interesting-people message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


  • From: Dave Farber <>
  • To: ip <>
  • Date: Mon, 14 Apr 2003 14:25:10 -0400

------ Forwarded Message
From: Bob Frankston <>
Date: Mon, 14 Apr 2003 14:18:08 -0400
To:, "'ip'" <>
Cc: "'Bob Hinden'" <>, "'David P. Reed'"
<>, Richard Shockey <>

We already have IP address portability! It's called the DNS. The problem is
that people remember the old days when the IP address was really an address
and not a routing path or circuit ID. This is also my problem with IP
mobility -- it is like trying to provide reliability at the IP layer rather
than the TCP layer. The persistent connections should be defined at the
layer above IP between end point identifiers. The DNS can serve as source of
identifiers, especially since it maps those to IP addresses.

There may be a reason for some local mapping of IP addresses among a set of
proximate end points but trying to create a new mechanism for IP mobility is
akin to creating another DNS focused on the accidental properties of one
application when the current DNS address the issue. (I can go into more
detail if there is interest).

While I welcome cell phone portability it is a special case that diverts
attention from a simpler approach already being proposed. Once again the
cellular phone industry is trying to create its own mechanisms that presume
there is some special about cell phones. Maybe there is since special is a
euphemism these days.

Using the DNS (ENUM) to translate phone numbers is far simpler. The problem
is indeed political since if the ITU doesn't manage to lockdown the complete
hierarchy then it becomes too easy to leave the PSTN (not just jump to
another carrier) and there would be little reason not to. Imagine having
your own NS records in the ENUM space -- you could define your own local
phone numbers and to your own follow-me implementation. Free phone calls are
a minor threat compared with the ability to redefine telephony to be much
more useful and effective. The only thing interesting is that the "names"
would be limited to digits on phones as an accommodation to legacy devices
with legacy interfaces.

ILECs realize that this is extremely threatening since the only reason they
exist is the difficulty of interconnecting the PSTN to user-defined VoIP.
Unlike the RIAA there is surprisingly little outrage enough the economic
consequences are far far worse. I guess it's too hard for most people to
recognize that telephony is a trivial app.

Given all this I guess I do need to belabor the point that the problems with
the DNS are largely due to the secondary and flawed agendas (meaning and
authority) rather than the mechanism itself (though it is far form perfect).

------ End of Forwarded Message

You are subscribed as
To manage your subscription, go to

Archives at:

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Search: Match: Sort by:
Words: | Help


Powered by eList eXpress LLC