[2682] in Commercialization & Privatization of the Internet
RE: Settlements
daemon@ATHENA.MIT.EDU (Donald E. Eastlake, III, LJO2/I4 +)
Fri Apr 10 16:51:45 1992
Date: Fri, 10 Apr 92 16:50:46 EDT
From: "Donald E. Eastlake, III, LJO2/I4 +1 508 486 2358 10-Apr-1992 1651" <dee@ranger.enet.dec.com>
To: drw@bourbaki.mit.edu
Cc: com-priv@psi.com
Apparently-To: com-priv@psi.com, drw@bourbaki.mit.edu
I don't know about the Internet but for POTS (plain old telephone service) the
cost of the local switching plant and the copper running to people's houses is
well over half of all plant costs. This would argue that no less than half the
costs of the Internet should be allocated on a local connection capacity basis.
Donald
--------------
>From: US1RMC::"drw@BOURBAKI.MIT.EDU" "MAIL-11 Daemon" 10-APR-1992 14:49
To: com-priv@psi.com
CC:
Subj: Settlements
From: HORN%athena@leia.polaroid.com
Why does everyone assume that packets (bytes, whatever) should be counted?
Settlements should correspond to costs, so measure and settle based on the
things that cause costs. Costs are operations labor, capital equipment,
communications line leases, etc. These are capacity related, not usage
related. Usage is just an indirect factor.
Well, for the "last mile" (the wire between your organization and the
first external router), costs are based on the capacity you have, not
the usage. But once you get to that router, the extra capacity needed
to support you is much more dependent on your average usage than your
peak usage.
Dale Worley Dept. of Math., MIT drw@math.mit.edu
--
Computers are like Old Testament gods; lots of rules and no mercy.
-- Joseph Campbell
