3.723 NeXT, UNIX, and the promised future (256)

Willard McCarty (MCCARTY@vm.epas.utoronto.ca)
Thu, 9 Nov 89 19:05:48 EST

Humanist Discussion Group, Vol. 3, No. 723. Thursday, 9 Nov 1989.

(1) Date: Thu, 9 Nov 89 12:52:46 GMT-0800 (143 lines)
From: "Oliver G. Berghof" <oberghof@next.acs.UCI.EDU>
Subject: NeXT

(2) Date: Wed, 8 Nov 89 11:44:23 EST (93 lines)
From: Jeffrey Perry <JEFF@PUCC>
Subject: (1) NeXt query; (2) UNIXEXPO '89

(1) --------------------------------------------------------------------
Date: Thu, 9 Nov 89 12:52:46 GMT-0800
From: "Oliver G. Berghof" <oberghof@next.acs.UCI.EDU>
Subject: NeXT

Lest it should seem that I have abandoned the discussion I started I
would like to respond to some of the contributors.

1. James H. Coombs, JAZBO@BROWNVM

This was the posting that promted my response. I am very concerned that
HUMANISTs who read Jim's impressionistic comments will be dissuaded from
trying out the NeXT themselves and forming their own opinions. I am
very grateful for this contribution for it provides a perfect example
for the kind of underinformed criticism that will keep HUMANISTs shying
away from the NeXT machine.

I take issue with all of Dr. Coombs' criticisms:

1. It's very slow. Compared to a Mac II running A/UX and a Max
(sic) toolbox application ... Much of the time it seems unusable ...
the Frame application. Too slow for me.

I have a Mac IIci standing three feet away from me. For computationally
intensive tasks the NeXT leaves the Mac limping in the dark.

If you're really interested in getting a notion of the speed of a NeXT,
try the Mandelbrot - demo in the NeXT-Developer library. It compares
compilation in C with compilation using the Digital Signal Processor and
displays the runtime result in two adjoining windows.

Wordprocessing is not likely to use the Digital Signal Processor, but
then you don't buy the NeXT to type away at your christmas card in
WordPerfect. FrameMaker takes a long time loading. The newly released
version 1.0 should amend this. But then try loading any desktop -
publishing package on an 80386 - you will be grateful that there is such
a thing as the NeXT !

2. Feedback is poor. Some messages are inappropriate, such as
something about "file system error" when one modifies a document but
does not have the proper permissions . Even worse, one often does
not know whether the system is working on something or not...

Compared to the nonexistent feedback on a Mac I find the NeXT's error
messages a godsend. Jim Coombs' remark gives the argument nicely away:
had the NeXT not included the explanation that he did not have the
proper permission to modify a document he would not even have known what
to complain about ... ! As far as the distinction between active and
inactive applications is concerned: each of the application icons on the
right hand side of the screen (in the "workbench") contains three little
dots when it is inactive. The only way for you not to know on a NeXT
what applications are active is to actually HIDE them with the HIDE
option which is located right beyond the QUIT option on the menu. But
of course, if you're a Mac mouse freak you are likely to miss one button
for the other.

3. The mouse is infuriating. The acceleration when one
moves it quickly is insignificant ... There is also a > problem with
spurious clicks... ... I guess the biggest need is for faster hardware
to drive everything they are trying to do.

Try clicking on the icon in the upper right hand corner. It is there
for the sole purpose of customizing your machine - and that includes the
mouse ! As far as the spurious clicks and the faster hardware are
concerned I would hazard a guess that in this case the problem was
neither the software, nor the hardware ....

- - - - - - - - - - - - - - - -

2. David Carpenter, ST_JOSEPH@HVRFORD Malcolm Brown,

I am very much indebted to Malcom Brown for his elaborate account of his
experiences with the NeXT. Both he and David Carpenter seem to express
limited enthusiasm for the software available although they deplore the
scarcity of what is on the market.

I fully agree with the descriptions both have given of various NeXT
programs such as WriteNow, the Digital Librarian and FrameMaker.
However, I feel that the real issue, given the visionary concept of the
hardware, is not so much what there is already as what there might be in
the future. When I used the phrase "a fair amount of tools for
HUMANISTs" I referred specifically to Object-C and Allegro Common-Lisp.

Although HUMANISTs with David Megginson's expertise in Unix are a
minority and although our power, as a network of brains is more akin to
that of ants than that of eleph-ants, this does not have to remain the
state of affairs. The insects were barred from an expansion of their
evolutionary domain by chitin ("a horny polysaccharide that forms part
of the outer integument esp. of insects and crustaceans" - for those who
don't have a DigitalWebster) -do HUMANISTs have to be barred from
evolving any further by lack of funds and initiative ???

The NeXT is currently priced at $ 6500.- for institutions in higher
education. Considering the current prices for optical storage and the
estimated value of the software included in this price the NeXT machine
is probably the most underpriced piece of computing equipment currently
on the market. This being so I was a little depressed to hear from
Malcolm Brown that his initiative to get support for the development of
a text analysis tool met with little encouragement from NeXT. Maybe
this attitude will change only if HUMANIST's become visible to NeXT as a
group of software developers seriously to be reckoned with. Before that
happens, though, we will have to know more about Unix, Object-C, Lisp,

I suggest that HUMANISTs interested in becoming third party developers
write directly to NeXT (if Malcolm Brown is willing to share some e-mail
addresses). I will also continue to clamour for support from them , and
try to share whatever information is available.

To end opimistically, with one of those intolerable "wouldn't it be
nice"-phantasies, does not seem to be appropriate, but let me have a try:

Imagine you're a classicist who has to walk over to the library to get a
print-out from the Thesaurus Linguae Graecae on an Ibycus system that is
incapable of anything but text-retrieval and boolean string-searches.
Now imagine the same classicist sitting in her snug little home with a
NeXT machine in front of her, flipping the CD-ROM with Herodotus, and
Thucydides, and Galen, and whatnot into her cube, happily incorporating
quotations into the paper that she is writing. To say nothing of
text-analysis, immediate communication with other scholars or customized
programs... .

_ _ _ _ _ _ _ _ _

I wrote the above comments on Nov. 4th, before I received David
Carpenter's and Malcolm Brown's second comments. Apologies for

Oliver Berghof
Department of English and Comparative Literature
University of California, Irvine

(2) --------------------------------------------------------------101---
Date: Wed, 8 Nov 89 11:44:23 EST
From: Jeffrey Perry <JEFF@PUCC>
Subject: (1) NeXt query; (2) UNIXEXPO '89

(1) NeXt Software availability query
Steve Jobs of NeXt recently spoke here. He handed out a very
slick catalogue of software and peripherals said to be available
for the NeXt. I was quite excited to see products like Framemaker 2.0
listed, but at the bottom of the page describing Framemaker it said
"Availability: Contact Manufacturer." Does this mean that Frame
Technology Corp. is committed to producing a version of Framemaker
for the NeXt sometime soon, or does it mean that Framemaker 2.0 is
available right now to NeXt users? Come to think of it, most of the
software listed in the NeXt catalogue included the notation "Avail-
ability: Contact Manufacturer". I hope this doesn't mean that third-
party software support for the NeXt is still mostly in the future.
(2) UNIXEXPO '89
On Thursday Nov. 2 I attended UNIXEXPO '89 in New York. The sheer number
of vendors and volume of information was overwhelming; a keynote address
by Bill Joy of Sun Microsystems provided some background and perspective
that allowed me to keep from feeling totally at sea.
I'll begin by summarizing Joy's talk, since that was definitely the
most important part of the day for me. Joy's main message was that
powerful, cheap, easy-to-use Unix machines will be ubiquitous very soon,
and that they will create a new computer-use paradigm, to wit networks
of desktop machines connected to file servers, each work station able
as a matter of course to communicate locally to all of the others in
the network as well as to distant networks and servers. Interactive
tasks will be handled by the desktop work stations individually, more
computation-intensive tasks being farmed out to one's friendly neighbor-
hood server(s). Joy predicted that computers will be 'mostly invisible'
in the 90's, and said that the 'frontier' of computing has already
shifted from the realm of hardware and operating systems to the realm of
applications, and to what he called the 'sociological' aspects of
computing. This point of view was a welcome corrective to the gadget-
mongering then taking place in the exhibitors' displays. Joy said that
with the shifting of the frontier, 'trade shows like this one will
become rather silly' since Unix computing will be so ubiquitous and
easy to do that a show centered around Unix will make as little sense as
a trade show centered around electricity.
Joy predicted that with the arrival of a standardized version of Unix
(release V.4), the DOS world will soon be absorbed into the Unix world.
He predicted a somewhat more viable future for the Macintosh, but
predicted that eventually the latter will be doomed as well by its
inability to do what all Unix machines soon will be able to do, i.e.
talk to any other Unix machines, participate in the new distributed
computing paradigm, and run any applications that any other Unix
machine can run.
It was a well-focused talk that made me feel willing to swear fealty to
Unix on the spot. Indeed, there are now Unix-based WYSIWYG word
processors /document creation packages, like Framemaker 2.0, that seem
to do everything any of the snazziest Macintosh products can do vis a vis
presentation; there's a version of the vi editor that lets you enter
text (and address the operating system) in Arabic and other languages as
well as in English, and of course databases, graphics software, and
other packages that should be quite well suited to doing whatever the
computing academic needs to do.
And yet...
The total absence of vendors offering educational applications at the
show made it hard for me to get any potentially useful information from
anyone about any topic whatsoever. I would ask a vendor about (for
example) large textual databases, or non-roman alphabets, and would
get a momentarily glazed stare folowed by PrePackaged Spiel #34 on How
We Can Modernize Your Company's Payroll Accounting. Fine, so it wasn't
an educational show; still, with the coming of The New Order, I have
some fears that this sort of thing did little to allay.
First fear: who is going to develop the educational software? Will
academia (and anyone not part of the corporate mainstream) become a
stepchild of Unix? Sure, things will be so easy and standardized that
if this happens schools and scholars will probably be able to catch
up on their own with less trouble than was the case when previous
'universal standards' were introduced. It might've helped if NeXt had
come to the show, but they didn't.
Second fear: what happens to the mountains of scholarship and data
ammassed and manipulated in the pre-Unix Dark Ages? Are any vendors
or developers interested in at least some kind of conversion/emulation
for IBM Mainframe, DOS and Apple software? Will I be able to tell
Nota Bene users, or Hypercard users, or SPIRES users not to worry,
that they'll be able to get to their documents after the Revolution?
Since this sort of Ellis Island for old software, old documents and
old data isn't likely to be a big moneymaker, I very much fear that a)
the old DOS machines, mainframe applications, etc. will be with us for
a long time, making headaches for everyone, and b) this will retard
progress towards the 'invisible' computing that Bill Joy promises for
the '90s.
I'd love to see a Unix version of Apple's Macademia, which I'm afraid
would inevitably be called - you guessed it - 'Unixversity', where such
concerns could be addressed. All in all, I am mostly sold on Joy's view
of the future, despite the lack of a non-corporate perspective at the
show. Until someone shows me otherwise, I will have to assume that this
reflects a similar lack in the Unix software development world as a