Default score for UPPERCASE_75_100

[Fran Fabrizio <fr...@cis.uab.edu>]

What's the default score for UPPERCASE_75_100?  This test does not 
appear to be documented at http://spamassassin.apache.org/tests_3_1_x.html.

I am examining a mail to figure out why it got a false positive.  User 
has no user_prefs or even .spamassassin directory.  If the default is 0, 
where else might this be set?  This is the last test I can't account 
for, and the score is still 1.2 or so higher than the sum of the rest of 
the tests that hit.  Setup is spamassassin 3.1.0 invoked through amavisd 
on a linux postfix server.

Thanks,
Fran

-- 
Fran Fabrizio
Senior Systems Analyst
Department of Computer and Information Sciences
University of Alabama at Birmingham
http://www.cis.uab.edu/
205.934.0653

Re: Default score for UPPERCASE_75_100

[Fran Fabrizio <fr...@cis.uab.edu>]

> Please don't use Reply if you're not replying. It makes a mess of the
> threading.

In Thunderbird I right-clicked on a prior mail to the list and chose 
Compose Mail To, which starts a new mail, not a reply.  So probably is 
something in your setup if it is showing otherwise.  That's odd, though.

> Have you looked in /usr/share/spamassassin/50_scores.cf or similar?

Thanks for this.  I have been using spamassassin for a while, but 
frankly, it's been working so well I've never needed to analyze a mail 
rule by rule till now.  I was looking everywhere but there. :-)

This particular message, although legit, was trying REALLY hard to look 
like spam.  Encoding the text message with base64, having a mime file 
attached that had no name, the subject and body both in all caps, etc... 
but it was helpful to tell the recipient exactly why, rule by rule, that 
message was caught by his spam rule.  It was from a travel agent.  What 
is it with the travel industry still using all capital letters 
everywhere? :-)

Thanks for the help!

-Fran

Re: [OT- threading] Default score for UPPERCASE_75_100

[Magnus Holmgren <ho...@lysator.liu.se>]

Matt Kettler skrev:
> Magnus Holmgren wrote:
> 
>>Please don't use Reply if you're not replying. It makes a mess of the
>>threading.
> 
> 
> Magnus.. While that is a good suggestion, I don't see how it applies here.
> 
> There's no References: or In-Reply-To: in Fran's original message, and my copy
> of Thunderbird shows it as a top-level parent, not a child of any thread..

It may be my Thunderbird that's screwed up. My apoligies.

(Chunks that should go in the index files (.msf) regularly end up in the
mailbox files and vice versa, it seems. My $HOME is NFS-mounted but I
don't know if that's part of the reason, or if it's just that the old
box I use as server produces stochastic errors (but only in that
particular way, as far as I've noticed), or if something entirely
different is to blame.)

-- 
Magnus Holmgren

Re:[OT- threading] Default score for UPPERCASE_75_100

[Matt Kettler <mk...@evi-inc.com>]

Magnus Holmgren wrote:
> Please don't use Reply if you're not replying. It makes a mess of the
> threading.

Magnus.. While that is a good suggestion, I don't see how it applies here.

There's no References: or In-Reply-To: in Fran's original message, and my copy
of Thunderbird shows it as a top-level parent, not a child of any thread..

I don't think this thread was started with a hacked-up reply as you allege it
was, unless my copy of Thunderbird 1.0.7 is wonky and stripped the headers on
me.. (highly unlikely).

Re: Default score for UPPERCASE_75_100

[Magnus Holmgren <ho...@lysator.liu.se>]

Please don't use Reply if you're not replying. It makes a mess of the
threading.

Fran Fabrizio wrote:
> What's the default score for UPPERCASE_75_100?  This test does not
> appear to be documented at http://spamassassin.apache.org/tests_3_1_x.html.

At least in my installation (Debian package) it's:

score UPPERCASE_75_100 1.394 1.040 0.809 1.371

> I am examining a mail to figure out why it got a false positive.  User
> has no user_prefs or even .spamassassin directory.  If the default is 0,
> where else might this be set?  This is the last test I can't account
> for, and the score is still 1.2 or so higher than the sum of the rest of
> the tests that hit.  Setup is spamassassin 3.1.0 invoked through amavisd
> on a linux postfix server.

Have you looked in /usr/share/spamassassin/50_scores.cf or similar?
(Location depending on installation method of course.)
If there is no score set the default is 1.0 for a normal rule, 0.0 for a
rule beginning with "__" (for use in meta rules) and 0.01 for a rule
beginning with "T_".

-- 
Magnus Holmgren

Re: Default score for UPPERCASE_75_100

[Matt Kettler <mk...@evi-inc.com>]

Fran Fabrizio wrote:
> 
> What's the default score for UPPERCASE_75_100?  This test does not
> appear to be documented at http://spamassassin.apache.org/tests_3_1_x.html.

I would STRONGLY suggest not relying on the information on that page for
anything other than a gross overview of SA.

If you want to know the score of a rule, grep 50_scores.cf for it:

grep "UPPERCASE_75_100" /usr/share/spamassassin/50_scores.cf
score UPPERCASE_75_100 1.394 1.040 0.809 1.371


> 
> I am examining a mail to figure out why it got a false positive.  User
> has no user_prefs or even .spamassassin directory.  If the default is 0,
> where else might this be set? 

If the default was 0, you'd never see the rule hit. Rules with 0 score are
completely disabled and are never evaluated.

> This is the last test I can't account
> for, and the score is still 1.2 or so higher than the sum of the rest of
> the tests that hit.  Setup is spamassassin 3.1.0 invoked through amavisd
> on a linux postfix server.
> 
> Thanks,
> Fran
>