[Raven] Thoughts on the monitoring problem.

["Turner W. Rentz, III" <treyr@atr.net>]

The monitoring problem isn't a problem for protocols, it is a directory
service problem -
that for some bizarre reason, the fed still hasn't tried to work through.

What does the net offer to telephony today? Well apart from shredding up the
voice
over ip stuff with routing delay, and cloud based architectures (as opposed
to switched)
-  a zillion passwords , having to sustain twenty different authentication
schemes every
time you have to connect to another person (as opposed to another +machine+
where the dns is nicely propagated and working fine, thanks!)  and never
getting near anything like quality of service !

 For any type or level of telephony (historically, a high quality of service
endeavour)
or any other transaction which has to be mission critical
or have some trust basis, the current net environment is unsustainable
and unusable, just like voice recognition.

There is still, however, only one person in all of those databases,
and only one person who needs to ethically use the network
infrastructure: whether regulated or not - the end user. Currently,
Mr. Anonymous.

Solution.
An ITU compliant, standards based international directory service
for people where we can +safely+ and with some degree of +trust+
and +constancy+ store our biometrics and/or public keys - and
the simple integration into directory services so that people
can connect to people and not have to deal with ip addresses.


The government(s) can check crypto  with their own hardware and leave the
ISPs out of
it. They will have to fund it, because it will be a high transaction/ master
/ replica
type problem. ISPs can simply run the additional service on a small server
somewhere
(even 20 million users is still not that much seek time for an X500
database), that the feds should
offer for free, to keep the replicas intact and distributed.

The protocol problem is worked out well in the X 500 models for database
transactions and protocols such as LDAP and standards such as
X509 work well with all the necessary elements

Summary.
My recommendation is simple. Give the net a scalable, enterprise ready
database of users, with privacy and trust, and subsidize it by the tax
payer.
Keep the ISPs out of it, and let a tiny little bit of social structure play
across the net where we can let the feds play in their sandbox and
maybe even counteract the current insanity gripping the user base - and
follow historical precedent in organizing telephone numbers, but this
time - indexing them to real people and not just land lines. A global X500
directory.







_______________________________________________
raven mailing list
raven@ietf.org
http://www.ietf.org/mailman/listinfo/raven