On Thu, 17 Jun 2010, [email protected] wrote:
You hereby play foul to those who want location-based routing, because you educate people to understand something else than a locator ought to mean. And you pave the ground such that people would get confused when someone uses the word locator in its true meaning.
I don't think ILNP changes the notion of "locator" at all. The "address" continues to have a topological component as far as routing is concerned.
ILNP provides a clear line as to what part of the address may be considered for topological routing. Changes beyond that are just host-side, to decouple ULPs from the locator portion - that's intrinsically a good thing, methinks.
Ratifying an L/I split, and that ULPs should not bind the I to any L for ILNP seems intrinsically a good thing. It would not prejudice any routing work, I think.
The ony thing that's unclear to me is whether the ILNPv6 space should be a new address family or a defined subset of the IPv6 number space. Indeed, it's unclear to me what the ILNP authors think on this topic. It perhaps would be nice to see worked examples - the papers seem not to have them.
I ask this because it would be good to strive for maximum compatibility as much as possible. E.g. it'd be nice if ILNPv6 sockets would have IPv6 addresses and could be used to accept IPv6 connections, and connect out to legacy IPv6. There are a lot of details here that don't yet seem to be fully described.
Also, it's perhaps unfortunate that the links on the main ILNP to some of the more detailed papers go to academic journal paywalls. It would be good if there were direct links to the PDFs concerned.
regards, -- Paul Jakma [email protected] Key ID: 64A2FF6A Fortune: God requireth not a uniformity of religion. - Roger Williams _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
