Re: Comments about draft-hain-templin-ipv6-limitedrange

[Keith Moore <moore@cs.utk.edu>]

> For those who were not in the room this evening, my comments were that 
> the
> complaints on the list about the draft fall into a couple of broad
> categories:
>
> One class basically intends to tell people how to run their networks. 
> The
> IETF defines how the protocols work, not how individual networks are 
> run.
I've responded to this in a separate message.

> Another class basically ignores the reality that network managers will
> deploy address space for local use, no matter what the IETF does.
you might be right, but our job is produce designs that actually work.  
what you are proposing is that we abandon that task, on the assumption 
that users will break what we do anyway.  if you feel that this effort 
is a waste of your time feel free to concentrate your energies in other 
directions, but don't expect us the rest of us to give up simply 
because of your speculation.

> One comment I did have though was the delay of the milestone as 
> presented by the
> chairs due to comments seems inappropriate since neither Fred nor I 
> know of
> any outstanding issues.
other than that the set of constraints you wish to impose are 
unworkable?

> This document is detailing the reasons that
> individual network managers use private address space. It really 
> doesn't
> matter if the IETF believes they are wrong, this is what they do and 
> will
> continue to do to meet their requirements.
this is just a re-expression of the myth that everything in IPv6 has to 
be done like it was done in IPv4.  managers can choose to use IPv6 or 
not based on whether they think IPv6 suits their needs.  but if they 
choose to use IPv6 and expect IPv6 to work well even though they are 
using IPv6 in a way that conflicts with its design, they're deluded.  
individuals and vendors who encourage such delusions are not doing 
their audiences a service.

> Since consistency plays a significant role in recurring operational 
> costs, to achieve any deployment, the various proposals suggesting 
> that the IETF provide a pile of
> different approaches for segments of the problem space will need to 
> show
> these network managers how such a collection is cheaper than the 
> increased
> training and generally higher clue factor of the operations staff 
> needed to
> run them.
The notion that this is a single problem space is an illusion, or an 
artifice.  The IP address is a bit like Maslow's hammer - just because 
you can awkwardly use it as a tool to solve a particular problem does 
not mean that it's well-suited for all of those problems.

Keith


--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------