inline.
Michael Richardson wrote:
I agree with Melinda.
I would very much like to be able to let the desk clerk at the hotel
know that I won't be paying for their "Internet" service, because it
wasn't RFCxxxx compliant. (I now wish that someone did get the trademark
on that word, and would deny it to locations that offer only NATwork service)
Well, I suspect the hotel clerk won't understand and certainly won't
care what an RFC is. But, their service provider does, and I'm more
hopeful that the RFP's and RFI's they produce and send to their vendors
includes a statement that their NATs need to be RFCxxxx compliant,
because it increases the ability of their networks to support
applications that are ultimately driving demand for those networks.
That's the only value I see in this situation.
For the the vendors that have a clue, and will likely be involved in
this process, they are likely already compliant. For those that aren't
compliant, they won't be there. That's tough for them.
Actually, it's surprisingly more complicated than one might think to
determine what the "right" thing is, in terms of NAT behavior and
treatment of basic protocols such as UDP. Even clueful vendors appear to
have differing treatments even within the same product families [1].
For some of these behaviors, such as the UDP binding timeout interval,
there is value in trying to shepherd everyone to arrive at a common
minimal value (as in, the binding timeout MUST be greater than X). For
that example, it's not an issue of clueful or clueless implementations,
but rather, getting folks at a consistent value. That is exactly what
standards are for.
Thanks,
Jonathan R.
[1]
http://www.ietf.org/internet-drafts/draft-jennings-midcom-stun-results-01.txt
--
Jonathan D. Rosenberg, Ph.D. 600 Lanidex Plaza
Chief Technology Officer Parsippany, NJ 07054-2711
dynamicsoft
jdrosen(_at_)dynamicsoft(_dot_)com FAX: (973) 952-5050
http://www.jdrosen.net PHONE: (973) 952-5000
http://www.dynamicsoft.com
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf