On Wed, Aug 21, 2002 at 09:37:50PM -0500, Igor Izyumin wrote:
> Well, yes, but it does rely on keywords and headers to score the email.  
> Sometimes, even a wrong header combination will cause it to trigger.  It 
> doesn't know the context of the email message (you'd need AI for that), yet 
> it scores email based solely on its content (unlike razor-type tools which 
> compare a message against a database).  

Yeah but spamassassin uses every piece of information about the message
it can to come up with a score.  Some of that is content based.  It can
even use Razor to get some of that score.  Plus with autowhitelisting
spamassassin gets better and better about false positives.  I can't
remember the last time it grabed an email that I really wanted and put
it in the spam folder.  Fact is I rarely even looka the spam folder.

> My point wasn't that it sucks, just that it is far from foolproof, does have 
> false positives, and doesn't have a "reasonable" default configuration.  If 
> it is incorporated into Mandrake as an easily enabled option, you would 
> probably have to provide a configuration drake for it.

Well I don't really think it has a bad default config.  I've made very
few alterations to mine.  A few blacklist and whitelist entries where I
disagree with it about what is spam.  And a couple extra rulesets.  But
for the most part it's the default config.

Now most of my users are using the default config.  And I rarely get
complaints about it.  I've been running spamassassin for a while now for
all my users.  So I think I have enough experience to make these
statements.

But a drake type tool or hell any tool to help users config spamassassin
woudl be a good thing...

-- 
Ben Reser <[EMAIL PROTECTED]>
http://ben.reser.org

If your love has no hope of being welcomed do not voice it; for if it 
be silent it can endure, a guarded flame, within you.
- The Wisdom of the Sands

Reply via email to