I would like to suggest a significant change to they way whitelisting works.  What I've noticed over time is that some senders might always fail a particular test (e.g. AOL and NOABUSE, or a company whose employees all use Outlook and fail the CMDSPACE test), so it occurred to me that if we could whitelist a sender or from domain/address for a particular test, but test against all others, we could cut down significantly on spam getting through because of a whitelisted from address.
 
So the syntax might be
 
FROMFILE
------------------
sampledomain1.com # whitelists the domain for all tests
[EMAIL PROTECTED] # whitelists the email address for all tests
 
sampledomain2.com NOABUSE  # whitelists the domain for only the NOABUSE test
[EMAIL PROTECTED] NOABUSE  # whitelists the domain for only the NOABUSE test
 
Whitelisting with no test name would then act just like current whitelisting
 
The same could apply to an IPFILE as well for checks against sending mail server.
 
I would also like to reiterate the desire to have internal wildcarding in the whitelist entry.  I see a lot of newsletter senders these days that use patterns like
 
XYZnews-<random string>@XYZcompany.com
 
or
 
<random string>[EMAIL PROTECTED]
 
In either case if we could put a whitelist entry like
 
 
or whitelist with the implied leading wildcard as
 
 
it would help significantly.
 
Currently a whitelist entry of [EMAIL PROTECTED] will not whitelist [EMAIL PROTECTED]
 
Thoughts?

Darin.
 
 

Reply via email to