Re: A follow up question

Tony Hain wrote:
> > Most of us -- MOST OF US -- have a clue.  That you refuse to 
> > recognize 
> > it and respect those opinions is unfortunate.
>
> Where is the document that shows what requirements need to be solved? 
How about starting with the IPng archives?!  They're actually more 
persistant than, say, an internet draft, anyway.
> How would you correlate the stated need from the network manager for:
> Stable addresses, both during ISP changes, and for intermittently
> connected networks.
I wouldn't.  I would decompose the network manager's stated need for 
stable addresses.  That's what responsible people do when they see a 
customer requirement with a bad solution.  But see below...
> with:
> Why is it that stable addresses are so necessary?  
> Could it be because renumbering is painful?  
> Could that be one source of our ills?
> &
>
> Fix the underlying problem.  Making renumbering easy.  
> If we don't do that, IPv6 is no better than Ipv4.
> ???
>
> While I agree that we need to get at the underlying problem, handwaving
> that easy renumbering will solve it is not helpful. In many cases the
> requirement for stable addresses is being handed to the enterprise
> network manager by unreasonable Sr. VPs who want to lower their internal
> development cost & don't care about anything more than getting their
> product out the door. Claiming that easy renumbering will solve that
> internal business issue is just an IETF fantasy.
This strikes me like the old adage, "You're safe if you buy IBM" (no 
offense to IBMers).  You were until you weren't because the processes 
you used were outmoded, only you didn't think about it.  Frankly, the 
fantasy here (and I love your charged use of words) is that IPv6 
provides any -- ANY -- added benefit to the customer base.(*)  That same 
Sr. VP ought to be asking why he is spending what will for the forseable 
future be a huge amount of money to cut over to IPv6 when IPv4 will suit 
the vast percentage of enterprises and users just as well as IPv6, given 
their intended use of SLs, as stated by some of those very same network 
managers.
Eliot

(*) Indeed I question whether even the mobility features will be 
properly taken advantage of, given the stated network manager need for 
stable addresses reconciled with their need for Internet connectivity 
(i.e., NAT).  What do you then use for your MN address?  If that mobile 
node wants to have a stable identifier, then it's using site-locals. 
Nonsense?  But if they don't use site-locals then they have all the same 
problems those network managers are complaining about, and you don't 
have a stable address for applications.