Pyzor, Razor, and DCC? {Scanned by HJMS}

[Peter Bonivart]

I hear you. I realized after my post that I might have started a flame
war if this list wasn't so nice. Anyone on the qmail list? ;)

I know they all are different but I still think they are similar enough
to make me choose one to begin with and I went with DCC because of
simplicity and efficiency which I'm really fond of. DCC doesn't need 10
Perl modules like Razor and it doesn't need Python like Pyzor. It just
works out of the box with SA if you follow their 4 line install guide
(including test), it doesn't need patches like Razor and it only needs 1
open port in the firewall, to have port 7 (echo) open like Razor is not
popular with FW admins.

I'm sure they all work well and those who chose to run two or three of
them are probably well of. I just tried to explain to someone who didn't
run any of them why I chose *one* and which one of them.

Thanks for the replies, I have learned more about Razor by reading them.
Of course I could still be completely wrong about most things. :)

/Peter Bonivart

--Unix lovers do it in the Sun

Sun Fire V210, Solaris 9, MailScanner 4.23-11, SpamAssassin 2.60 + DCC
1.2.9, ClamAV 20030829

Malcolm Ray wrote:
>>I think that since they all share the same idea you shouldn't use more
>>than one, just bump the score instead. At least in theory they should
>>give the same result most of the time. You wouldn't query the same RBL
>>twice, would you? You would double the score instead if you trusted it.
>>Less things to break and upgrade. Less network traffic which means
>>better performance.
>
>
> They're not *exactly* the same idea.  Razor's aim is to catch known spam
> (I assume the same is true of Pyzor, though I haven't used it).  DCC's
> job is to tell you when a particular message has been received by many
> recipients.  It doesn't distinguish between solicited and unsolicited
> bulk mail, so you need to be careful about whitelisting or risk tagging
> legitimate mailing list traffic.
>