Re: DNS RRTYPEs, the difficulty with

John Levine wrote:
> > Er, so? If the tool to bundle up the needed bits for SPF-as-text and
> > SPF-as-binary are similarly trivial, why should we care if people who
> > want to use a feature of the Internet can't persuade their provider to
> > make a trivial change.
> Because there's no point in checking records that don't exist.  RFC
> 4408 has been out for six years, and people who run large mail systems
> report that the number of type 99 records is basically zero.  To
> several significant digits, all type 99 queries are wasted traffic.
However, that was the original migration path expected which has 
finally materialized - Getting clients to support both record types 99 
and TXT first before recommending type 99 only publishing can be 
considered.  It would not have been feasible nor reasonable to 
recommend type 99 only publishing in 4408.
To the surprise of many, clients are doing the expected behavior of 
dual type queries - all expected to come with a short term waste.
Now its time for the BIS update to declare, pardon the pun, "Route 99" 
is now paved and safe to drive (begin publishing type 99 records).
--
HLS


_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf