Re: WG Review: IPv6 Operations (v6ops)

[Pekka Savola <pekkas@netcore.fi>]

Thinking back after discussions with Jim et al, I have one additional 
issue to highlight:

On Thu, 5 Sep 2002, Jun-ichiro itojun Hagino wrote:

> >Your just plain wrong.  Do you think users are stupid?  Do you think
> >they even care about what we say here in the IETF work?  No they do not.
> > What they want are IPv6 solutions.  And they will all want different
> >ones.
> 
> 	(chair hat off)
> 
> 	the comments i get from the people most is that, they have big trouble
> 	picking which ngtrans mechanisms to pick (for instance, which one
> 	to implement on their router, which one to deploy to their network).
> 	so what we need to do is cut down number of deployment mechanisms,
> 	and help people deploy network in actual use.

...

Most importantly:
 
> 	as i said in IETF54 plenary, there's no need to deploy fancy IPv6
> 	network.  keep it simple, stupid and robust.

There are two cases, I think:

1) network (ISP/core) transition.  Folks who feel stuff like IGP or BGP 
tunneling is really needed etc.

Most probably agree that complexity in the core networks is highly 
undesirable.

2) site/home transition.  Current "approved" toolbox offers no solutions
to e.g. provide IPv6 access through NAT's (well, it's a double-edged
sword..), and some other solutions like ISATAP may be useful especially in
large networks e.g. where admins wouldn't want to bother to build proper
dual-stack infrastructure.

-- 
Pekka Savola                 "Tell me of difficulties surmounted,
Netcore Oy                   not those you stumble over and fall"
Systems. Networks. Security.  -- Robert Jordan: A Crown of Swords