spamassassin 2.54 released

Julian Field mailscanner at
Tue May 13 12:01:09 IST 2003

At 21:10 12/05/2003, you wrote:
>At 20:36 12/05/2003, you wrote:
>I am currently running roughly 60,000 messages through 2.54, and will then
>do 2.53, to see what the spam score distribution looks like. This will tell
>us if it will detect more spam, and whether you need to move your threshold
>Will post later when I have some results.

Attached is a gif of the distribution of spam scores you get with my 60,000
message test set.
Basically it generates a lot less "<=0" values and a lots more 1, 2 and 3
values to compensate. Once you get up to 5 or so, the differences between
the 2.53 and 2.54 are pretty minimal.

So you shouldn't hopefully need to adjust your threshold scores, unless you
use particularly low scores (<5).

Bear in mind that my test set is getting a bit old now, so these stats
won't include the much-improved results from spammers defeating 2.53 rules.

Hope that's some use to someone...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Book2_26737_image001.gif
Type: image/gif
Size: 5579 bytes
Desc: not available
Url :
-------------- next part --------------
Julian Field
MailScanner thanks transtec Computers for their support

More information about the MailScanner mailing list