Hi all,
The thread about Kim's .INTERNAL draft reminded me of a different draft I wrote
a while back.
In draft-jabley-sink-arpa I proposed that SINK.ARPA be reserved by the IETF as
never existing because I thought there were protocol reasons to have a name
that reliably didn't exist. For example, you might want to specify a NOTIFY
target for a zone in its SOA.MNAME field as something that doesn't exist and be
clear about what you are doing.
That draft proposed a registry, but in this case it was about assignment policy
("reserved names") not special handling. Assignment policy in the ARPA domain
is owned by the IAB so it made sense to me to document this small part of it in
a protocol registry.
Nobody liked this idea at the time and it withered on the vine. I seem to
remember one reaction being (paraphrasing) "this is a draft that literally
recommends doing nothing, we don't need a draft for that" which I don't quite
agree with but which made me smile at the time.
INTERNAL itself (that actual DNS name) doesn't fit the gap I thought I saw at
the time, since (a) it's not a hostname, and (b) you can't reliably guarantee
that names under INTERNAL don't exist from all vantage points (the point of
INTERNAL is that names in that domain probably do exist, e.g. in internal
networks).
Just thought I'd mention it in case anybody is feeling renewed enthusiasm for
names that don't exist and is willing to give it another look. I don't think
this is an idea that will change the world but I do think it has non-zero
value, even if it's pretty close to zero.
Joe
_______________________________________________
DNSOP mailing list -- dnsop@ietf.org
To unsubscribe send an email to dnsop-le...@ietf.org
