> It's a bad idea to use SPFPASS.
Agreed.
> SPF Fail should in theory give you similar results to SPAMDOMAINS,
> though it's not very widely adopted at present so the hits should be
> lower.
SPF FAIL is much, much more robust than SPAMDOMAINS, and there's no
deployment difficulty that SPF FAIL suffers from that doesn't affect
SPAMDOMAINS as well!
For just the area that SPAMDOMAINS covers, sender-domain-to-sender-PTR
mapping, SPF offers real-time updating of data from the authoritative
source, while SPAMDOMAINS is a moving target using "best-guess"
entries based on a local corpus and some community involvement.
> ...this test would likely just cause more false positives,
> especially when domains like aol.com are in both my SPAMDOMAINS file
> and making use of SPF records.
First, AOL is using a ?all in their record, so there should be no
duplication.
Second, I do have a couple of new feature suggestions in this regard:
- The exclusive use of an SPF record if it exists, overriding any
SPAMDOMAINS entries for the same domain (and perhaps a setting
DNSBEFORELOCAL OFF that would reverse the preference).
- The ability to use SPF syntax in a SPAMDOMAINS-like local file
(SPFLOCAL), which by default would be used only if no public SPF
record exists for the domain (with preference reversible, as above).
Since the Declude SPF parser is already in place, I'd think that
reading SPF-style entries (like example.com v=spf1...) from such a
file could be made possible.
--Sandy
------------------------------------
Sanford Whiteman, Chief Technologist
Broadleaf Systems, a division of
Cypress Integrated Systems, Inc.
e-mail: [EMAIL PROTECTED]
SpamAssassin plugs into Declude!
http://www.mailmage.com/download/software/freeutils/SPAMC32/Release/
---
[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.