terry(_at_)ashtonwoodshomes(_dot_)com wrote:
-----Original Message-----
Subject: Re: Unencumbered Checking
<snip>
There are many instances were the check will be done after the SMTP
transaction. SpamAssassin is an obvious example.
Plus, the MUA checking is what makes this valuable against phishing.
Why?
If the email is rejected before being written to users mail spool files, it
still doesn't get to the
MUA. But rejecting at the MTA also saves bandwidth etc.
So why isn't checking at the MTA valuable against phishing?
Let me more clear - it can in theory be used as a check on the MUA level
or past MTA for phishing in those cases where the underlying MTA does
not support Sender-ID yet but the MUA or a filter running behind the MTA
does. Obviously it is better to reject stuff at MTA level. Another
possibility is when the MTA doesn't reject it out right btt assigns a score.
Of course this is in theory, take a look at William Leibzon's earlier
reply to this of why this is might not feasible.
Yakov