On 6/15/2004 4:32 PM, Eric A. Hall wrote:
SPF is worth publishing, but the value for checking is marginal and
will probably stay that way.
ps--if the value for checking is proportional to the cost (which it is),
then the value goes down as:
* record sizes require fallback processing
* firewalls/bugs/etc break fallback processing
* large numbers of large records make caching expensive/difficult
* multiple RRs of the same type (but not the same meaning) require
deeper analysis (= more processing, = more blocking)
I know some folks have said that they will actively work against large
RRs, but for me I'll just not implement the checker. Almost every mail
that was ever trapped with SPF was also trapped via other analysis (such
as spamassassin checks). Slightly improved heuristic trapping just isn't
worth the extra expense, especially at large scales.
--
Eric A. Hall http://www.ehsco.com/
Internet Core Protocols http://www.oreilly.com/catalog/coreprot/