3.737 NeXT, Ibycus, UNIX, and the future (287)

Willard McCarty (MCCARTY@vm.epas.utoronto.ca)
Sun, 12 Nov 89 17:24:02 EST

Humanist Discussion Group, Vol. 3, No. 737. Sunday, 12 Nov 1989.

(1) Date: Fri, 10 Nov 89 22:39:39 EST (21 lines)
From: unhd!psc90!jdg@uunet.UU.NET (Dr. Joel Goldfield)
Subject: "Humor & common sense in Vol 3, No. 729"

(2) Date: Sat, 11 Nov 89 07:45:21 EST (85 lines)
From: David Megginson <MEGGIN@vm.epas.utoronto.ca>
Subject: Revelations

(3) Date: Sat, 11 Nov 89 12:41 EST (64 lines)
Subject: Ibycus

(4) Date: Sat, 11 Nov 89 15:56 CST (87 lines)
From: John Baima <D024JKB@UTARLG>
Subject: NeXT and UNIX

(1) --------------------------------------------------------------------
Date: Fri, 10 Nov 89 22:39:39 EST
From: unhd!psc90!jdg@uunet.UU.NET (Dr. Joel Goldfield)
Subject: "Humor & common sense in Vol 3, No. 729"

Ah, the computer quarrels are heating up, again, eh? I enjoy the give
and take here, despite the concomitant "MY machine's better than YOUR
machine," and especially enjoy learning something about most of these
micros I've never been able to try for more than a few minutes. Dr. Jim,
thanks for the most hilarious Humanist message I've read all semester!

Now, will somebody summarize all of these astute remarks, good common
sense of Eli, et al., for those of us who are still in the dark ("glowing"
or not) about what the latest incarnations of Mac's, NeXT's, etc.,
*REALLY* do well and what we should be aware of in problematic features
(or lack of them)? Yes, I've read the _BYTE_ reviews, but I find the
specific Humanist comments much more interesting and to the "humanist point."

Joel D. Goldfield
Plymouth State College (NH)
(2) --------------------------------------------------------------89----
Date: Sat, 11 Nov 89 07:45:21 EST
From: David Megginson <MEGGIN@vm.epas.utoronto.ca>
Subject: Revelations

Here are some of my ideas about where computing will be in 5-10 years.

1) MSDOS will die completely. When Visicalc came out, the business
community moved as a body to computers. When MSDOS came out, and they
were all using CP/M, they moved as a body to MSDOS. In the next few
years, they will move as a body to Unix for the sake of decent
multi-tasking and inter-process communication. We will remember OS/2
only as a brief abhorration. As CP/M users discovered, the fact that
_everyone_ is using your operating system does not guarantee that
operating system's future.

2) The Macintosh, the Amiga and the ST will fade into insignificance.
Musicians will continue to use the ST for its MIDI capabilities
and some video enthusiasts will hang on to their Amigas. Basically,
though, these machines will linger on only because their
manufacturers are including Unix with the high-end versions
(Atari TT, Mac II, Amiga 2000).

3) There will be a single, common Unix graphic interface, like the one
on the NeXT, only more advanced. Although most people will own Unix,
they will have no idea how it works or even that it is there. They
will delete files by dragging them to the trash, not by typing "rm."
However, for hackers or other line interface enthusiasts, sh, csh
etc. will be running underneath the graphic environment, and they
will be able to do all the command-line typing they want.

4) We will have a completely different idea of how computer programs
should work. Each program will execute one task well, and it will
call other programs for other specialised tasks. Your DTP application
will call your resident dictionary to check hyphenation, open up
your resident editor to allow you to edit its text, and call your
resident image processor to crop pictures. As a result, you will
be able to customise your system, using the same text editor for a
your database, your text-retrieval system, and your mail reader.
Your natural-language parser will use the same dictionary/ies as
your spell-checker.

5) With new, high-speed communications lines, this idea of inter-process
communications will extend to other computers. Your natural language
parser may choose to call your own dictionary for semantic information,
or it may choose to call a dictionary in Tokyo. From your point of
view, there will be no difference (except in the dictionary itself).
If you want files from another machine, you will be able to open a
directory window and drag the files to your machine with the mouse,
the same way you would copy a file from one directory to another with
the Mac.

6) Screen resolution will increase by several magnitudes, until we will
no longer strain our eyes reading 10pt or even 8pt text on the computer
screen. Screens will also grow (a 12" monitor is too small for

7) With Unix and a common graphic interface, _all_ computers will be source-
compatible. That means that once a program exists for one machine, it
will take seconds to reproduce it for any other, even though the machine
code is different. Every program will be available for every computer.

8) Keyboards will begin to disappear, as voice-readers become cheap, accurate
and readily available. Keyboards may be an oddity in ten years. Perhaps
the mouse, too, will disappear, and computers will track the user's eye

9) Text scanners will work at at least 1200 dpi and will have an effective
0% error rate, barring mechanical failure, on typed or printed text.
Scanners which read handwritten or calligraphic material will operate
with as low an error rate as possible given the clarity of the original

10) Computers will read and "understand" books, and will be able to acquire
basic knowledge from them.

11) Everyone will have at least a gigabyte of disk storage, probably many.

I am philologist, not a computer scientist, so beware of citing me as an
authority. I hope that someone will keep this and show it to me again in
five years.

David Megginson, Centre for Medieval Studies, Toronto
(3) --------------------------------------------------------------71----
Date: Sat, 11 Nov 89 12:41 EST
Subject: Ibycus

I neither own nor have regular access to a NeXT machine, but,
because I am fortunate enough to have an Ibycus and a Macintosh
IIx, I can compare their relative usefulness in my work. With
apologies to R. Kraft, then, I must admit that the Ibycus remains
for me chiefly a tool for searching the TLG, while everything else
(wp, db, telecom) is done on the Macintosh. Although it is doubtless
POSSIBLE to do much more with the Ibycus, the ways in which
these things may be done are either a good deal more difficult (less
"user-friendly," if you will) or not always made clear to the
end-user -- this end-user, at any rate, who belongs to the
IBYCUS-L and will confess, in the current sardonic style, to being
a relatively stupid bozo. Thus, whereas it is probably possible, as
Kraft mentions, to access e-mail through the Ibycus -- indeed, I
have heard of the existence, somewhere, of VAX terminal emulation
software -- neither this software nor the details of cabling a
modem to the machine is readily obtainable by me.
This admission of ignorance and/or incompetence may elicit
clucks from the cognoscenti. I will risk them. The potentials which
I have left unrealized are left so, I imagine, in consequence of the
fact that I am not particularly well-connected, not in the right
place at the right time. While several Ibycus users have generously
created and distributed valuable utilities for that machine, none of
these ameliorates -- or can, I dare to venture, hope to ameliorate
-- significantly the "spartan" quality of the interface. The
Macintosh is so much easier for this bozo to use and to manipulate
(with ResEdit and/or HyperTalk) that there is no serious incentive
to develop facility with Ibyx, the very fine but rather specialized
programming language of the Ibycus. The lack of incentive, for me
(and perhaps only me), is enhanced by the fact that the inevitable
problems which the very mediocre DOS or Macintosh hacker
encounters may be resolved in many "etheric" places, the former
group having a vast number, the latter at least an impressive
number of fora. Ibycus users, however, have IBYCUS-L, and,
essentially, nothing more. I find this situation daunting; however, I
bought the Ibycus in full awareness that there would be a small
amount of long-distance support. The machine is very reliable and
well made. Yet, I cannot let pass the suggestion that, upon having
obtained one, the user can then kick back and roar with a wry grin
into the 21st century. This, in my experience, is a significant
exaggeration of what most users will encounter.
Let no one think, however, that the Ibycus is not an excellent tool
-- still the very best there is (so far as I, the you-know-what,
know) -- to use with the TLG disk. I am very grateful to have been
lucky enough to obtain one. However, having had the opportunity to
use Pandora (in a beta version) on the Macintosh, I cannot help but
look forward to the day when I will finally be able to integrate the
TLG disk into a single work station.
A NeXT machine capable of being a real scholarly work station is
still yet to arrive, if I may judge from the brief encounter I had
with one a few months ago. It was, however, under system .8 or .9,
immediately easy to navigate through much of the interface without
any help from a manual or salesbeing. That certainly could not be
said of the Ibycus, whose true virtues, in contrast to the NeXT (and
the Macintosh), are a very reasonable price and rapid searching of
the TLG and CCAT disks.

W. McCarthy
Wash., D.C.

(4) --------------------------------------------------------------91----
Date: Sat, 11 Nov 89 15:56 CST
From: John Baima <D024JKB@UTARLG>
Subject: NeXT and UNIX

I was quite interested by Oliver Berghof's comments about NeXT, and as
a software developer, I would like to respond to some of his comments.

He states, "I suggest that HUMANISTs interested in becoming third party
developers write directly to NeXT". Well, I did just that this spring.
And the program I had in mind was his "dream" for NeXT (TLG access),
although good access exists via Ibycus, Pandora on the Mac and Lbase on
IBM-PC's (my program). Anyway, last Spring developers had to submit
applications and wait in line for both a machine and "boot camp".
Developers could buy the machine at the university price only if they
went to a several day long "camp" (at the developers expense) to teach
them about the NeXT machine. The review of NeXT that I read was from a
developer and he thought that the large hard disk was necessary for
development work. When you start adding all that up, the NeXT is not a
bargain ($15k) to develop software for, especially low volume, academic
software (and for people who demand that their software be
inexpensive). So, I was discouraged by all that and I will not be
developing any software on NeXT in the future unless one floats down
from heaven and lands on my desk.

One of the really nice things about developing software on the NeXT is
the Objective-C (not Object-C, please) environment. I think that
Objective-C is the best hybrid, object oriented language available
today. However, NeXT's Objective-C compiler will soon be available in
source code form from the Free Software Foundation! It seems that NeXT
linked in some of the GNU-C compiler to make the Objective-C compiler
and thus are bound by the FSF's "copyleft" agreement which requires the
distribution of source code. This point has been *extensively*
discussed on gnu.misc.discuss. Thus, I hope to see Objective-C
compilers for UNIX machines from the FSF in the coming months.

UNIX, of course, is another story. I use UNIX quite a bit, mostly to
run FrameMaker, mail and to make Smalltalk. I like what UNIX can do,
but I do not like UNIX *at all*. I doubt that anyone who really likes
to be a UNIX sysop on a network has time to do anything else :-). It
can be a bear. Yes, I know, NeXT will provide UNIX for the rest of us.
I just don't believe it.

While Bill Joy is one of my heroes because he thinks that C is a large
mistake for large systems or applications, he would never make it as a
prophet. UNIX will not replace DOS or OS/2. There are over 30 million
DOS machines out there and 20 million more could be shipped next year.
UNIX will increase, but the UNIX market is in total disarray. System
V.4 may help, and OSF may merge back to provide a unified UNIX
community, but that is not enough. Business is behind DOS and OS/2 all
the way and that will not change, for better or worse. Market share is
not based on technical excellence alone and Bill Joy (one of the
founders of Sun Microsystems) is not exactly a neutral observer anyway.

As for vi and the rest of the UNIX tools being able to do 15 years ago
what is now available in today's desktop publishing programs, I like
Bill Joy's comment about his program, "If I had know that vi would
become so popular, I would never have written it." vi??? Gack!!!

As for NeXT and Humanists, I don't expect to see NeXT make a serious
effort to capture humanist's hearts and pocketbooks. Given the software
currently available for NeXT, the educational market is the only one
they could sell to. Oh, yes, many businesses will buy a NeXT to play
with it, but not to solve any problems. Not today, at least. If NeXT
wants to survive, it must get into the business market and then
academics will fade away. Did you know that Apple sold fewer Macs
to higher education in 1988 than they did in 1987, although the total
number of units sold has soared?

What's the point? I doubt that any of the hardware vendors are going to
try very hard for the HUMANIST market. I would love to develop software
for NeXT, except for the fact that it is expensive and the market is
tiny. Remember, us independent software developers have to pay for our
machines and our time (have any of the NeXT people out there bought the
machine out of their own pocket?). Either schools will sponsor software
(like U Toronto) or HUMANISTs will have to pay for it. Statements like
"this machine is great, why don't some of you software developers get
us some software for Humanities" is not very helpful. I also agree with
Elli Mylonas that it is unreasonable to think that humanists who do not
have an extensive background in programming will become effective
programmers. Objective-C is great, but it takes the average
*programmer* 2 months to become proficient enough with Objective-C to
begin designing a significant application. How many HUMANISTS can
invest that amount of time and money just to get going?

John Baima