Hi Janeth,

You could have a rule to look for vowels a, e, i, o, u and numbers 1 thru 0
in the message, if none are found then it probably could be your spam
message.

Ken

Hi Janet
I did read your other post but you did not qualify the type/format of the
spam you were getting, otherwise I would have given you my input(for what it
is worth)
I now believe what you are getting is the .gif type spam i.e. it is to all
intents a picture and NOT text - hence message rules as commonly used will
have no effect.
Some months back when this type became more prevalent, I posted the
suggestion(not mine)that message rules could work, but it was a very
'long-winded' job.
In message rules, you stipulate blocking any message that does 'not contain'
a letter of the alphabet, since this type of email has no letters. You need
to do one rule for each letter.
You can add numbers one to ten if you wish.
This does work - I have not tried it though - according to feedback.
The other alternative is to contact your ISP - they may provide a spam block
service where all suspect spam is put into your Mail Box.
The odd one or two might get through, but if you get 20+ per day you will be
more than happy to only get one or two.
Not all ISPs can block .gif spam - yet.  Some still have no spam blocking
service at all.
Those ISPs who do have this spam blocking service, as the norm, expect you
to go to your mail box and check for any genuine emails that may have got
pulled in with the spam.
My ISP will delete all spam in my mail box every 7 days, by date age.
Rgds
Antioch

I see a number of varieties of spam. Much is "multipart", with an HTML
component that carries the payload, and a text component that carries a
"hashbuster"; random text designed to be changed with each message in a
spam run. The purpose is to break the "hash" created by anti spam
systems such as DCC and Razor.

Some carries images, in a base64 encoded block of text, and, again,
"hashbuster" text.

I am getting spam to a MyRealBox account which is so minimal in payload
content that, even with the capability of the Mercury/32 filters to
check against header lines which MSOE ignores, it is difficult to
filter. E.g.:

The rule I settled on for this one looks like this:

That is from the rule file in Mercury/32; this program uses a text file
for the rules.

First line says to skip the next line if my actual name appears in "To:"
or "Cc:".

The second line, which would be skipped on a match in the first line,
says to move email which has my MRB email address to a different account
("A/S" means "Anti Spam"; but the account really has a given Christian
name).

I use Pegasus Mail, in conjunction with the Mercury/32 MTA; and both
have a rich ruleset. I don't use MSOE, so I don't know how to accomplish
similar with MSOE.

OTOH, the rules I created will only work with this account because it is
so rarely used, and only by me for certain tests. I can use a technique
like this because I don't have to worry about correspondents sending me
email without the exact given name I have specified in my rules.
Otherwise, I would have to try and create a scoring rule in the Content
Checks, which applies points to words in the message body, and/or
headers, where the points would add up to enough to cross a "spammy"
threshold for this message, while not crossing the threshold for a
non-spam message.

MSOE is just not capable of anything like that, at all. In fact, the
primary reason that I moved to Pegasus Mail over MS Outlook Express is
the richness of the filters. And, with the latest release, Pegasus Mail
4.41, Pegasus Mail has included a Naive Bayesian spam filter.

I further switched from relying on third parties for email to running my
own server, for my own domain, because that gives me better control over
the SMTP transaction, which is where the spam is injected into the SMTP
system.

This is not for everybody, though; and, if you insist on staying with MS
Outlook Express, you will have to use the tricks commonly posted here;
"Filtering In" (whitelisting), and such.

Sometimes, MSOE can be finagled to work with an anti spam filter proxy,
which interposes between MSOE and the POP3 server. I have used POPFile
with succes, but K9 seems to be more popular. Both use Naive Bayesian
statistical analysis of the words appearing in email. They "learn" as
you click on the choice of "spam", or "ham"; and the corpus used for
analysis is unique to each user, thus foiling spammer efforts to
"poison" the filter with hashbusters.

K9:  http://keir.net/k9.html
POPFile:  http://popfile.sourceforge.net/

Other than that, getting effective spam rules in MS Outlook Express is
very difficult; it was never designed for filtering spam.