Hi Masataka,
Thanks very much for your response.
Returning to the discussion of the LISP protocol, I received the
following in an off-list message:
And no, none of the LISP advocates have ever claimed that LISP was the
only Locator Identifier Separation proposal or protocol.
I think this claim is implicit in the opening words of the main LISP
protocol draft http://tools.ietf.org/html/draft-ietf-lisp-15 which have
been present since version 08 of 2010-08-13:
This document describes the Locator/Identifier Separation Protocol
(LISP), . . .
I think the claim is also implicit in the title of this draft:
Locator/ID Separation Protocol (LISP)
which has remained unchanged since draft-farinacci-lisp-00 of 2007-01-17.
I think the claim is also implicit in the absence of any references to
other Locator-Identifier Separation protocols: GSE, HIP and now ILNP and
lesser known RRG proposals: GLI-Split, Name-Based Sockets and RANGI.
(See RRG msg06219.)
All these Loc-ID Separation protocols involve new host protocols to
implement the central principle of Loc-ID Separation: a new (ID)
namespace for uniquely identifying hosts. This requires a new
stack<->application interface and for all current applications to be
substantially rewritten - though ILNP is claimed not to require either.
The LISP protocol operates on totally different principles, which I
consider a good thing - as I wrote in the first message in this thread.
Does anyone have arguments as to why the LISP protocol is a Loc-ID
Separation protocol like those just mentioned?
Alternatively, does anyone argue why the definition of "Loc-ID
Separation" should be extended to include the LISP protocol's approach?
I believe that to extend its meaning to include this approach - and
therefore the approaches of Ivip and IRON - would fudge an important
distinction and render the term meaningless.
- Robin
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf