[2151] in Commercialization & Privatization of the Internet

home help back first fref pref prev next nref lref last post

Re: low-end users

daemon@ATHENA.MIT.EDU (Dave Crocker)
Tue Jan 28 17:48:52 1992

To: Richard Mandelbaum <rma@tsar.cc.rochester.edu>
Cc: bjp@sura.net, com-priv@psi.com, rtennant@library.berkeley.edu
In-Reply-To: Your message of Tue, 28 Jan 92 14:45:25 -0500.
Date: Tue, 28 Jan 92 14:47:41 -0800
From: Dave Crocker <dcrocker@Mordor.Stanford.EDU>

Richard,

While it certainly is necessary to make sure that the discussion on
low-end economics starts with a reasonable economic analysis of base
costs, such as properly calculating the fully loaded cost of a full-time
person, I hope that there can be some movement towards consideration of
the real limitations to achieving reduced costs.

Of course, one line of thought is that it can't be achieved.  But rather
than pack up and go home, I'd like to suggest looking for other
views.

We pay modest rates for telephone service.  Even if you look at
commercial phone service, rather than home (since home service tends to
be subsidized by commercial) then we still see reasonable rates.  And the
service is not constrained to single-function (equivalent to limiting
yourself to email, rather than full IP.)  Better still, the quality of
customer support tends to be quite high.

Well, one point is that they DO have service limitation, in that you
only get local calls for the flat fee, but it is worth noting that the
current IP providers don't have to pay packet charges to other packet
carriers (true?) so that some of the logic dictating a similar
local/long split would seem not to apply.

My feeling is that the cause of the current cost problem, due to 
the cost of support, is that too much support is needed.  That is, we
haven't done anything approaching an adequate job of making hosts
and nets plug-and-play.  And I don't see anything coming out of the
IETF to fix that anytime soon.  Yet I keep thinking that this is a
problem -- well, a set of problems -- that CAN be solved.

Seems like there is an opportunity, here, though not an easy one.

Dave

home help back first fref pref prev next nref lref last post