Some comments inline. But I will also respond on the new thread. Heiner
-----Ursprüngliche Mitteilung----- Von: Tony Li <tony...@tony.li> An: heinerhum...@aol.com; rrg@irtf.org Verschickt: Fr., 16. Apr. 2010, 1:26 Thema: Re: [rrg] Proposal for recommendation language Hi Heiner, I’m not sure that I see the point in going through these arguments yet again, but in the interests of completeness, I’ll recapitulate. ILNP is an IPv6-only solution. IPv6 may do what so ever - for more than a decade. No one really cares. You’re speaking for yourself. You don’t care. There are now lots of people who do care, in part due to the IPv4 address space runout. H: You are half right. I don't see that IPv6 is a future-prone protocol and you don't either (see your own next comment). However, it would be against RFC4984 to ignore the IPv4's scalability problem incl. the associated Moore's-law-resistance. Just supporting ILNP means ignoring the task given by RFC4984. Imho: ILNP is free to go ahead ( just as LISP was free to go ahead). I’m not sure what you’re referring to, but the fact of the matter is that we currently have exactly the same routing architecture for v6 and v4 and that the problem that we have been chartered to solve is to address that routing architecture. H:Exactly, it has the same architecture as v4, only more address space (enough to identify each square micrometer on this planet, why????) and more inappropriate overloads: not only more of the misconcepted multicast addresses, but even anycast addresses!!! As requested by 4984. H: Read what 4984 write about the hopes on IPv6! What is the ILNP locator all about? Is it routable? Or is it only mappable just like the MAC address or the current IPv4/v6 address? Is it at least PI ? What is it really and precisely? It is a topologically sensitive namespace for ‘subnets’, topologically allocated and not associated with a host. H:.. and will repeat the same mistakes as of PNNI - that protocol I once fervently supported, but which disproves itself, by requiring nodal topologies (the so-called complex node representation). How about the basic requirement "incremental deployability" ? Is it a non-issue for ILNP because the few IPv4-addressers aren't worth to be considered? Six years have been wasted so far and the actual decision of the chairs is blantly ignoring the task to come up with an architecture that fits a future internet. ILNP is incrementally deployable. Any host can individually deploy it and it will continue to interoperate with legacy hosts. H: If, as you say above,a new topologically sensitive namespace shall be provided, you have to change the existing addresses, right? When will you do that ? No flag day? Tony _______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg