Received: by eff.org id AA15057
  (5.65c/IDA-1.4.4 for pub-infra-exploder@eff.org);
 Fri, 13 Dec 1991 15:17:38 -0500
Reply-To: pub-infra
Precedence: bulk
To: pub-infra
Date: Fri, 13 Dec 1991 15:17:35 -0500
Message-Id: <199112132017.AA15051@eff.org>
From: <frankston!Bob_Frankston@world.std.com> (by way of mkapor@eff.org (Mitch
 Kapor))
Subject: Re: What is ISDN Good For?


It seems that the main attribute of interest in ISDN is the (relatively)
high 
bandwidth.  Actually, this it is modest compared with other media (such as 
cable TV and newspapers), but still better than standard phone lines.

But there are other aspects that I'm more concerned about including:

* Ideally, ISDN should provide peer protocols between my premises equipment

and the central offices in terms of controlling connections and accessing 
network resources.  I should also be able to provide services.  A simple 
example is call forwarding -- I should be able to implement this locally.  
One advantage of a local implementation is the ability to add policies.  
These protocols should support queries about pricing and other service 
aspects.

* For data transmission I need to be able to connect to multiple services 
using either a datagram capability or multiple virtual circuits.  This
should 
be a standard offering so that I can build software that takes advantage of

it.

* Addressing (phone number) issues should be cleaned up over the current 
system so that I have a uniform way of addressing resources on the network.
 
I.e., no dial "9" for an outside line.  But this should be extended to 
subaddressing as in specifying an extension number, a person or service.  
This is similar to DID but universal without the high costs of a DID trunk.

This is extending the view of what ISDN is.  For example, I should be able
to 
use standard ISDN protocols to pass the additional data bits to a pager. 
(Of 
course, the character set should be Unicode and not just digits).

* Of course, pricing should be such that casual connection to network 
services for many hours should be comparable to the price for an unmeasured

service line.  The current message unit pricing assumes that the use of the

phone is an unusual event.  If it is our connection to the communications 
infrastructure then the default is to be connected and use of a modest size

information pipe should be negligible  Perhaps a better price comparison 
would be to the newspaper.  Getting the newspaper delivered via ISDN should

not cost more than getting some ink laden woodpulp hand carried to my
house.

In reading the notes about the physical wiring for ISDN, I'm concerned
about 
ISDN being an MIS manager dream and an end user's Procrustean bed.