COMSAT@MIT-MC@sri-unix (11/28/82)
From: Communications Satellite <COMSAT at MIT-MC>
A copy of your message is being returned, because:
"BAK-JUNK-MAIL" at MIT-MC is an unknown recipient.
Message not sent.
Failed message follows:
-------
Date: 27 Nov 1982 1235-EST
From: Mel Pleasant <WORKS at RUTGERS>
Subject: WORKS Digest V2 #89
Sender: PLEASANT at RUTGERS
To: WorkS: ;
Reply-To: WORKS at RUTGERS
Works Digest Saturday, 27 November 1982 Volume 2 : Issue 89
Today's Topics:
Queries - 68K Virtual Memory Systems,
Hardware - Terminals & Trackballs & The Corvus Concept,
Education - Utah and WICATs
----------------------------------------------------------------------
Date: Mon, 22 Nov 82 08:00:02 EST
From: barkley at NBS-SDC
Subject: 68k virtual memory systems
Does anyone have a SUN workstation from SUN microsystems? Is
there any other 68k system which will run Berkeley Unix? Any
information would be appreciated.
thanks
------------------------------
Date: 22 Nov 1982 1809-EST
From: WITTMAN at RU-GREEN at RUTGERS
Subject: Re: PR1ME computer
To: DREIFU at WHARTON-10
Henry,
Thanks for your comments. All PR1MEs terminals are "expensive"
compared to independent suppliers and they pretty readily admitted
that. I think the major problem is not the expense of the terminal
but that the OA systems seemed not to partake of PR1MEs "software
first" philosophy. There were many incarnations with many problems,
and they were external products to begin with. PR1ME builds systems
tools, NOT applications packages.
I'm unfamiliar with the capability of the 2250 (Rabbit) but know it
is more than a Workstation (and less that the traditional
time-sharing PR1ME) so I can't make many comments at this juncture.
You'll probably see it before I do; what do YOU think of it.
Regards,
Barry
------------------------------
Date: 22 Nov 82 14:33:00-PST (Mon)
From: decvax!utzoo!utcsrgv!ralph at Ucb-C70
Subject: Trackball followup
Some time ago I put a request out on this news group (I think) for
information on sources of inexpensive track balls. There were
several good suggestions, but I did seen any sources that got the
price down as low as I would think possible.
Subsequently we have found several CHEAP (!) trackballs, but some
were not real trackballs (it didn't roll). The following supplier
makes fairly good trackballs that are inexpensive. We have been
quoted just under $200 (Cdn) in Canada (i.e., after duty, currency
conversion, etc.). In the U.S. I would think <$100 in quantity one
should be expected. (In case you are wondering what "fairly good"
means, it can be translated into "good enough for us". We have
ordered one).
Disc Instruments
102 E. Baker Street
Costa Mesa CA 92626
(714)-979-5300
------------------------------
Date: 23 Nov 82 14:49:57-PST (Tue)
From: harpo!ihnp4!ixn5c!inuxc!pur-ee!CSvax.Pucc-H.Physics.hal at Ucb-C70
Subject: Corvus Concept review
I posted this review of the Corvus Concept about a month ago but
since there have been a couple of requests for information since
then and I can't get mail through to the ARPA net, here it is again.
A couple of new items are included (in particular, the predefined
nonstandard Pascal type "string").
I've had my Corvus Concept since August, so here are a few
comments. Here's a list of what's in my system:
Corvus Concept workstation (256Kb memory)
Corvus 5.7Mb hard disk
Corvus 8" disk drive (SSSD)
Okidata ML84 printer
First, the bad news:
The Corvus advertising says that the 8" diskette is 1.0Mb
(formatted). It isn't. It stores 256Kb. When I talked to the
Corvus marketing people, it was the first they had heard about it!
So much for communication between engineering and marketing. I
settled with Corvus when they agreed to supply a Mirror interface so
I can backup the hard disk on my video cassette recorder.
The 8" drive is capable of double density operation and Corvus
says they will have a DD driver for it in the future.
For the rest of this review, I'll divide things up into Hardware,
Software, and Documentation. I will make a few comments on the
printer at the end of this review.
Hardware:
The workstation uses a Motorola MC68000 microprocessor at 8MHz.
It comes with either 256Kb or 512Kb of memory, clock/calendar
with battery backup, 2 RS-232C ports, an Omninet interface
(RS-422 1Megabit/ sec.), 4 interface slots (Apple compatible), a
15" 35MHz display monitor.
Sockets are already on board for the second 256Kb of memory.
The chips used are MC6665L20 64K DRAMs.
The battery backup for the Clock/Calendar does not work
properly. This is true not only of my unit but also for the
unit purchased by a friend at the same time. On my unit, the
clock appears to run at about half speed when the power is
turned off. On the other unit, the date jumps about two days
when the power is turned off.
There are two serial ports on the Corvus but only one of them
can be used at a time. Bummer!
The display screen may be used in either a horizontal or
vertical orientation. Horizontally this yields 56 lines x 120
chars/line. Vertically: 72 lines x 90 chars/line.
Display updating uses DMA but with a 50Kb display that means a
lot of bus cycles are being stolen from the processor. Since
this is a workstation and not a multiuser system, however, this
isn't too bad and I haven't noticed any objectionable delays in
running programs.
Software:
The operating system claims to run up to ten processes. It does
not do time multiplexing however. What it does is let you do is
nest program executions ten deep. Unfortunately, the suspended
programs are not swapped out to disk but remain in memory. So,
the deeper you nest the less memory space available.
IO redirection works as it does in the UNIX shell.
There is only one level of directories.
Device drivers can be loaded while the system is running.
Pipes are implemented using temporary files written to a
directory called /PIPES.
A very nice character set editor is provided for defining
character fonts. New font files can be loaded while the system
is running.
The Pascal compiler supports most of the UCSD extensions. In
particular, UNIT IO and separate compilation (using units).
Unfortunately, the compiler enforces the strict rule of all
const definitions first, then all type definitions, then all var
declarations, etc. This means the use of include files loses
most of its benefits.
The terminal driver supports window management and graphics.
Pascal units are provided to support these.
The predefined type "string" is not only a predefined type but a
reserved keyword as well!! This means you cannot pass variables
of type "string" as procedure parameters since the syntax scan
of the compiler expects a parameter type to be an identifier not
a keyword. I worked around this, by defining:
type pstring = string[80];
Documentation:
The documentation provided is pretty good as far as it goes.
However, it took some requests to Corvus to get to get system
documentation: description of operating system, LINKER and
LIBRARY manuals, ASM68K manual, description of Pascal Units for
accessing system functions.
Some of the documents are missing chapters. For example, the
EDWORD (EDitor/WORDprocessor) manual contains seven chapters but
has references to chapters up to ten!
No information on memory locations of hardware interfaces, etc.,
etc., etc.
Okidata Printer:
The Okidata ML84 printer is great! Some features:
Fast. 200cps (bidirectional, logic-seeking)
Near Letter Quality mode. (unidirectional, two passes per line).
This not only overlaps the dots but puts the little sariphs on
the characters.
Underlining, emphasized, super/sub-scripts, graphics.
3 print densities and double width.
Down loadable character fonts. (Doesn't work!)
Electronic VFU and horizontal tabs. (No way to set or recognize
Bottom-of-Form!)
SUMMARY
Overall, I am quite pleased with the Concept hardware and rather
disappointed with the software. Function keys and menus are fine
when you are learning a new system but a pain once you know what you
are doing and have nest down thru several windows to get the
functions you want to execute. I am presently working on
implementing Software Tools using their Pascal compiler.
These are a few points that come to mind. If you have any
questions about the system, send me some mail and I will be glad to
answer them.
Hal Chambers
pur-ee!pur-phy!hal
pur-ee!Physics:hal
------------------------------
Date: 26 Nov 1982 2305-MST
From: JW-Peterson at UTAH-20 (John W. Peterson)
Subject: Computers for school
Those of you who reacted to CMU's plans to require students to
purchase computers may be interested in this one:
The "World Institute for Computer Assisted Teaching" (better known
to WORKS readers as WICAT) currently has a proposal before the Utah
state board of Education asking the state of Utah invest $15 million
in Wicat, for them to develop a standard CAI system to be used in
most/all of Utah's public schools. The state would be the sole
owner of the program and receive the royalties (6% of the sales
commissions) on the software. According to the proposal "the
royalties would continue [at >$3 meg a year] for a 15-year period
... meaning the state fund would be fully reimbursed and the state
would receive $30 million for the hardware acquisition."
The courseware would include subjects such as "English, Writing,
Calculus, Biology, History and Foreign Languages (with Audio)."
These programs are to run on a "System 300 (the Hydra System)" that
has 30 terminals "with audio, graphics and animation" and a CPU with
an 80 meg disk. Price (w/ discount) is given at $67,000.
WICAT's proposal also states they would be willing to "translate"
the programs to other hardware vendors at the states option (the
state would pay extra for this service).
While the proposal does list some of the reservations about such a
move (such as WICAT's current marketing capability, and having one
company as a sole source), Utah's executive directory of
administrative services claims "there are strong, positive feelings
about WICAT's offer and it's potential role in assisting the State
of Utah to develop and utilize CAI materials."
The proposal is set to go before the '83 session of the Utah state
legislature. The source for the above quotes is a Utah State Office
of Education newsletter.
-jw peterson
------------------------------
End of WorkS Digest
*******************
-------