[1341] in Commercialization & Privatization of the Internet
Re: What is the Definition of Infrastructure?
daemon@ATHENA.MIT.EDU (James B. Van Bokkelen)
Mon Sep 16 13:57:03 1991
Date: Mon, 16 Sep 91 13:57:53 -0400
To: Ittai Hershman <ittai@shemesh.ans.net>
From: jbvb@ftp.com (James B. Van Bokkelen)
Reply-To: jbvb@ftp.com
Cc: com-priv@psi.com
Size is irrelevant with regard to educational connections. The base
price includes all educational institutions, including K-12.
True enough, but only for educational customers. "For each additional
research or commercial connection, there will be a charge per attachment
based on the connection bandwidth" (from ANS Gateway Pricing Details and
Example). It appears that this charge is based on the bandwidth of the
connection from the regional to the non-educational site.
First, ANS wants to compete with the regionals to provide end-user
connections.
Wrong. Since ANS has both direct attachments and gateway attachments,
we naturally compete with the mid-levels. The paper is the result of
a collaboration among ANS and the mid-levels to create an environment
in which both mid-levels and ANS could co-exist and thrive.
The stated pricing policy is size-neutral only to mid-levels that aren't
interested in handling commercial traffic. How much it actually affects
commercially-interested regionals depends on the "charge per attachment".
Second, ANS wants to avoid regional-regional interconnects
that bypass ANS.
Wrong. The COMBit mechanism allows us not to care about "backdoor"
connections between mid-levels. We only care about, and recover costs
for, traffic which crosses the ANS backbone
The COMBit percentage is just a premium on top of an Attachment Price
that is already linear with the number and bandwidth of non-educational
leaf nets. If it was 0%, an example mixed CO+RE regional with 25 T1
connections of each kind would pay $155,000 (presumably per year?) for
a 56Kb ANS gateway. One with 250 of each would pay $1,055,000.
ANS can *say* many things, but what it *means* is most likely to be
evident in its published pricing. I can't say how deliberate it was,
or how much of ANS or NSF or FARNet is aware of it, but when you offer
financial incentives (or disincentives), the world tends to move to
conform to them.
If ANS is in sole control of the Unit Prices in Attachment B, then ANS
controls whether it is economical for any given commercially-interested
regional to connect, expand, upgrade existing connections and install
bypasses. I can only guess how $2000 compares with the profit that a
PSI or a CERFNet expects from a 56Kb leaf connection, but I expect it's
a significant chunk.
James B. VanBokkelen 26 Princess St., Wakefield, MA 01880
FTP Software Inc. voice: (617) 246-0900 fax: (617) 246-0901
