Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!bellcore!faline!ulysses!sfmag!sfsup!glg
From: glg@sfsup.UUCP
Newsgroups: comp.society.futures
Subject: Re: Bioproduced nanocomputers
Message-ID: <2375@sfsup.UUCP>
Date: Tue, 17-Nov-87 18:48:25 EST
Article-I.D.: sfsup.2375
Posted: Tue Nov 17 18:48:25 1987
Date-Received: Sat, 21-Nov-87 07:46:34 EST
References: <8710290154.AA09880@agent99.wedge.com> <21526@ucbvax.BERKELEY.EDU> <18947@yale-celray.yale.UUCP>
Reply-To: glg@/guest4/glgUUCP (xmpj20000-G.Gleason)
Organization: AT&T Information Systems
Lines: 70

In article <18947@yale-celray.yale.UUCP> robertj@yale.UUCP writes:
>In article <21526@ucbvax.BERKELEY.EDU> oster@dewey.soe.berkeley.edu.UUCP (David Phillip Oster) writes:
>>Imagine a system as easy to use and as complete as the average university
>>library, but where any document can be annottated by anyone at anytime.
>>(And any reader can choose to turn off any subset of annotations he does
>...
>>Conclusion:
>>How can we evolve usenet news into such a system?
>>(There Nick, I've tied nanotechnology to Stuart II.)

>In Stewart Brand's book about MIT's Media Lab, there is a passage
>in which the director of the Lab is telling an audience that AT&T
>currently has the capability to put fiber-optic cabling into every
>home in America.  "What would be the bandwidth of such a system?"
>he is asked.  "500 gigabits per second," he calmly replies.

It would be interesting if this were to happen, and it would have
implications far beyond hypertext.  It is not as far-out as it sounds,
I can easily imagine stringing fiber could be as cheap as the coax for
cable (is it already?), and although 500Gbps is not current technology,
it is likely to be sooner than you think.  The major stumbling block 
that I see is that AT&T (and the rest) would want to make too much money
(think of what this kind of bandwidth would do to voice telephone rates)
on it, so it would take a long time to be widespread.

>This is the only way that a hypertext authoring system such as the
>one above could become real.  (The supercomputers to hold all the
>works themselves don't exist yet, but that's not my point.)  Nearly
>universal access to this hyperlibrary would be essential to a full
>implementation of this hyper-authoring scheme.

I don't think real time is all that important.  Sure, it is frustrating
to wait, and there are major benifits if following links can be done
in real time, but think about how much better it would be that trying
to hunt down references in a library.  Many people do not even have
access to a large enough library to do any real research.  Even if it
takes overnight to download a reference to your local system, I think
it is worthwhile.

Also, I can think of other ways to get a global hypertext system linked
into your home with existing technology.  The upload link does not have
to be that fast, since an individual contributer can only type so fast,
phone lines and modems can handle this.  Downloading of large files
could be done over broadcast media such as cable-TV channels.  Your
computer would dial in a request with your modem, and the file would
be captured and loaded into the computer from a box hooked on to your
cable-TV.  News type stuff could be broadcast continually, so it could
be waiting for you (if you told your computer to capture it that is).

Yes, current machines are not quite up to the most ambitious hypertext
projects, but they are more than enough to make a very useful system.
Besides, by the time a system is designed and built, the machines will
probably be ready.  In _EoC_ Drexler talks about "designing ahead",
that is, doing the groundwork in anticipation of future technology.
Laserdisks already have enough capacity for storage of large archives,
communications bandwidth is more of a problem, but even for this, 
technical solutions exist, we only have to wait for economic feasibility.

The is a danger of forming standards too early, before enough is known
about the problems, but lack of standards slows things down, and isolates
comunities of users within their local standard.  Enough is probably
known now to set standards for storage and transmission of text and
links, and for the interactions between archive-servers and terminal
machines.  The existence of these standards would clear the way to
set up some initial servers, to prompt development of user interfaces
for terminal machines, and to start dealing with issues like copywrite
protection and reimbursment of authors (probably the most significant
issues to encourage publication of quality material in this medium).

Gerry Gleason