welcome to the one size fits all Internet discussion swamp
scharf at vix.com
Sun Aug 23 06:41:58 UTC 1998
I think we're being more than a little silly here. Person A says that the
financial model for internet peering should be client funded, and B says no is
should be content provider.
folks, you can both be right for various situations. Most newspapers and
magazines are largely paid for by the advertiser. Most subsciption services
are paid for mostly by the clients. In each case the line is not simple, and
the delivery media adapts to the money flow and vice versa.
The problem comes from one of two major insanity swamps.
a) deciding that there is one model that should be used for pricing.
b) trying to decide how to classify a given flow, and the contractual concepts
involved. Maybe I'm just hardened to it all, but anything I can't describe in
a contract with terms, costs and redress is too soft for me to chage money.
That's why we have these simple minded peering agreements. Anything beyond the
mindless trading starts you onto a slippery slope from which I see no return.
Phone settlements are 100x simpler than this stuff, and you see as the cost of
haul drops this will become a major part of total cost of a phone call. I can
understand bandwidth times distance hauled in simple terms, so saying I may
not accept nearest exit routing from someone else is within my scope of simple
my last mail on this whole subject,
More information about the NANOG