Technology and Trends
 USENET Archives
  
Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utcsrgv.UUCP
Path: utzoo!utcsrgv!perelgut
From: perel...@utcsrgv.UUCP (Stephen Perelgut)
Newsgroups: net.micro,net.research,net.cse
Subject: First Summary of PC's in Education Survey
Message-ID: <3459@utcsrgv.UUCP>
Date: Tue, 6-Mar-84 13:57:17 EST
Article-I.D.: utcsrgv.3459
Posted: Tue Mar  6 13:57:17 1984
Date-Received: Tue, 6-Mar-84 14:09:59 EST
Organization: CSRG, University of Toronto
Lines: 95

(..)
 \/

Here are the preliminary results of my survey about personal computers in
a formal education environment.  With the possible exception of Dartmouth,
nobody seems to really know what to do about student purchased pc's in the
classroom situation.  

I am presenting the results.  A follow-up questionnaire will be posted
separately.
----------- Summary (it really is a summary, just not concise) -----------
University of Toronto
    - We are currently looking into teaching introductory computing
      on a pc based system (8086-family, MS-DOS).  The language will
      be Turing (developed at the U. of T. and accepted for all intro.
      computing courses here.)
    - Initial plans are to develop a lab of 15 pc's and use it to teach a
      course.  Other problems being investigated are networking and file
      servers.
    - The University is trying to work out a deal for making computers 
      cheaper for students to purchase.  A minimal machine will be
      designated and a good price for a particular version will be
      negotiated with the manufacturer/dealer.
Rose-Hullman @ Indiana
    - Classrooms, study areas, dorms, etc. all have computer ports
    - Students required to purchase a small, portable PC with AC/battery
      pack that can be tied into a large computer network.

North Carolina State
    - Introductory courses now taught on micro-computers using the
      Sage IV running a multi-user P-System.  Each Sage IV has 5 (?)
      terminal attached.  Other courses use an IBM 4341, Data General
      MV-8000, and a lab full of PDP-11's
    - Students can purchase a single-user Sage II running the UCSD
      P-System.
    - A Sage is based on the 68000 processor.  The Sage II has 2 
      floppies (DSDD), 512K (?), 2 RS-232 ports, a printer port, and
      an IEEE-422 (?) port.  The Sage IV has 1 floppy, 1 hard disk,
      1M RAM, 6 serial ports, 1 parallel port, and an IEEE-422 (?).
University of Waterloo
    - Many introductory courses use pc's in networks.  There are
      several Janet networks of IBM PC's and several Waterloo
      MicroNets using IBM Series 1's, PDP-11's, and big IBM's as
      central file servers.  Languages include Pascal, Basic, Cobol,
      and Assembler.
    - Students cannot use their pc's at home and cannot get copies of
      Watcom software.  Network pc's have no floppies, only shared
      file servers.
Looking Glass Software
    - currently designing a programming environment based on 
      syntax-direct editors to run on the ICON (Bionic Beaver) pc.
    - These will be purchased by high-schools in Ontario and are 
      not likely for students initially.
    - First target is a Pascal environment.
MIT
    - uses HP machines based on 68000's with 4M main memory for 
      introductory programming.
    - Project Athena at MIT will use lots of DEC Pro-350's and some
      IBM PC's, all donated.
University of Saskatchewan
    - currently looking for 50 networkable micros to replace the 11/70
      being used for RJE.  
    - Will probably standardize on MS-DOS and are looking into Modula-II
Simon Fraser University
    - May be doing same things as Sask.
Harvard 
    - Plans to use Macintoshes to teach introductory courses next semester.
    - Will be developing software over the summer.
    - Developing introductory Pascal and more advanced assembler courses.
      Pascal environment may eventually use three windows on the Mac -- 
      one for source, one for output, and one for condition of data
      structures.
    - Currently using IBM PC's in an extension course teaching Pascal.
      Ok experience but a pain with printers.
UCSD
    - has been using micros since 1978.  First used Teraks but they were
      too expensive for students ($8K).  Course switched to Apple ]['s and
      IBM PC's.
    - project involves groups (2-3) of students so usefulness of personally
      owned PC's is debatable.
Dartmouth
    - Any student can buy a PC or get financial aid to help buy the PC.
    - DCS assuming all students will have easy access to a PC.  
    - Introductory: will use True Basic on a Mac
    - Followup: will use Pascal on Mac
    - Assembly: will probably use Mac (68000 is a nice assembly lang.)
University of Alberta
    - Getting a lab of PC's but aren't quite sure what to do about it.
New Mexico State University
    - uses Terak (11/03) to teach UCSD P-system Pascal for intro. computing.
-- 
Stephen Perelgut   
	    Computer Systems Research Group    University of Toronto
	    Usenet:	{linus, ihnp4, allegra, decvax, floyd}!utcsrgv!perelgut
	    CSNET:	perelgut@Toronto

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: he...@utzoo.UUCP (Henry Spencer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3604@utzoo.UUCP>
Date: Tue, 6-Mar-84 19:18:28 EST
Article-I.D.: utzoo.3604
Posted: Tue Mar  6 19:18:28 1984
Date-Received: Tue, 6-Mar-84 19:18:28 EST
References: <3459@utcsrgv.UUCP>
Organization: U of Toronto Zoology
Lines: 22

There was an interesting CS seminar today at U of T, a fellow from
Brown University describing their "electronic classroom" work with
some references to future plans.  He made one interesting point,
which I sincerely hope the planners at U of T (and elsewhere) will
pay attention to (ARE YOU LISTENING, OVER THERE IN MCLENNAN LABS AND
SANDFORD FLEMING??):  don't expect miracles if your idea of a "personal
workstation" is an IBM PC.  He cited a "7 M's" definition of a reasonable
workstation, attributed to Bill Joy:

	Megabyte of memory
	Megapixel screen (800x1024 is close enough)
	MIP (million instructions per second) processor (68000 is ok)
	Mouse
	Megabyte/s network
	Menu-based interface
	Memory (large contiguous unsegmented virtual)

Note that a garden-variety IBM PC under MSDOS has none of these.
Brown uses Apollo workstations.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utcsrgv.UUCP
Path: utzoo!utcsrgv!perelgut
From: perel...@utcsrgv.UUCP (Stephen Perelgut)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3466@utcsrgv.UUCP>
Date: Tue, 6-Mar-84 22:17:37 EST
Article-I.D.: utcsrgv.3466
Posted: Tue Mar  6 22:17:37 1984
Date-Received: Tue, 6-Mar-84 23:24:06 EST
References: <3604@utzoo.UUCP>
Organization: CSRG, University of Toronto
Lines: 21

In defence of IBM-PC's (and their compatible buddies):
    - The 80186 is arguably as good as a 68000 and anyone with talent
      can create software to run on the 80186 and 8088 and all the other
      family members
    - 8088 based micros are the most popular.  It isn't fair to students
      to force them to buy another machine.  This may change with the
      Macintosh-madness.  (If the happens, the CSRG is capable of making
      the switch with little pain.)
    - Since Canada has no history of corporate support of universities,
      the choice of micro is typically dependant on who makes the best
      offer and what is available "north of the border."
    
Aside from this, th "7 M's" are true for Brown's "animated algorithms"
scheme, but they aren't essential for the simple task of introductory
computer programming courses.  An IBM PC (or Tandy 2000 or DEC Rainbow or ...)
is suitable, cheaper, available, and currently popular.
-- 
Stephen Perelgut   
	    Computer Systems Research Group    University of Toronto
	    Usenet:	{linus, ihnp4, allegra, decvax, floyd}!utcsrgv!perelgut
	    CSNET:	perelgut@Toronto

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: he...@utzoo.UUCP (Henry Spencer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3611@utzoo.UUCP>
Date: Wed, 7-Mar-84 17:51:59 EST
Article-I.D.: utzoo.3611
Posted: Wed Mar  7 17:51:59 1984
Date-Received: Wed, 7-Mar-84 17:51:59 EST
References: <3604@utzoo.UUCP>, <3466@utcsrgv.UUCP>
Organization: U of Toronto Zoology
Lines: 47

Steve Perelgut observes, in defence of the IBM PC and clones:

    - The 80186 is arguably as good as a 68000 and anyone with talent
      can create software to run on the 80186 and 8088 and all the other
      family members

I think Mark Mendell answered this one well.  In no way is a collection
of 16-bit address spaces as good as a single big contiguous one.  Sure,
there are many things that will run quite well in 16 bits, but when you
run out of address space, you're really stuck.  I can create software
that will run on a PDP-8 -- that doesn't mean that it's a good idea or
a worthwhile use of my time.

    - 8088 based micros are the most popular.  It isn't fair to students
      to force them to buy another machine.  This may change with the
      Macintosh-madness.  (If the happens, the CSRG is capable of making
      the switch with little pain.)

By this argument, you ought to be teaching the students Microsoft BASIC
as their first language.  Universities should be buying the equipment
that will be right for tomorrow, not yesterday.  As for making the switch
with little pain:  after you've bought a huge pile of 8088 machines?  No
way that investment is going to be written off without pain, screaming,
madness, committee inquiries, etc. etc.  When this much money is involved,
there is a very high premium on getting it right the first time.  Buying
a 16-bit-address-space machine is clearly a major mistake.

   Aside from this, th "7 M's" are true for Brown's "animated algorithms"
   scheme, but they aren't essential for the simple task of introductory
   computer programming courses.

Introductory computer programming courses are among the things Brown uses
their fancy lab for.  It is quite possible that an 8088 is adequate for
the introductory courses DCS teaches today, taught the way they are
taught today.  The availability of the new technology is likely to change
the way courses are taught.  It is a mistake to view the introduction of
per-student workstations as just new hardware for the same old purposes.
Maybe DCS is not interested in things like "animated algorithms" now;
in a few years they will be.  But by then it will be too late, if the
short-sighted advocates of the cheapest hardware available have their way
right now.  U of T will once again find itself stuck with antiquated
equipment and a second-rate teaching environment while everyone else
forges ahead.  How many years will it take to fix it THIS TIME?  Much
better to think ahead a bit now, and invest a bit of money in the future.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: he...@utzoo.UUCP (Henry Spencer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3612@utzoo.UUCP>
Date: Wed, 7-Mar-84 18:01:14 EST
Article-I.D.: utzoo.3612
Posted: Wed Mar  7 18:01:14 1984
Date-Received: Wed, 7-Mar-84 18:01:14 EST
References: <3604@utzoo.UUCP>, <3466@utcsrgv.UUCP>, <3611@utzoo.UUCP>
Organization: U of Toronto Zoology
Lines: 10

It just occurred to me that outsiders may not be aware of just what I
was referring to in my previous article, when I referred to U of T getting
stuck with a poor computing environment "again" and asked how long it
would take to get it fixed "this time".

Three years ago, U of T's computing environment for computer science
teaching was punchcards.  Only quite recently did this change.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utcsrgv.UUCP
Path: utzoo!utcsrgv!perelgut
From: perel...@utcsrgv.UUCP (Stephen Perelgut)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3475@utcsrgv.UUCP>
Date: Wed, 7-Mar-84 21:32:31 EST
Article-I.D.: utcsrgv.3475
Posted: Wed Mar  7 21:32:31 1984
Date-Received: Wed, 7-Mar-84 22:17:17 EST
References: <3604@utzoo.UUCP>, <3466@utcsrgv.UUCP>, <3611@utzoo.UUCP>
Organization: CSRG, University of Toronto
Lines: 43

.
I don't want too many people to think I am a promoter of outdated
technology or something.  So let me say a few more things in my own
defence.

1) I advocate 8088 based technology because it is there.  The reason I
    don't advocate MS-BASIC is that we can do better now.  Following Henry's
    argument, I would design my teaching environment based on a 16032 with
    all 7 M's.  However:    it isn't available in any quantity
			    it isn't tested (and no university is going to buy
				a pig in a poke no matter how pretty it is.)
			    it would cost both arms, one leg, and the other leg
				up to the second joint
2) I advocate IBM PC's because they are here, they have upper management
    support, and we have one to play with NOW.  Not next week or next year
    or whenever.
3) I realize that 16-bits isn't enough to write the worlds largest programs,
    but I seriously doubt any student will suffer from the "limitation" in a
    first computer course.
4) "Animated algorithms" may be the wave of the future, but there is no 
    support here, now.  Would anyone recommend that we wait until the future is
    here?  I would rather do something than sit and wait for something better.
    There is always something better on the horizon.  After the 68000's are the
    16000's.  And then the next generations.  Etc.  I don't wanna wait.  I want
    to start conversion now.
5) The U. of T. is grossly underfunded.  When there is a commitment to pursue
    some project, turning it down is suicide.  If we look too far into the
    future, we will miss the present boat.

In spite of all that, I sympathize with Mark Mendell (who got stuck with the
8086 coder when I became too busy.  Just goes to show you what happens when
you're the best, and Mark is!  So he got stuck!)  I would love to use 68000's,
but I don't have one, let alone the 15-20 we need.  Anyone wanna give a gift
to a worthy university.  We even have means for giving out U.S. tax
certificates.  And we implement top-quality compilers in under a man-year.
(3 months for a production Turing compiler developed by 3 people, about
2 man-months for Turing/68000 when we go that route.)  And Turing is
complete, verifiable, and FAST (compile-time AND run-time.)
-- 
Stephen Perelgut   
	    Computer Systems Research Group    University of Toronto
	    Usenet:	{linus, ihnp4, allegra, decvax, floyd}!utcsrgv!perelgut
	    CSNET:	perelgut@Toronto

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: he...@utzoo.UUCP (Henry Spencer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3619@utzoo.UUCP>
Date: Sat, 10-Mar-84 20:37:57 EST
Article-I.D.: utzoo.3619
Posted: Sat Mar 10 20:37:57 1984
Date-Received: Sat, 10-Mar-84 20:37:57 EST
References: <3604@utzoo.UUCP>, <3466@utcsrgv.UUCP>, <3611@utzoo.UUCP>, 
<3475@utcsrgv.UUCP>
Organization: U of Toronto Zoology
Lines: 105


Steve Perelgut counters with:

   ...I advocate 8088 based technology because it is there.  The reason I
    don't advocate MS-BASIC is that we can do better now....

And we can most assuredly do better than the 8088.  Even stipulating that
we couldn't, this would NOT be an argument for buying 8088's now; it would
be an argument for postponing purchases until we could get equipment with
reasonable characteristics and a reasonable useful life.

    ..........I would design my teaching environment based on a 16032 with
    all 7 M's.  However:    it isn't available in any quantity
			    it isn't tested (and no university is going to buy
				a pig in a poke no matter how pretty it is.)
			    it would cost both arms, one leg, and the other leg
				up to the second joint

If you are willing to settle for a 68000 instead of a 16032 (I know, this
is a poor deal, but it's not too visible from high-level languages), it is
available, in quantity, tested, right now.  True, it's expensive.  The
price can come down a lot if you're willing to settle for something that
doesn't meet the 7M specs but nevertheless comes lots closer.  Also, the
prices keep dropping.  I would not call the Macintosh "both arms, one leg,
and the other leg up to the second joint"; more like half an arm.  Also
remember:  "there ain't no such thing as a free lunch".  If we want a
decent teaching environment, as opposed to a half-obsolete one, it may
cost some money.  Punchcards are real cheap, but we appear to have finally
decided that they are no bargain.  The general lesson does not seem to
have sunk in, though.

   ...I advocate IBM PC's because they are here, they have upper management
    support, and we have one to play with NOW.  Not next week or next year
    or whenever.

They indeed are here now.  So are better things.  I know people who have
Macintoshes to play with right now -- Toronto-area people with no special
"in" at Apple.  True, it took a bit of effort; they think it was worth it.
As for "upper management support"...  [long string of expletives deleted].
If upper management is making a colossal mistake, it is your *DUTY* to
tell them so, attempt to convince them, and resign if it doesn't work.
I am prepared to do exactly this if *my* management makes a sufficiently
large mistake; fortunately, this does not seem likely.

   ...I realize that 16-bits isn't enough to write the worlds largest programs,
    but I seriously doubt any student will suffer from the "limitation" in a
    first computer course.

Of course the students aren't likely to notice.  It will be the instructors
and their programmers who will notice, as soon as the teaching environment
starts to shift towards really *using* the new technology.  Incidentally,
I haven't heard any provision for giving the upper-year students anything
better.  You know as well as I do that if 8088s are the official U of T
teaching environment, they'll be stuck with them too.  (Barring the few
lucky courses that have access to something better.)

   ..."Animated algorithms" may be the wave of the future, but there is no 
    support here, now.  Would anyone recommend that we wait until the future is
    here?  ...

No, we're supposed to be trying to MAKE the future happen, not just sitting
waiting for it to hit us in the eye.  Which it will if we buy 8088s.  It is
the duty of people who are making decisions to try to avoid doing things
in ways that are short-term wins but long-term disasters.  Nobody ever said
that it would be easy.

   I would rather do something than sit and wait for something better.

So we are supposed to sit with our 8088s and wait for something better
to magically appear?

    There is always something better on the horizon.  After the 68000's are the
    16000's.  And then the next generations.  Etc.  I don't wanna wait.  I want
    to start conversion now.

Of course.  But it is nevertheless possible to pick equipment that is not
going to be cripplingly obsolete in a few years, especially when we are
talking about something that may well represent a long-term commitment to a
specific teaching machine.

   The U. of T. is grossly underfunded.  When there is a commitment to pursue
    some project, turning it down is suicide.  If we look too far into the
    future, we will miss the present boat.

Believe me, I know about U of T's funding.  I refer you back to my previous
comments about trying to make things happen rather than just accepting a
rotten situation, and about the possibility of waiting in order to get
decent equipment.  Even if it meant getting only half as many (say)
Macintoshes (an assumption I think implausible -- fewer, maybe, but not
that big a difference), that would still be a better investment.  Buying
by the lowest bidder makes sense only if the products offered are roughly
equivalent in functionality.  Not so here.

Sorry if some of the above seems harsh.  I really do think that an 8088-
based teaching machine would be a disastrous mistake, and that it could
be avoided.  We've had the @&$&%est time trying to get rid of punchcards
here, and I'd prefer not to see a repeat.

P.S. to all the other readers:  I know this is starting to seem like a
personal debate between Steve and myself, but I think the issues are of
sufficiently wide interest to justify continuing to use "f" rather than
"r" for my rebuttals.  Let me know if you disagree.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utcsrgv.UUCP
Path: utzoo!utcsrgv!perelgut
From: perel...@utcsrgv.UUCP (Stephen Perelgut)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3501@utcsrgv.UUCP>
Date: Sun, 11-Mar-84 18:26:49 EST
Article-I.D.: utcsrgv.3501
Posted: Sun Mar 11 18:26:49 1984
Date-Received: Sun, 11-Mar-84 19:21:02 EST
References: <3619@utzoo.UUCP>
Organization: CSRG, University of Toronto
Lines: 36

[Rebuttal**n]

Let me make things clear and concise.
1) I don't love 8088's but they aren't as bad as Henry seems to think.  If
all courses had to use IBM PC's, they would be almost as good
2) The situation at the U. of T. is so bad that the students cannot complete
assignments.  1/2 as many Mac's would be WORSE (100% certainty).  Offloading
the intro courses to PC's would allow full use of the COMPUTING power of the
IBM's and Vaxen for courses that need the power.
3) The situation requires action.  8088's are ideal for the current needs and
will be used.  68000/16032 systems are the unquestionned future but we need
them now.  

Summary
	- support is there
	- machines are available in quantity
	- software is being prepared by the best in the business.  (Let's seee
	    anyone else produce a compiler for a complex, verifiable language
	    with simplified syntax in only 3 months!  And the code (excluding
	    run-time checking) is as good as any other VAX compiler with very,
	    very few exceptions)
	- students own THESE machines, almost none (none?) own promises of
	    Mac's.
	- the machines are cheap and upgradable.

Conclusion
	We aren't stupid, we're just making the best of the situation.  Talk
	to me again next year and we'll be up on 68000's, 16000's, and almost
	anything else there is a crying need for.  But 8088's are the 
	(qualified) BEST choice NOW!  And if you don't agree, try teaching a
	course on one of the currently overloaded machines!
-- 
Stephen Perelgut   
	    Computer Systems Research Group    University of Toronto
	    Usenet:	{linus, ihnp4, allegra, decvax, floyd}!utcsrgv!perelgut
	    CSNET:	perelgut@Toronto

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site utcsstat.UUCP
Path: utzoo!utcsstat!geoff
From: ge...@utcsstat.UUCP (Geoffrey Collyer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <1765@utcsstat.UUCP>
Date: Tue, 13-Mar-84 00:53:29 EST
Article-I.D.: utcsstat.1765
Posted: Tue Mar 13 00:53:29 1984
Date-Received: Tue, 13-Mar-84 01:43:11 EST
References: <3501@utcsrgv.UUCP>
Organization: U. of Toronto, Canada
Lines: 15

Steve Perelut has recently made several references to overloaded U of T
teaching machines.  Just so the general audience understands, the VAX
780 which is taking 2/3 of the undergraduate ``Computing Disciplines''
students on the main campus is not overloaded, it is *broken*.  We are
having trouble getting it fixed because, as usual, DEC diagnostics have
run without failure.  At this moment, Ian Darwin and a DEC repairman
are trying to get the machine to fail DEC diagnostics.  One symptom of
the failure could perhaps be confused with heavy load because the
machine just stops running for minutes at a time and only runs for a
few minutes at a time.

This may seem like a local U of T argument, but the mainframe situation
is not as bad as it has looked for the last week or so.

Geoff Collyer, U. of Toronto Computing Services

Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: he...@utzoo.UUCP (Henry Spencer)
Newsgroups: net.micro,net.research,net.cse
Subject: Re: First Summary of PC's in Education Survey
Message-ID: <3638@utzoo.UUCP>
Date: Tue, 13-Mar-84 19:58:23 EST
Article-I.D.: utzoo.3638
Posted: Tue Mar 13 19:58:23 1984
Date-Received: Tue, 13-Mar-84 19:58:23 EST
References: <3619@utzoo.UUCP>, <3501@utcsrgv.UUCP>
Organization: U of Toronto Zoology
Lines: 76

Well, here's my rebuttal to Steve's rebuttal...

   1) I don't love 8088's but they aren't as bad as Henry seems to think.  If
   all courses had to use IBM PC's, they would be almost as good

Senior courses using non-trivial software packages will be badly hurt
by the 8088's awful 16-bittisms and its dismal speed.  I don't agree
that they are suitable for all courses; they are only marginally suitable
for the lowest-level courses.

   2) The situation at the U. of T. is so bad that the students cannot complete
   assignments.  1/2 as many Mac's would be WORSE (100% certainty).  Offloading
   the intro courses to PC's would allow full use of the COMPUTING power of the
   IBM's and Vaxen for courses that need the power.

   3) The situation requires action.  8088's are ideal for the current needs and
   will be used.  68000/16032 systems are the unquestionned future but we need
   them now.  

What does "now" mean?  If we are talking about major offloading of the
central machines, my impression is that this *isn't* going to happen
overnight.  If nothing else, the administration is going to have to be
talked into buying *lots* of personal machines, and this administration
has a history of grossly underestimating the amount of interactive access
needed for intro teaching.  (My impression is that this is as big a problem
as overloading, although admittedly the delays due to overloading worsen
the finding-a-free-terminal problem.)

Given that there is going to be a non-trivial delay before personal machines
have a large impact on local computing, it might well make sense to opt for
a better personal machine.  Macintoshes are available *now* -- they are the
present, not the future.  By the time we have enough personal machines
to have a large offloading effect, it may be very hard to change machines
because of the immense inertia.

Which leads to:  just how wide is the conviction that "68000/16032 systems
are the unquestioned future"?  Once upper management at places like UTCS
gets its collective mind locked into "IBM PC", or some such similar track,
it's going to take a long time to get it unlocked.  If we cannot get
decent machines right away, it is *vital* that there be an immediate and
strong commitment to upgrading as soon as possible.  Otherwise we will
still be buying PCs in 1999.  Do not forget what a powerful effect the
IBM marketing hype has on "professional managers".

	- students own THESE machines, almost none (none?) own promises of
	    Mac's.

So we're going to encourage students to buy a machine that we plan to
abandon in a couple of years anyway?  This sounds like a recipe for being
stuck with the PC for the indefinite future to me.  And if the machines
owned by students now are considered a major factor, what about those
poor souls who aren't rich enough to own one?  Do they get screwed?  If
not, why are the existing student-owned machines so important?

	We aren't stupid, we're just making the best of the situation.  Talk
	to me again next year and we'll be up on 68000's, 16000's, and almost
	anything else there is a crying need for.  But 8088's are the 
	(qualified) BEST choice NOW!

Note what I said above about the time required to have a major impact.
By the time we can get enough 8088s to help, they will be obsolete.

	And if you don't agree, try teaching a
	course on one of the currently overloaded machines!

No thanks, I do have some idea what it's like.  I also have some glimpse
of what it'll be like teaching 8088-based courses 15 years from now.
I'm glad I'm not teaching at all.

I have continued to post to the world, rather than follow Brian's
suggestion of more localized discussion, because I continue to think
that much of this stuff is of wide interest.  The mail I've received on
the matter has agreed with me.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,linus,decvax}!utzoo!henry