On Aug 31, 2005, at 2:37 PM, John Rudd wrote:
On Aug 31, 2005, at 1:49 PM, Matt Kettler wrote:
Bowie Bailey wrote:
Ok. I thought that might be the case, but I wasn't sure. The
--add-addr-to-blacklist entry in the man page states that it is added
with a high score, but the --add-addr-to-whitelist entry doesn't
mention score.
The use of the word "whitelist" in the man page is a bit confusing
since the entire database is called a "whitelist". For example:
--add-addr-to-BLACKLIST adds the address to the WHITELIST ... huh?
:)
Bowie
Quite frankly, using either white or black in *any* reference to the
AWL is a
complete misnomer in the first place. The AWL isn't really an
"Automatic White
List", it's just named that because that describes one of it's
behaviors.
It's really more of a "[SpamAssassin] Score Handicap List".
What about:
--set-high-handicap-score-for-message
(replaces --add-to-blacklist)
--set-low-handicap-score-for-message
(replaces --add-to-whitelist)
--remove-message-from-handicap-list
(replaces --remove-from-whitelist)
--set-high-handicap-score-for-address
(replaces --add-addr-to-blacklist)
--set-low-handicap-score-for-address
(replaces --add-addr-to-whitelist)
--remove-address-from-handicap-list
(replaces --remove-addr-from-whitelist)
Oh, and in the name of promoting "Abstract Data Types" (esp. since it
seems that the AWL database type has changed more often than the phases
of the moon), I would also like to propose two more commands for
"spamassassin" itself (not for sa-learn*, not for yet another program
that will disappear one day, etc.):
--show-handicap-list
(display every entry in the list, with its handicap score, and all
relevant information, in a nice presentable fashion)
--clear-handicap-list
(to follow up nicely with the main purpose of this thread in the
first place, and which means that the owner of that list doesn't have
to try to track down where their system is keeping it, etc.).
(* though, all 8 of these options ought to be available through
sa-learn as well, so that sa-learn can also be a one-stop shop for
managing both bayes and the handicap list)
