On Thu, Sep 16, 2004 at 02:46:17PM -0500, Kyle Hasselbacher wrote:

> 1. People sometimes spam forged to appear from my domain, and they
> write back with a "screw you" email (and I'm talking about literally

Gee, I'd think catching hate mail from joe-jobs is a perfect use of
TMDA. And if the end-user can't even be bothered to validate their
flame, I can't imagine why you'd think it useful to see it in your
inbox.

> 2. I've seen messages blocked and stay blocked because the return-path
> on the message is bad.

Again, this is a perfectly valid reason to block mail. There is *never*
a valid reason for email to have an invalid return path. People using
professional ISPs should never have to worry about it, and people
running their own mail servers should know better. So where's the
problem?

And the argument against using it in a business environment is spurious.
I use TMDA for all individual accounts in my business, and just bypass
TMDA for well-known corporate addresses where I don't want C/R. Why do
you think TMDA "needs work" in that regard?

I certainly wouldn't recommend TMDA to the technically illiterate, nor
is it necessarily the right tool for every environment. But I run a
number of anti-spam measures, including Bayesian filters, and honestly
find TMDA to be the single-most effective tool in my arsenal. 

Bogofilter gets it wrong about 15% of the time around here. On the other
hand, I have had only a handful of false positives from TMDA, and those
are almost always related to sender-domain addresses where the mail
server doesn't share the domain for which it was generated, and is easy
to fix for respondents I don't *want* to block.

Your mileage may vary. *shrug*

-- 
Find my Techno-Geek Journal at http://www.codegnome.org/geeklog/
_____________________________________________
tmda-users mailing list ([EMAIL PROTECTED])
http://tmda.net/lists/listinfo/tmda-users

Reply via email to