Summary: a provider independent addressing solution is proposed so that
site-locals are not necessary.
One of the chief reasons proposed for the use of site-locals is for stable
addressing (especially if you need to change ISP's). A nicer solution, that
has so far proven unimplementable is provider independent addressing. The
problem with provider independent addressing is that, because routing is no
longer aggregated, it becomes too difficult for our current hardware limits.
I've also heard proposals for allocating prefixes to each geographical point on
earth. I propose a modification of this idea. Allocate a /32 (or some
suitable number) to each region (where each region is about the size of
California, but is based on politics rather than size). Then each ISP in that
region could coordinate to delegate prefixes out of that slice of that /32.
Hence, we have aggregation on the region level, but within that slice, the
routing would be flat (i.e. prefixes are not associated with any ISP). Flat
routing on a region level is doable if the region isn't too large. I can see
that some people might complain that physically relocating to another region
would cause renumbering, yet I propose that physically relocating to another
region is more difficult in and of itself than just renumbering.
Of course, I haven't done all my reading yet, so forgive me if I'm missing
an obvious problem.
-jj
--
def qsort(l): # Purely functional implementation of QSort in Python.
if not len(l): return []
return (qsort([i for i in l[1:] if i < l[0]]) + [l[0]] +
qsort([i for i in l[1:] if i >= l[0]]))
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
