I would rather have a 6 month pushback from the IESG than a spec that is
dependent on the next release of Windows Server to become viable.
The only problem is that I don't know what it means.
Someone was predicting another round of DNS wars, of the define a new
RR vs.reusing TXT variety, with hostilities breaking out via the IESG.
My position, which I believe is widely shared among the DKIM group, is
that the _domainkey subdomain introduces a new namespace within which
all TXT records must be DKIM keys, so we're not overloading any
existing TXT record. We might want to add a sentence or two warning
people that they'll be sorry if they put any other kind of TXT record
under _domainkey, but that's cosmetic. This is a fairly important
principle, since we're doing the same thing for the SRV records in CSV
and will probably do a lot of other namespaces that redefine existing
record types in the future.
The sentiment up at >>> is that a new RR type is such a bad idea that
it'd be worth a six month fight with the IESG to prevent it, but I
think we all agree that a much better outcome would be to go ahead
with the technically sound plan we have without the detour.
R's,
John
_______________________________________________
NOTE WELL: This list operates according to
http://mipassoc.org/dkim/ietf-list-rules.html