished auditing my email logs for new spamtrap addresses. Addresses
that have never existed for legitimate mail but still receive regular spam
include webmaster@domain, sysop@domain, e@domain, jobs@domain,
hostmaster@domain, sales@domain, [EMAIL PROTEC
hether
someone actually entered SAtalk as "someone they admire." (I tried to log in
as SAtalk to find out who admires us, but someone has already registered
that address and I couldn't guess the password.)
--
Michael Moncur mgm at
g their spam and reporting
false positives.
Since users are hopefully going to be reporting messages as spam or nonspam
for Bayesian purposes anyway, it would be pretty easy to hook this up to the
AWL.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
&
at score to set for a new phrase?
The scores were calculated by a program that scans through a spam corpus
looking for phrases. I don't think there's a non-spam component, which is
why "from our" has a high score, among others.
I don't think it's meant for adding ph
eep them because SA takes
care of 99% of the problem.
I cringe every time someone says "what's the big deal, just press Delete."
Try that when you have 15X as much spam as real email...
--
Michael Moncur mgm at starlingtech.com http://www.starl
hould add
.78 points to every single nonspam message, I have set the score for this
rule to zero on my system.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Every crowd has a silver lining." --Phineas Taylor Barnum
-report feature for quite some time. (was it
2.30 that changed this?) The developers agree with you.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Pity the meek, for they shall inherit the earth." --Don Marquis
--
ng the SA lists for some reason)
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Spare no expense to save money on this one." --Samuel Goldwyn
---
This sf.net emial is sponsored by: Influence the futu
d
20_head_tests.cf. There's an online copy of this if you haven't downloaded
SpamAssassin:
http://www.spamassassin.org/dist/rules/
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"The covers of this book are too far apart." --Ambrose Bierce
-
positive. I then move
the valid spam to my spam corpus.
20+: "definitely spam" - dumps into a mailbox file on the server where I
scan through it every couple of weeks just in case, then add it to my spam
corpus.
(In about 4 months using this system, I haven't yet had a 20+ false
po
e a probably-spam threshold of 7 and a definitely-spam threshold of 20)
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"It is easier to forgive an enemy than to forgive a friend." --William Blake
---
This sf.net em
10 had a negative AWL
adjustment.
> I'll check in the reversion now... I guess this means we'll need to
> do a 2.43 soonish.
This does seem to be causing some confusion.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
&qu
e who sends more than x
messages must not be a spammer? That's far from true in my experience. I
often get 5-10 copies of the same spam from the same sender address. Not to
mention the fact that this is a very easy way for spammers to whitelist
themselves, just by doing what they do bes
I can submit a patch for the
docs if nobody else does.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Famous remarks are very seldom quoted correctly." --Simeon Strunsky
---
This sf.net email is sponsored b
uot;all")
> I think the following is correct:
>
> ok_locales en # Disable all of ja, ko, ru, th, zh
> ok_locales all # Allow all of ja, ko, ru, th, zh
> ok_locales ja ko # Allow standard english locales, disable ru, th, zh
That makes sense.
--
Michael Moncur mgm at s
> ok_languages en
>
> It must be listed, that's where I got it ;-)
That's nice, but I was talking about ok_locales. Completely different thing.
Anyone know whether "en" is valid for ok_locales? I thought it was at one
time.
--
Michael Moncur
marked as
possibly
> being spam in a foreign language.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"Students achieving Oneness will move on to Twoness." --Woody Allen
---
This sf.net email is s
looks for an IMG tag and not much else might
be good too.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"No human thing is of serious importance." --Plato
---
This sf.net email is spo
; The point is, these are the ones that were negative in the stock
> scores that really seem like they should be positive.
These are virtually all fixed in the new scores, and using those will get
you much better results than making up your own scores. Wait for the 2.42
release in a few da
om zone that only lists SPEWS,
we could use that in a meta rule.
--
Michael Moncur mgm at starlingtech.com http://www.starlingtech.com/
"It is better to be quotable than to be honest." --Tom Stoppard
---
This sf.net email is sponso
X-Spam-Flag
header. I had the best results detecting "X-Spam-Level: *" (or your
preferred number of stars).
I finally just used procmail to send the spam to a different POP mailbox and
let Outlook filter based on the account it came in on. Probably not an
option for everyone. :)
--
M
now would be to make it
a rawbody test and strip the tags myself.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Nobody can be exactly like me. Even I have trouble doing it."
-- Tallulah Bankhead
mazon and registered for lots of things
at Yahoo. I get 2-300 spams a week and not a single one has EVER been from
any of those companies.
Nonetheless, I'm currently unwhitelisting Amazon and eBay because spammers
frequently fake their 'from' addresses using these two.
--
michael mon
am.
Yeah, right.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Everywhere is walking distance if you have the time."-- Stephen Wright
---
This sf.net email is sponsored by:ThinkGeek
Wel
;t be
a problem from a SpamAssassin perspective - the score would be under 5 - but
I'm worried that if you do set up an RBL for things like this, some
misguided ISPs will start using it to dump email wholesale... As happens
with every other RBL.
Maybe this would be better as a DCC-type system
> It's a few bits of perl I threw together. I didn't make it public mostly
> because I just didn't get around to it. ;)
Thanks! I'll have to play around with that. Now I just wish there was some
way to report messages from Outlook without losing 90% of the headers.
e easily using procmail or your local equivalent.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"It was a book to kill time for those who like it better dead."
-- Dame Rose Macaulay
---
This sf.
> mysql> select COUNT(*) from reported where date > '2002-06-26
> 00:00:00' and rules like
> '%RAZOR_CHECK%';
Hey, cool. How did your reported spam end up in MySQL? Any chance you can
share the script that's doing that?
--
michael moncur mgm at starli
L: Received via a relay in orbs.dorkslayers.com
[RBL check: found 157.90.38.194.orbs.dorkslayers.com.]
* 2.0 -- RBL: Received via a relay in relays.osirusoft.com
[RBL check: found 1.191.99.210.relays.osirusoft.com.]
* -2.0 -- RBL: Do not double penalize for being an open relay on O
all_spam_to, in fact, because whitelist_to
isn't enough for some of the messages that discuss spam.
If spam to this list actually became a problem I suppose I'd have to do
something else.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Nothing is said
the mailing lists I subscribe to,
but we shouldn't expect most users to bother with things like that.
Of course, having negative rules for common mailing list software / systems
would accomplish the same thing. (Not hugely negative, just enough to claw
back a few points.)
--
michael moncur
's email.
The only challenge is finding a way to say things like "you're receiving
this message because you subscribed to our mailing list" and "follow the
link below to unsubscribe" without triggering SA. Just another way spammers
make it difficult for those of us s
mail and
participate in the mass-check process next time. If the GA had a few of your
non-spam mails that mention "viagra" in mind, it wouldn't give the rule such
a high score.
As for "incest", it's a word that appears in many discussions of a serious
topic, a few of
to spamtraps. All of
them had been bouncing for the last year or so. A few days later, I have 164
spam messages collected from those addresses. Bouncing addresses don't seem
to be of any concern to the spammers. Considering that most of them don't
even include a valid return address, thi
ported to DCC.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"When ideas fail, words come in very handy."-- Goethe
---
This sf.net email is sponsored by:ThinkGeek
Caffeinated soap.
bly don't have, but that doesn't make it spam. I get a
couple of those every month but I wouldn't call them spam.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Women who seek to be equal with men lack ambition." -- Timothy Leary
--
score X_AUTH_WARNING -0.703
score X_ESMTP-1.662
score X_MSMAIL_PRIORITY_HIGH -0.886
score X_NOT_PRESENT -1.920
score MAILTO_TO_REMOVE -1.669
All in all, I believe the GA is really smarter than I am this time. :)
--
mic
F_8BITS rule for 8-bit
subjects. People can score it however they want. The unexpected thing is
that every 8-bit subject also matches the SUBJ_ALL_CAPS rule, which it
shouldn't.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"My one regret in life is that I am not
a larger score than it deserves. The SUBJ_ALL_CAPS test should
probably ignore 8-bit characters.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don't change direction soon, we'll end up where we're going."
he last line looks for = followed by two digits and
replaces it with just the two digits.
This is the only thing I see after the tag that has any remote link with the
bug I was seeing...
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don't change directio
nd let it update) may have a problem, but 2.30 looks OK.
Weird.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don't change direction soon, we'll end up where we're going."
-- Professor Irwin Corey
d new one. Can someone with the latest 2.21CVS try it?
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don't change direction soon, we'll end up where we're going."
> On Wed, Jun 12, 2002 at 03:45:48PM -0600, Michael Moncur wrote:
> > > http://www.hughes-family.org/bugzilla/show_bug.cgi?id436
> >
> > What happened to the = sign in id436? This has been happening
> for several
> > messages.
>
> Looks fine to me! I thin
zor and/or DCC. Just not as many as the English-language spams...
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don't change direction soon, we'll end up where we're going."
e it a hard link and it
> works now.
>
> Hmmm... It looks like other people are going for it, too :-) Have fun,
> all!
You know us. We can't resist the opportunity to get more Spam. :)
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"If we don'
hitelist_to worked fine too,
but I stopped using it since 6 points isn't enough to guarantee that
something gets through.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"You can pretend to be serious; you can't pret
nd SAdev message, doesn't
require a new rule, and works in individual user preferences.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"You can pretend to be serious; you can't pretend to be witty."
-- Sacha Guitry
_
s? Doesn't matter much, the character-set rules catch these
anyway.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"You can pretend to be serious; you can't pretend to be witty."
-- Sacha Guitry
_
}->{languages_filename}) || die "cannot open
languages: $!\n";
>>>>>>> 1.2
Those >>> and <<< symbols aren't mine, they're in the file. There's also a
spurious = on line 57. Removing them made it run and the languages stuff
seems
> > To answer my own question after some testing, "en" is the correct value.
>
> Yep, it's in the documentation too, BTW.
Ah, found it in man Mail::SpamAssassin::Conf. I was looking in the (older)
manual on the Web site before.
--
michael moncur mgm
ules_filename}
And it's coming out empty, so the "languages" file isn't found. I tried
hardcoding the path to the rules directory in TextCat.pm and now it works.
Not sure what the real solution is.
Oh, I also had to manually copy the "languages" file from CVS to the r
least that's what one of your
original messages about the patch said. This isn't in the documentation yet
so I wasn't sure of the syntax.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"A good listener is
dy?
I'm not sure whether this is a good idea or not. On the one hand, it seems
like if any five low-scoring rules are enough to guarantee that something is
spam, the GA would have given them higher scores. On the other hand, maybe
it wouldn't hurt to let the GA figure this out.
e this (in Outlook) as my only spam protection and it caught 90% of spam,
but then the spammers got smart and it stopped working very well at all.
That's when I switched to SpamAssassin...
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"I have often depend
ore or less random score because
there weren't many non-spam ebay questions in the corpus. There certainly
isn't a single spam message with "Question for seller" in the subject in my
archive of the last 5,000 spams I've received.
--
michael moncur mgm at starlingtech.co
O_BE_REMOVED_REPLY 0 # -2.150
score TO_UNSUB_REPLY0 # -1.996
score WEB_BUGS 2.0 # -0.823
score X_MSMAIL_PRIORITY_HIGH0 # -1.356
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"
EAVE_SUBJECT Spam unsubscribe (LEAVE)
score LEAVE_SUBJECT 2.5
I don't recommend them for SA in general since they're only matched by one
spammer who are bound to change their name (again) before too long.
--
michael moncur mgm at starlingtech.com
DVERT_CODE Subject =~ /(?:^\s*|\s+|\[)(?:ADV|cc)[:\]]/i
describe ADVERT_CODESubject: contains advertising tag
This would also catch the "cc:" that many spam subjects start with. I see
more of these than ADV: these days.
--
michael moncur mgm at sta
me as whitelist_from but
can have its own score, and use default_whitelist_from in 60_whitelist.cf.
That way (a) anyone can turn off the default whitelist with a single score
entry in a preference file, and (b) spam reports will refer to the "default
whitelist" so it's easy to diag
too, but I doubt spammers bother.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"My sources are unreliable, but their information is fascinating."
-- Ashleigh Brilliant
___
Have bi
m,
@paypal.com, and @amazon.com addresses are becoming all too common in
spam...
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Research is the process of going up alleys to see if they are blind."
If they're all coming from a similar source (i.e. a mailing list) you might
be better off whitelisting rather than changing scores.
If you're getting false positives with really high scores, you might let us
know which rules they're hitting, as the rules might need work.
-
Same here. Appears to be a typo - here's the fix:
Line 81:
if ($self->{conf}->{check_mx_attempts) < 1) {
Should be:
if ($self->{conf}->{check_mx_attempts} < 1) {
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"Research is the process
;t use whitelist_to because a score of 6.0 isn't enough to
completely "opt out".
Have you considered that SA in its default configuration is an "opt in"
system anyway? Since it marks messages and doesn't filter them, you could
run it sitewide and let users o
t looks like it. I
don't know why this is missing from the right-click menu.
I personally use a special IMAP mailbox to report spam. I move the spam to
that folder, and a cron job grabs it every few minutes, runs spamassassin -r
on it, and returns it to my spam archive folder.
--
m
occasions where the AWL causes false negatives.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"The ships hung in the sky in much the same way that bricks don't."
-- Douglas Adams
th a lot of email clients, the second you make *any* font change --
> italics, color, bolding -- it becomes an HTML message.
I believe the "HTML check" in question is a check for Content-type:
text/html ONLY. Outlook, in particular, always includes a text version in a
multiple-part messa
ad
an optional eval rule that calls FPROT or sophos or another easily
available, regularly updated virus scanner. Heck, there's even an
open-source one in development:
http://www.openantivirus.org/
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"The ships hun
e.
This is not necessarily a bad thing but the threshold should definitely be
something larger than 3.
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"We forfeit three-fourths of ourselves to be like other people."
-- Arthur Schopenhauer
gs over the threshold.
Perhaps after testing it might be good to have a separate LOTS_OF_PORN_3
rule that checks for a higher number...
--
michael moncur mgm at starlingtech.com http://www.starlingtech.com/
"We forfeit three-fourths
Perhaps a count
higher than 3 would be better?
Second, the @porn_words Daniel posted doesn't include all of the words that
PORN_3 does. It's missing everything from, er, "whore" to "titties" in the
current CVS.
I'm sure it's better than the current PORN_3 regar
I ran some quick macros on the current RATWARE test and came up with the
attached "20_ratware.cf" file with each mailer in its own rule. I haven't
included scores - perhaps this could be thrown into the next GA run if it's
worthwhile.
--
michael moncur mgm at sta
1 - 100 of 181 matches
Mail list logo
