At 07:58 PM 4/2/2003 -0500, Aaron Sherman wrote:
Notice the use of the word first. You cannot forge the first header. Or
rather, you CAN forge the first header, and then the receiving MTA will
promptly add its own as the NEW first header. Thus the first header is
always added by the local MTA. If S
On Wed, 2003-04-02 at 14:23, Mike Vanecek wrote:
> > su -c "perl -MCPAN -e shell"
[...]
> > install CPAN
> > [it does its thing]
> > install Term::ReadLine::Gnu
> > [no time to get coffee, but a short wait]
> > reload cpan
> > [optionally update "install LWP::UserAgent"
Version 2.52
timelog_path /path/to/dir (default: NULL)
If you set this value, SpamAssassin will try to create
logfiles for each message it processes and dump infor?
mation on how fast it ran, and in which parts of the
code the time was sp
On Wed, Apr 02, 2003 at 12:57:12PM -0500, Matt Kettler wrote:
> >2. Bayesian learning simply is not happening.
You were complaining that there was no BAYES_* rule results. That's
different than learning. Are there any bayes* files in the SA dir?
How many ham/spam have been learned so far?
--
R
I'm having a large number of problems with SpamAssassin presently, and the
docs don't say _anything_ that's helpful.
1. My users' user_prefs files are not being read. In local.cf, I have
allow_user_rules set to 1, and I've restarted spamd at least 3 times since
making that change, with no change i
On Tue, 1 Apr 2003, Michael DeSimone wrote:
> Knowing exactly what SA does to flagged mail would also be nice - can I make
> it bounce?
SpamAssassin is kind of like Statler & Waldorf, the old guys who sat in
the high box seats and made catcalls in the old "Muppet Show" -- like S&W,
SA does nothin
On Wed, 2003-04-02 at 09:18, Ian MacDougall wrote:
> No suggestions or ideas, but the same thought occured to me. We seem to
> be seeing a lot more spam recently (and since starting to use
> SpamAssassin). I'm putting it down to there just being an ever
> increasing amount of spam.
The mailing lis
