[AFS3-std] Encoding IPvN addresses (was Re: first draft: ubik update proposal)

[Jeffrey Hutzelman]

--On Tuesday, February 08, 2011 03:09:23 PM -0500 "Matt W. Benjamin" 
<matt@linuxbox.com> wrote:

> Hi,
>
> I am probably misunderstanding something, but is it the case that the
> address types of the future problem is the same one addressed by RFC
> 5665?

No, I don't think so.  That document is part of the ongoing solution to the 
"we (IETF) have too many protocol fields that lack proper registries" 
problem; it creates and provides initial contents and registration 
procedures for a registry for an existing protocol field used in ONC RPC. 
It's really not a good idea to try to interpret it out of that context.

Now, we might look at the origins of that field and how it is used in 
NFSv4, but I don't think it necessarily solves the same problem we're going 
after, I don't think representing everything as strings is a wonderful 
idea, and just because ONC RPC tied network and transport together in a 
particular context doesn't mean that doing so is appropriate in every case 
where an Rx application wishes to transport a network address.