Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!microsoft!edwardj
From: edwardj@microsoft.UUCP (Edward JUNG)
Newsgroups: comp.sys.next
Subject: Re: NeXT: Please Consider 2MB -> 20 MB floppies
Message-ID: <57380@microsoft.UUCP>
Date: 12 Sep 90 23:17:16 GMT
References: <344@kaos.MATH.UCLA.EDU>
Reply-To: edwardj@microsoft.UUCP (Edward JUNG)
Distribution: na
Organization: Microsoft Corp., Redmond WA
Lines: 56

In article <344@kaos.MATH.UCLA.EDU> barry@pico.math.ucla.edu (Barry Merriman) writes:
>The latest Byte (15th Anniversary Issue, pg 188) has a product
>review of a new type of floppy drive that is the same size
>and shape as a standard 3.5" drive, but the floppies hold
>a whopping 20MB. If NeXT has to switch to floppies, this is
>the type of drive I could live with. (You might actually
>fit TopDraw on one floppy!).

[further descriptions deleted]

>(I forget the name---something like Xor---but they're based in GA)

[stuff deleted]

The companies that actually make this kind of hardware are Brier
and Insite.  NeXT would deal with them rather than an OEM.

It is unfortunate that the industry failed to pick up and leverage
the large ODs, but it is clearly difficult to produce that kind
of magnitude of useful data for your average application, esp.
when the data is accessed with the rates, reliability, cost, and
noise of the NeXT optical drive.  Some day, not too far in the
future, we will need this because voice annotation and telephony
and sound will be routine and large full-color rendered graphics
will be common and easily-produced, but this is still not mainstream
and probably will not be for a few more years (when the technology-
cost point comes down).

From a purely marketing point of view, things would be different
if the media cost $10, even at 1/2 the capacity.  But failing that,
I wouldn't mind a CD ROM drive; I tend to produce far less information
than I consume, and publish even less of it (and when I do, it isn't
very often), so a $2 read-only medium is a reasonable compromise.

I propose that there are two access patterns that, for today and for
most people, are distinct:

  1. Frequent exchange of information
  2. Exchange of huge sets of information

For the next few years at least, the first will tend to be smaller
datasets, and the latter will be less frequent.  By virtue of the
technology and cost-effectiveness of the technology, a hybrid of cheap
large read-only transportable media with fast read-write fixed media
and slow networks can meet most of my needs.  Yes, a fast, transportable,
cheap read-write media would be better.  Yes, FDDI+ optical LAN would be
better.  But they are still expensive.  NeXT was just too far ahead of
the technology.

My 2 cents.

--
Edward Jung
Microsoft Corp.

My opinions do not reflect any policy of my employer.