On Thu, Apr 03, 2003 at 06:59:19PM -0500, Margaret Wasserman wrote:
> If the consensus to deprecate them does not hold, I'm not quite sure
> what we are going to do...  We certainly don't have consensus to do
> any further work on site-locals, and it isn't clear how we can
> ever finalize the scoped addressing architecture without some type
> of decision on this issue.  Perhaps we can break out the non-contentious
> parts and advance those parts?
> 

It seems to me that the way out of this is to find some mechanism by which
site-locals can be considered reasonably unique.  I think it would be
sensible to define several different mechanisms for this (perhaps each
with a different prefix) as cases can be made for choosing uniqeness
based on time, geography, mac address etc.

e.g. geographical prefixes allow a degree of control and aggregation,
time or mac-address prefixs are more suiteed to zero-conf etc

If they're recommended and easy to use then I see no reason why network
managers would choose not to use them, as it's a benefit for them.

I admit I'm not an expert with regards to the scoping issues here, but
personally I don't see that these private-unique prefixes need to be
given any special treatment other than that they're the least preferred
prefix used.

Personally I do not give the "site locals + nat will be used because
re-numbering is hard" argument much credit, that's a re-numbering
issue not a site-local one.  I think it's reasonable to expect an "ipv6
renumber" command in routers, at least if re-numbering is going to be
a major concern, and for DNS I'm sure something similar can be done.

Well that's my take on it anyway :)

Cheers,

Mike
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to