On Wed, 20 Mar 2002, Kerry Nice wrote:
> I did email Chris Prillo of Lockergnome and tried to enlighten him.
> His response basically was that he was mad that people were using
> something that they didn't know how to use and it was too powerful.
Hey, that's a reasonable objection. I would complain if someone near me
were misusing an axe, and that's still a useful tool. ;)
[...]
> Maybe the newsletters I get are probably more spammy looking than a
> lot of others on the list.
More likely, most of the list are typical hacker-types who don't really
subscribe to many newsletter type lists. Certainly I don't -- they tend
to be far too fluffy for me. :)
[...]
> Does it seem worth while to create a corpus of legitimate mailing
> lists,
Yes. In fact, one of the best ways that you could both contribute to the
development of SpamAssassin is by helping Craig out with data for the
Corpus.
(Craig is busy right now and not paying attention so don't take this as
actually speaking for him, please. :)
A while back (check the archives) there was a call for people to run
their collections of email through the mass-check routine and supply the
details of that to Craig.
This is then fed into the GA to generate the scores for the next
release.
If your body of email, including the false positives, is part of that
evolutionary system then SpamAssassin will try very hard to learn the
difference between the SPAM and the non-SPAM email you get.
[...]
> I guess it depends on what the focus is here, do you want something
> that works great for a largely US based group with mostly technical
> email or is there a wider goal?
Both Justin and Craig have regularly expressed a desire to expand beyond
the US-technical market, as it were. :)
> Do you go for 100% spam catching with some false positives or do you
> miss some because you never want a false positive?
The default setting of five is intended to minimize the false positives
while still catching a lot of SPAM. :)
> I've also noticed, I get a lot of mailing lists sent to me as a
> digest. That pretty much guarantees that they will get tagged since
> the body is multiplied many times and there are more chances to get
> something spammy in them.
Maybe you should have procmail explode them for you before you feed them
into SpamAssassin...
That said, maybe submitting a rule that matched digest mail would be
useful. That way it would be likely to score down since, overall, it's
not actually going to be SPAM mail.
Daniel
--
Among those whom I like, I can find no common denominator,
but among those whom I love, I can: all of them make me laugh.
-- W.H. Auden
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
