ATMARP and NHRP

[Andy Malis <[e-mail]>]

I would like to respond to the recent messages on multiple ATM ARP
servers on the ip-atm list by stepping back to a bit of a
meta-discussion.

I see the possibility of a lot of otherwise unnecessary work going on,
first in the IP/ATM WG and then in people's implementations, to
bullet-proof a limited-lifetime protocol (one that we've already
agreed isn't going to go past the "Proposed Standard" stage).

I will admit up front that I have two vested interests here:

As ROLC chair, I would MUCH rather see the brain cycles we have
available in the IP/ATM and ROLC WGs spent on specifying ATMARP-NHRP
interactions and transition, and bullet-proofing NHRP, than spent on
bullet-proofing ATMARP.

As a network equipment vendor, I want several things: to have to do
less, rather than more, development in order to deliver the right
products to my customers, and to prevent the possibility of confusion
in the marketplace (including the endstation vendor community).  To
the extent that ATMARP and NHRP cooperate and have a solid transition
plan, rather than both continuing to be developed (even though we've
already committed to the transition), work is conserved and confusion
is minimized.

Am I the only one that thinks this makes sense???  Please let me
know!!!

I cced the rolc mailing list because the ROLC WG has to be involved in
the transition planning, and not everyone on the list may be aware of
the ongoing ip-atm discussion.  My apologies to those of you that will
receive this thread twice.

Looking forward to your feedback,
Andy
__________________________________________________________________________
Andrew G. Malis   [e-mail]             [phone]
Ascom Timeplex    289 Great Rd., Acton MA 01720 USA   FAX: [phone]