On 6/9/2004 1:53 PM, Andrew Newton wrote:
1) During the MARID interim meeting, Ted Hardie suggested a dual record
type approach whereby TXT would be used in servers incapable of
supporting a new record type, but more capable servers would use a new
record type (specifically defined for MARID). Do you feel this is a
workable solution?
The correct approach is for TXT to be used for proto-typing but a binary
RR to be defined in the final spec. An informational document can define
mechansims for transition and fallback processing. Folks with broken
servers and firewalls and funky DNS-RPC gateways can survive as
second-class citizens until they're able to get up to speed with the
binary RR. It would be beneficial to the network if these RRs shared a
common name where they can be fetched with QTYPE=ALL.
2) Do you feel a MARID solution needs the capability of DNS wildcards?
Yes, but I don't think this requirement has been defined very well yet.
3) If you answered "yes" to both of the above questions, then is it
reasonable to expect DNS wildcard capability only with the new record
type and not the TXT usage (because TXT may be defined to use a
prefix)?
_MARID.*.domain.dom should be workable as long as fallback is adequately
defined (limited to the zone cut, as signalled by the SOA, which is how
many other DNS applications work, and which should not be a problem with
new applications like MARID).
--
Eric A. Hall http://www.ehsco.com/
Internet Core Protocols http://www.oreilly.com/catalog/coreprot/