John,
You can't use STATISTICS - you must use protocol methodology - LOGIC.
You can't tell the world RFC 5617 is available as a DKIM "Add-on"
technology YET support the intentional ignorance of its existence or
hope that no one will support it.
You can't have it both ways. Either you will support RFC 5617 or you,
as the author of RFC 5617 propose to get rid of it.
Either way, as a SMTP and LIST SERVER developer, it will at least give
us and in the general sense, all SMTP and LIST SERVER developers, a
protocol consistent methodology to increase DKIM adoption and unleash
it to our customers.
--
HLS
John Levine wrote:
Before you report your personal experiences, could you include data about
the
receivers, please?
Well, here's a little bit of my data. Over the past day, my busiest
DNS server has seen 2014 lookups of _domainkey records, of which only
38 were for _adsp.
A surprisingly high 806 requests from 486 hosts were for plain
_domainkey with no selector. I can't tell if they were looking for
the old DK policy record, or if their DKIM implementations were
broken. I also saw 541 requests for _policy._domainkey, which was
used in a few drafts back in 2005, and 244 for _ssp.
Somewhat to my surprise, the 38 _adsp requests were from 34 different
hosts, but it's still fair to say that the fraction of hosts doing
_adsp lookups is very small, so it's not surprising that there's no
observable effect, good or bad.
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html