From: "Steve Thomas" <[EMAIL PROTECTED]>
> On Thu, Mar 11, 2004 at 06:15:05PM -0800, jdow is rumored to have said:
> >
> > I read all the above stuff on their site...
> > <snip>
> > ...I do not see any evidence outside of their website that they exist,
> > win lawsuits, deter spammers, and so forth.
>
> Erm...
>
> http://www.habeas.com/companyPressPR.html#victory1
>
> Give the new kid on the block a chance to play. Any business model based
on the legal system is not going to provide results overnight - give them a
few years to prove whether or not their model will work. So far, it looks
like it's working reasonably well.
I may have given it a +1 score for now. Do note that I review all the
marked spam by subject and sender before discarding it. I also sometimes
review the SpamAssassin applied wrapper. (Sometimes I review due to an
expected astronomical score and other things because of some doubt about
the potential it was a ham message. At the moment I am running with no
false positives for spam for well over a month now. I've had several
false negatives popping up from time to time. I feed them back into a
newly automated spamlearning trick using Outlook Express, imap, and a
tiny C futility I built so SA does not repeatedly "learn" on the mbox
format message 1 that my available IMAP tool insists should be there.
Once I train with a message I seldom have Bayes fail to catch it when
I retest the spam explicitly.)
So if I start to see X-Habeas marked false positives you can bet I will
modify the scoring. I would like to see them win. However, the latest
craze of spam generating viruses makes it rather hard to track down the
authors of the spam to nail them. And going after the manufacturer of
the spammed product is also a bad thing because they might have a
competitor or enemy who wants to injure then indirectly. A worst case
analysis of the Habeas model suggests it's worthy of only a small score
plus or minus in the long run. (Please do be aware of the fact that I
an a well known slightly paranoid searcher for worst case scenarios
when presented with new ideas. So I zero in on the flaws rather quickly,
unless, alas, they are my own ideas. If I could only go after my own
ideas as harshly I'd be better at my job - fewer bugs.)
{^_^}
