At 22:54 22/07/2005, Brian E Carpenter wrote:
I wouldn't change a word in RFC 2826.
The problem with RFC 2826 is that it links (for information) a unique
domain name resolution (what we want) with a unique authoritative root file
(we do not care it is "unique", we want the one we use to be pertinent).
Confusing the description with the described space was a way to protect the
name space, but it unfortunately lead to open roots confusion and to
alt-root suspicion (I only know one: ICANN with .biz) and to the lack of
preparation in front to PAD (private roots).
Now, I agree with Stephane and ICANN that a lot is/can to be done. We just
have to remember IMHO the namespace is the same as geographical space: the
map does not build the geography and no one thinks that geography depends
on the map he uses.
Except may be politics.
But there may and is to be a lot of innovative thinking. ICP-3 is
excellent, starting with a good review of RFC 2826, rooting into RFC 920
which is the true basis of the DNS as we live it, and calling on
experimentation and proposing avenues for the research and development with
classes (the way to the externets).
jfc
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf