1996-05-18 - Re: “Too cheap to meter”

Header Data

From: jim bell <jimbell@pacifier.com>
To: cypherpunks@toad.com
Message Hash: 343911eabda97de0b03dfca19d88d1a25e97178dc9cfe2d33c8180836e7a4baa
Message ID: <199605180723.AAA25334@newmail.pacifier.com>
Reply To: N/A
UTC Datetime: 1996-05-18 11:20:36 UTC
Raw Date: Sat, 18 May 1996 19:20:36 +0800

Raw message

From: jim bell <jimbell@pacifier.com>
Date: Sat, 18 May 1996 19:20:36 +0800
To: cypherpunks@toad.com
Subject: Re: "Too cheap to meter"
Message-ID: <199605180723.AAA25334@newmail.pacifier.com>
MIME-Version: 1.0
Content-Type: text/plain


At 06:54 PM 5/17/96 -0700, Timothy C. May wrote:
>At 2:38 AM 5/16/96, Alan Horowitz wrote:
>>Hey, let's build faster and faster fiber-optic networks. Let's create
>>bandwidth so cheap that it won't even pay to meter it.
> "Too cheap to meter" goes away pretty quickly.

Don't be so sure about that, Tim.  While it is probably true that service 
must be rationed, one way to do that is simply to charge based on the data 
rate of your modem.  When, soon, fiber normally runs at that 20 Gb/sec rate 
of this new AT+T fiber, that represents about 700,000 connections at 
28.8kbps, solid.  Based on normal statistical useage, it's probably closer 
to 3-4 million connections.  And that's only one fiber.  

Data-transmission companies need to make money, but they don't necessarily 
have to make that money by measuring actual transmitted data.  This is 
similar to cable-TV companies who (with the exception of pay-per-view) don't 
charge based on how long you watch TV.  The reason, obviously, is  that it 
is no more expensive for them  if you watch your TV 24 hours per day, than 5 
minutes per day.  Since fiber optic systems don't wear out with usage, and 
their capacities are exceedingly large, it would certainly be possible for 
these companies to start charging based entirely on maximum transfer speed.

 




Jim Bell
jimbell@pacifier.com





Thread