The Internet is expensive

Robert Robbins rrobbins at GDB.ORG
Mon Sep 20 13:42:15 CDT 1993

On Mon, 20 Sep 1993, Una Smith wrote:

> I agree.  Let's make the most of the TAXACOM archive on James Beach's
> <beach at> anonymous FTP/gopher server on

The TAXACOM archive contains copies of messages already sent via TAXACOM.
Using it, except to avoid resending entire messages already distributed
via TAXACOM is unrelated to the size of TAXACOM messages.  Another option
would be to figure out a way that long messages could be archived without
sending them, and then sending just a short announcement calling attention
to the existence of the longer note (much as Crust suggested initially).

> Robert, you haven't understood Ricardo's problem.  You are fortunate to
> live in a high-tech part of the world where competition among service
> providers has allowed most (not all!) universities to negotiate flat-rate
> fees for Internet connections.  But they still pay a lot for the service.

Universities pay a lot more for telephone service.  Granted, internet
access is not free, but neither is it a back-breaking burden for US
institutions.  I wasn't claiming a moral judgement about what is better.
Crust asked if others have experienced his problem and I was answering in
an effort to explain why for most workers in the US the answer will be no.

> You mention universities charging researchers for access to the Internet.
> At most US universities, every student and researcher pays indirectly for
> this access, whether or not they use it.  This subsidizes the costs for
> those researchers who do use it, and encourages use.  At others, research-
> ers are charged flat-rate fees for the *luxury* of having full Internet
> access in their own offices, to subsidize the cost to the university of
> installing campus-wide networks.

For most universities, maintaining the campus networking infrastructure is
several orders of magnitude more expensive than acquiring and maintaining
the internet connection.  At some sites the set-up cost for acquiring an
internet connect might be a few tens of thousands of dollars and the
recurring annual cost about the same.  At JHU a recent estimate placed
upgrading the network infrastructure on campus somewhere in the $1-5
million range, with recurring annual costs in the vicinity of $1 million

> But Ricardo Crust said his university does not charge him for his e-mail.
> As is the case in most less-developed countries, the only available access
> to the Internet is via a service provider that insists on charging per
> byte transmitted.

Here you use the word transmitted, but the discussion so far has implied a
fee for bytes transmitted AND for bytes received.  The telephone company
chatges for long-distance calls placed.  Incoming calls are free to the
recipient.  In telecommunications, the general rule on pricing is either
that people pay for services that they initiate or they pay a flat rate
(like for cable television).

> > The logic for not charging on incoming is just to avoid the kinds of
> > problems described in the posting -- the totally unpredictable nature
> > of incoming volume and the inherent unfairness in charging for it.
> There is nothing illogical here, this is simple economic reality.

I disagree.  There is a fundamental illogicality here.  There is also a
departure from standard economic practice.  I pay to send mail, not to
receive it.  I pay to place telephone calls, not to receive them.  I pay
to send FAXes, not to have them arrive on my machine.  Being required to
pay for incoming communication, when the source is outside your control,
is simply fundamentally illogical.  Occasionally email systems go berserk
and send out hundreds or thousands of duplicate email messages in an
endless loop.  Being in Crust's situation and being on the wrong end of
such an accident could break the bank.  Losing one's shirt because someone
accidentally sends you a lot of mail just doesn't make sense and strikes
me as fundamentally unfair.

Now, there is one kind of telecommunication service for which one does pay
both sending and receiving -- mobile telephone service.  For that reason,
it is a good idea to keep your car phone number unlisted.  Once it is
listed, however, it is a bit of closing the gate after the horse is gone
to worry too much about the volume of incoming messages.

I am very sympathetic to Crust's predicament.  I, too, would be very
unhappy if I had a lot of junk mail arriving postage due.  I just am not
certain what the remedy might be.  His posting, my response, you response
to my response, and now this message are all costing him money.  The best
solution would be for Crust's institution to attempt to renegotiate the
service on a combination flat rate, plus charges for outgoing packets,

More information about the Taxacom mailing list