[2177] in Commercialization & Privatization of the Internet
Re: Low end users
daemon@ATHENA.MIT.EDU (Dave Crocker)
Wed Jan 29 11:30:14 1992
To: Richard Mandelbaum <rma@tsar.cc.rochester.edu>
Cc: perry@mcl.unisys.com (Dennis Perry), HORN%athena@leia.polaroid.com,
In-Reply-To: Your message of Wed, 29 Jan 92 10:48:24 -0500.
Date: Wed, 29 Jan 92 08:21:39 -0800
From: Dave Crocker <dcrocker@Mordor.Stanford.EDU>
Why worry about support costs? Simple. If low-end service is to be
offered and give legitimate service, then we need to look for the
cost factors in current operating styles and either change the style(s)
fundamentally or find a way to optimize them. (I view this stage of
the effort to be little more than equivalent to finding out where the
bottlenecks are in a piece of software. That is, you can't optimize
everywhere, so look for the biggest payoff.)
In services industries, support is well-known to be a very, very major
slice of the pie. Hence, let's look for ways to reduce it. This will,
I believe, lead to 2 major lines of work (no. no drum roll, at this
stage of discussion. Wait until we have results from the lines of
work):
1. Enhancements to TCP/IP specs to allow some degree of automatic
configuration, definitely for first-time users and possibly for fully
mobile users; and
2. Vastly easier installation/configuration procedures for users.
The phone company, which I cited as a demonstration of low-price
service, unfortunately solves both of these issues by having a massive
relatively centralized support structure and hides most of the ugliness
from users. We really don't have the option, since the end-user will be
the one to install the software. Worse, the nature of the service is
far more complicated than is offered by the phone service. (Well, it
seems that way to me, but it would be interesting to have a well-
motivated skeptic review the assertion -- IS internet access inherently
more complicated than POTS? In particular, what lessons can we learn
from them?)
Dave
