>So far I've been very happy with JunkMail. I'm only running a few tests and
>it's catching a lot of spam and porn. However, I'm noticing the occasional
>legitimate email from badly formatted clients. For example, JunkMail caught
>a confirmation email from an online service that one of my co-workers signed
>up for. This was a good email but it had badly formatted headers.
>Fortunately, I'm not rejecting or deleting emails as of yet but eventually I
>will. How do you all deal with emails that fail the BADHEADERS test because
>of poor mail clients/senders but are legit emails that need to be delivered?
>I'm looking for my "next step" in configuring JunkMail. Any advice is
>appreciated.
I think that Jim's suggestion of relying on the weighting system is the
best answer.
My personal opinion, though, is that the BADHEADERS test should have a high
weight towards the weighting system, as no mail client should be sending
out E-mail with non-RFC-compliant headers -- that's very bad. Given how
much spam has increased lately, I think we're getting to the point where
broken E-mail headers can't be ignored any longer. Note that the problem
doesn't lie with the overworked mail server administrator on the other side
-- it lies in the company that designed the mail client, that they are
collecting money from.
The SPAMHEADERS test (headers that are technically RFC-compliant , but
spamlike) will catch E-mail from quite a few poorly designed web sites, and
*should* be fixed, but since the headers are RFC-compliant, a lower weight
should be used with the SPAMHEADERS test.
Scott
---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]
---
This E-mail came from the Declude.JunkMail mailing list. To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.JunkMail". The archives can be found
at http://www.mail-archive.com.
