On Fri, 2009-07-10 at 06:12 -0700, MrGibbage wrote:
> Just a little off topic here, but relevant.  When I test SA, I log into a
> bash shell.  I set my environment variables in .bash_profile (loading
> changes with the 'source' command).
>
I use spamc/spamd and the answer is simple. I use a script that runs
spamc to submit test messages to my live spamd. The only disadvantage I
can see is that this will probably distort any spam statistics you
generate from /var/log/maillog.

Now I have a second copy of spamd on this laptop, configured identically
to the live copy, and run it only when I need to test some new or
updated personal rules. The only additions were scripts to:

- copy local.cf and any associated files to /etc/mail/spamassassin
  and start/stop/reload spamd on the laptop

- install the revised local.cf on my mail host and restart its copy
  of spamd.  

> Sometimes when testing, I get different results than I do when processing 
> real mail.
>
Same here, but that is unimportant to me, since the differences only
affect standard rules: the sole purpose of this test setup is to confirm
that new and modified private rules work as expected and don't generate
FPs. As always ymmv.


Martin


Reply via email to