[2426] in Commercialization & Privatization of the Internet
re: Billing Systems
daemon@ATHENA.MIT.EDU (Craig Partridge)
Wed Feb 26 21:37:29 1992
To: Jack Haverty <jhaverty@us.oracle.com>
Cc: drw@bourbaki.mit.edu, com-priv@psi.com
Reply-To: Craig Partridge <craig@aland.bbn.com>
From: Craig Partridge <craig@aland.bbn.com>
Date: Wed, 26 Feb 92 18:33:25 -0800
> You have to consider cause and effect here. If it's too expensive to use the
> technology for non-constant situations, then of course its reasonable that
> measurements would indicate that current usage patterns are constant. We have
> many sites (countries) that do not have full-time access to our internet,
> because the costs (i.e., prices) of doing so right now simply are not justified
> compared to other approaches such as UUCP/dialup connectivity.
Jack:
I'm not sure I understand the gist of your statement here.
Are you saying that bandwidth is so expensive that people only buy
what they need and that's why a line speed is a useful measure? My
experience is that this is not true -- line speed is usually a tradeoff
between what is desired (i.e. peak bandwidth that would be useful) and
one's wallet and tends to cause folks to buy bandwidth well in
in excess of the average bandwidth they use.
Or are you pointing out that while billing for fixed bandwidth access
is nifty, there will continue to be a market of folks for whom fixed
bandwidth access it too expensive and how do we bill them? If this
is the issue, I'll observe that the phone company has a mechanism to
bill them for you (900 numbers). Further, having watched CSNET bill
customers on the per-phone call basis I can assure you that you don't
want to do that if you can avoid it -- very expensive process.
Craig
