Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: Notesfiles $Revision: 1.7.0.10 $; site sysvis
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxn!ihnp4!inuxc!pur-ee!uiucdcs!convex!sysvis!george
From: george@sysvis
Newsgroups: net.news
Subject: Re: Orphaned Response
Message-ID: <-1771099@sysvis>
Date: Wed, 22-Jan-86 14:28:00 EST
Article-I.D.: sysvis.-1771099
Posted: Wed Jan 22 14:28:00 1986
Date-Received: Tue, 28-Jan-86 06:00:53 EST
References: <2501@amdahl.UUCP>
Lines: 20
Nf-ID: #R:amdahl.UUCP:2501:sysvis:-1771099:000:1136
Nf-From: sysvis!george    Jan 22 13:28:00 1986


Re: Message costs (vs. National Debt  :-))

The use of large aggregate numbers for ALL sites is a bit misleading for some
readers and is totally inaccurate.  It might be better to display the USENET
AVERAGE score for cost/site.  In order to make some sense of this average, a
standard deviation might be needed.  As I understand it, there is no one who
actually pays $125/message ANYWHERE.  A per-site display would be much more
useful for comparing one's own costs to the network, rather than just posting
some meaningless large numbers.  I would think that some sites with only local
feeds might be well below average in cost and vice versa for sites with only
long-distance feeds.  There are probably two populations (multiplied by large
|mini|micro computer cost).  Total per-site costs should be computed with some
reasonable allocation of the cost of disk space and CPU usage at the site.
Just quoting a telephone bill does not tell one what percentage of the money
is spent for USENET information and what is not.

Can anyone propose an alternate form of USENET cost computation that is useful?

				...!ihnp4!sys1!sysvis!george