who's for a DNS resource record for representing sparse arbitrary
key/value data in BTREE format?
RFC 5507 reiterates the IETF's conventional wisdom about the DNS, which
is chronically somewhat detached from reality. The CW says that adding
new record types is easy, and reusing TXT is nasty.
On the Internet I use, adding new record types is painful, due to all
of the software that has to be upgraded. The DNS servers are the easy
part; what's hard is all of the cruddy web provisioning systems that
don't believe that there are any records other than A, MX, TXT, and
maybe CNAME.
Using TXT records in distinctive name spaces has its problems, most
notably that you can get very confused if someone uses a wildcard
you're not expecting, but the problems are way less than trying to
get everyone's provisioning system to handle a BTREE record type.
I've been doing some experiments with B-tree records, will try and
write them up when I get back from RSA. Short summary: the higher
your message rate, the better the cache behavior.
R's,
John
_______________________________________________
Asrg mailing list
Asrg(_at_)irtf(_dot_)org
http://www.irtf.org/mailman/listinfo/asrg