Greetings,I agree with jefsey, but don't really
think DNIC's are ideal, but there thinking is closer to the requirements for the
future, certainly.
When H.323 teleconferencing on broadband starts to be a serious
service; (as it surely will), I think synchronizing it to PSTN phone heirarchies
would be a liability. A recommendation might flow along these lines to find a recepient
in the new name space:
where a person is known to be associated with a server;
(ex.)
http://www.stanford.edu
then a ldap server is presumed to exist at
dir.stanford.edu
and
via HTTP and or LDAP base protocols presents named persons for messaging.
One
mapping lives at:
http://dir.stanford.edu
which is a human readable directory.
The
same applies for mapping SNMP addresses to servers for (I hate this term) "cable
telephony". If the email ws known to be jane@winterland.no then a source of a WWW
directory would be
http://dir.winterland.no
or
ldap://dir.winterland.no
note
that since an email address is unique, there lookup is trivial; Though it is contentious
how deeper domains might map this.
Internet to telephone gateways are nessessary
evils and the telco's will always encomber phone numbers them with intellectual property
rights; (like yellow page issues). Doesn't matter how many alleged watchdog agencies,
non-profit corps, etc that are interposed. The other problems include, is for instance
a "dialed" number is a NANP number (202) 345-6060 and it is mapped to a trivial secondary
identity; ex. {2023456060} how does the telecom switch decide whether to route the
last in organization part? By phone or through packet resources usually interfaced
to a computer?
Worse, services like call redirection, multi-user chats, etc have
been traditionally required hardware (like answering machines), or are fee
generating services. In the new heirarchy, costs should be driven to zero and replaced
with ownership. The DNS achieves this and building on it makes sense.
Finding persons
in a completely anonymous mode anywhere over a lifetime; (who want to be found),
I think is a slightly different problem. Even if the world starts with a LDAP slightly
geographic system, if all user/systems select a random very large number and associate
it with themseleves on the initial public servers, once a global roaming method is
on the planning horizon simple robots can walk the DNS and do whatever is required.
Security issues as usual are the most vexing. However, you perhaps can make the assumption
first contacts can be unverified, and PGP or some other crypto is the second high
certainty step.
I hope I''m not rambling, I hope I am thinking out loud.
Dan
Kolis