a 2nd potential solution
On Wednesday, July 16, 1997 12:28 PM, Andrew Partan[SMTP:asp at partan.com] wrote:
@ > Here's a copy. Please note that the existence of this work does NOT
@ > constitute unequivocal endorsement of the concept. There are significant
@ > business issues which are intrinsic to the model which would have to be
@ > resolved on a case by case basis. The technical footing is sound.
@ Please note that in order for Coalition or Geographic based address
@ allocation to work really well, you have to have restrictions on
@ your topology.
For many small ISPs, "geo-netric" issues are resolved very close to home.
Their main problem is that they get tied to an upstream provider
and therefore their customers get tied. Significant changes in the
relationship with that upstream provider now impact their customers.
In the Coalition approach, ISPs have to understand they will be
tied to the coalition. This seems like less of a problem for ISPs than
being tied to an upstream provider.
Of course, depending on the stability and greed factor of coalitions,
the ISPs may wish they were tied to an upstream provider. This is
one of the major concerns about organizations such as ARIN.
Supposedly, ARIN is being "given" some unknown /8 delegations
to manage. The companies that have allocations in those blocks
have not been given any say about this. What happens if ARIN
decides to enact policies and change companies allocations or
charge high fees ? Small ISPs will probably have as much say as
they did with the $50 domain taxes and the future of .COM, .NET
@ Addressing works really well when you do it on a topological basis.
@ Anyone who does not follow this needs to educate themselves in the
@ basic physics of routing and how it applies to CIDR based allocations
@ and keeping routing tables small.
Many people feel that the routing tables can be reduced
in size with better (and more fair) policies. Without experiments
to prove and document this, people will never know.