Re: VCC cost models .... (was Re: Limits

[Curtis Villamizar <[e-mail]>]

Steve,

In message <[e-mail]>, "Silverman, Steve" writes:
> 
> The point of the original message was that you can't view Internet
> links as "free" while assuming a charge for ATM links.  This makes
> makes the use of ATM uncompetitive.  The subsidies are disappearing
> (or are gone).  We need to create a system that optimizes in some
> sense the service the users get versus the total costs which the
> user's will have to pay in one way or another.  I think (with no real
> proof) that ATM is a component in this "solution" but an analysis to
> prove this (if possible) must be based on what will be not what was.

People are not saying Internet links are free, just substantially more
cost effective.  That holds true today for commercial use of the
Internet that no longer rely on any government subsidies when compared
to the cost of switched services using SVCs, particularly when making
lots of low bandwidth connections.

> I agree that these subsidies (of the Internet) created an industry and
> were a very good investment, far better than most of the other federal
> government's expenditures.
>  
> In the early '80s, I was able to identify several hundred million
> dollars/year in govt. subsidies of the Internet.  Most of what I saw
> was funneled thru DOD but I am confident I only saw a piece of it.
>  
> Steve

A lot depends on what you call a subsidy.

You could argue that the DoD by using the Intrernet is still
subsidizing it, but then you'd have to claim they are subsidising the
telephone companies because they use and pay for telephone service.

Today the NSF provides funding to colleges to give those colleges
access to the Internet.  It also funds state level programs to build
regional networks where there are none, though this too is going away.

The Internet was research in the early 1980s.  The telcos were well
funded and the capital investments made by telcos to upgrade their
equipment over the same decade completely dwarfs and funding provided
to the Internet.

Its a moot point anyway.  The issue is really which will be more cost
effective in the future and whether how well IP forwarding scales
relative to switching becomes an overriding reason to go with
switching or whether VC technical cost issues or VC pricing issues
dominate the decision.

It could go either way and we are probably wasting our time giving
opinions on which way it will go.

Curtis