[1237] in Commercialization & Privatization of the Internet
Re: impact of settlements on provision of free services
daemon@ATHENA.MIT.EDU (Erik E. Fair" (Your Friendly Postm)
Wed Aug 28 06:11:16 1991
From: "Erik E. Fair" (Your Friendly Postmaster) <fair@apple.com>
In-Reply-To: <m0kFKGx-000HeuC@heifetz.msen.com>
To: com-priv@uu.psi.com
Date: Wed, 28 Aug 91 03:10:59 -0700
I, as Postmaster and Internet Liaison for Apple Computer, will put it
in even more strong terms than Ed Vielmetti: we will not connect to a
usage priced TCP/IP network.
Every usage priced network I have ever seen has priced their services
so much higher than equivalent leased lines that reasonable budgeting
is impossible (i.e. doing the maximum is way too much money, and
guesstimating usage of such a service is a good way to get a mid-year
budget surprise). We still connect to one usage priced network, under
duress: SPRINT X.25. Budgeting for it involves a guess, and we *still*
get surprised, even after six years of experience with it.
It is also my belief that if the Internet goes usage-priced, you can
say goodbye to most experimentation with new network protocols and
services - what responsible administrator would allow casual
experimentation, without some estimate of the costs that will be
incurred by the experimentation? Beancounters are a notoriously
short-sighted lot, and their default answer is usually, "no."
We are now on CERFNET for two reasons: we believe that redundancy is a
good thing - when our primary regional network (BARRNET) is having
troubles (rare these days, but still...), we're still up; also because
CERFNET is a member of the CIX, we are on the right side of the rules
when sending commercial bits to other CIX interconnected sites (once
they set up their routing, that is). It is my expectation that many
other commercial sites will also choose a similar model of Internet
connectivity.
However, we will be watching this "settlements" business carefully.
Erik E. Fair apple!fair fair@apple.com
