> It doesn't matter how many times you write this, you can not make it
> become true. Brian keeps pointing out the simple case of an
> intermittently connected network getting a different prefix on each
> connect, but you keep ignoring it. STABLE ADDRESS SPACE IS A MAJOR
> APPLICATION REASON TO HAVE SL. 

Being the probable guilty party for introducing this thought back in
draft-*-site-prefixes-00.txt I can offer a slightly expanded perspective.

I don't think stable addresses per se is the key thing - it is
the robustness of the communication that is important.
This robustness has at least two factors that are relevant in this
discussion: the stability of the addresses, and the leakage of
non-global scope addresses. I think the question is how to weigh those
together.

In terms of the stability of the addresses one has to take into account
both stability as it relates to local communication and stability for
global communication. If you assume that the value/importance of local
communication is much higher than the value/importance of global communication
then site-locals make sense to explore. (FWIW that was my assumption
way back).
Such an assumption might make sense we say that a site is an
administrative concept (such as a company), but it makes less sense
if a site is a geographic concept and as I understand the original
thoughts a site was intended to be a geographic concept like a building
or a campus.
In any case, for a home user I suspect that the value/importance of
local communication would typically be less than the value/importance
of global communication. Thus the ISP offering a service with unstable
global addresses I don't think it would be that satisfactory
for the peer-to-peer communication that we wish to enable with IPv6,
even if there are stable site-local addresses so that the user can
communicate inside their home without a glitch.

Additionally, folks might want to use mobile-ipv6 and still get a quality
service. Since MIPv6 MNs are likely to use global addresses (presumably they
wish to be capable to move outside the home/site) their robust communication
require reasonably stable global addresses.

Finally, an enemy to robustness is complexity. Site-locals add complexity
in many places; applications, two-faced DNS configuration, etc.

So let's not loose sight of the fact that the goal is a robust network.

  Erik

--------------------------------------------------------------------
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