From: shamrock@netcom.com (Lucky Green)
Date: Tue, 13 Aug 1996 12:43:47 +0800
To: Brian D Williams <cypherpunks@toad.com
Subject: Re: FCC_ups
Message-ID: <v02120d03ae356356fa46@[192.0.2.1]>
MIME-Version: 1.0
Content-Type: text/plain


At 11:36 8/12/96, Brian D Williams wrote:

> First I would like to mention Lucky that I always enjoy your
>posts. Second I would like to say that as I have mentioned
>previously, the RBOC I work for filed for complete unbundling of
>the local loop in March of 93. We are not the one's holding up the
>show.

Thanks for kind words. I am happy to hear that your RBOC is so enlightened.
I did not mean to exclusively blame the Baby Bells. There is plenty of
blame to go around. Look at the FCC's incomprehensible decision to require
that an ISDN line be billed as two lines, arguing that it requires double
the upstream bandwidth.

I know that I am repeating myself, but here is a brief primer on telco
realities:

There is plenty of upstream bandwidth, since everything but the local loop
is already running on fiber. The problem is in the local loop. In many
cities, there is simply no room left in the ducts to run even one more
wire. Anything, such as ISDN, that can get more lines out of the same
number of wires is a Good Thing to the telcos and the consumers. The
alternative is trenching. An often prohibitively expensive proposal,
especially in a downtown area. The FCC should encourage, not discourage the
use of such wire saving technology.

>10 Mb/s ethernet and 100 ISDN B channels (64k each)? I would
>certainly like to here more! Unfortunately as I have also
>previously pointed out, point-to-point copper is a thing of the
>past, it is rare and expensive now. The current fiber-to-the-curb
>standard involves "slick 96" muxes which use 4 framed T-1's (1.536
>Mb/s) to provide 96 voice channels.

Given the cost of running fiber to the home, the near and intermediate
future is definitely fiber to the curb. The technology we were developing
works beautifully with this set-up. To be more precise, the system provides
a 10Mbps Ethernet, 96 64bps ISDN B channels, a D and an M channel if using
iso-Ethernet based technology. But the key to getting all this bandwidth
isn't iso-Ethernet, it was our chip. This can be done at ~$600/home and for
a low as $100 per node. Yes, this will increase upstream bandwidth
requirements, but as I mentioned earlier, there is plenty of upstream
bandwidth. Last office I worked in was facing an alley full of loading
docks and trash binns. Six feet away from the door ran MFS's OC-48 SONET.

There are of course exceptions. If you live in a very old building, you
might have to rewire. My current apartment uses three wire cabling. I have
no idea what the third wire is good for. You can't run two POTS over three
wires. In a building like this you'd have to rewire. But few office
buildings are likely to have such lousy wiring.

Anyway, I hate to be a tease, but that's about as far as my NDA will let me go.

[Any further dialog in private email, please. Time to take this tread off-list.]



-- Lucky Green <mailto:shamrock@netcom.com> PGP encrypted mail preferred.
   Defeat the Demopublican Unity Party. Vote no on Clinton/Dole in November.
   Vote Harry Browne for President.