bogofilter-0.13.0 available

[Jim Correia]

On Thursday, May 22, 2003, at 07:26  AM, Greg Louis wrote:

> Not necessarily.  Especially with the high minimum deviations we have
> been advocating before making these recent changes, errors may arise
> very quickly after a mistake in training.

Like David, I've been using -u because I find it convenient. I retrain 
on the mistakes daily (but if I am away it may be a few days.)

When I wrote earlier in this thread there was one particular example I 
was thinking about that reenforced  Greg's warning. (I had to dig to 
find this, because it was just in the spam pile with the rest of the 
stuff...

Perhaps it is an extreme example, perhaps not, but the different was an 
order of magnitude.

I got two identical messages in short proximity. The headers were the 
same (except for the date and message id as you would expect.)

The body was short, and differed only by the line

email : scm at vsn.com

vs.

email : his at tyk.com

The spam scores were

X-Bogosity: No, tests=bogofilter, spamicity=0.117262, version=0.12.2
X-Bogosity: No, tests=bogofilter, spamicity=0.011255, version=0.12.2

Jim