On Wed, 7 Nov 2001, Matt Crawford wrote:

> I still don't see it.  That an argument against unnumbered links, but
> it still doesn't imply any topological relation between the addresses
> of the endpoints.  Kernel tables and routing protocols are both able
> to cope with a p-p link from A::X to B::Y.  But, as I said, there may
> be operational concerns of which I am innocent.

I think that there is a routing issue - if I, as an arbitratry host, am
going to route packets destined for B::Y via A::X, I have to have a route
telling me this.

If the two nodes share a common prefix, A, then my route that gets me to
A::X will also get me to A::Y.

In the worst case with unrelated prefixes, I would need a host route
(/128) to get me to B::Y.

Ben

--
Ben Clifford   http://www.hawaga.org.uk/ben/
Telephone: United States 310 443-4485    United Kingdom 0709-227-5268



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