#Open Letter to CBM
179 messages in this thread
[ continuation ]
that Commodore would not bankroll big guys to write for the new platform as
Steve Jobs has done with NeXT developers.
I have little doubt that the Amiga will continue to be a principle source
of revenue for Commodore over the next several years. I value my A3000
very much as a professional work station. I delight in the fact that I can
run three operating systems on my Amiga (AmigaDOS, MAC System 6.05, and
MS-DOS 3.30 ala SOFT-PC under AMAX-II). I look forward to adding UNIX as
well. I understand that I might be atypical in requiring the utility of all
these operating systems, but I really do use them all regularly.
If the Amiga user base is to continue to grow, the technology must continue
to evolve and empower its users. Commodore should open a new tier of
Amiga-based computing and offer hardware upgrade paths for existing users.
Market share is critical. Distribution channels are also critical.
Commodore has both if they incorporate the new technology into an Amiga;
they have no guarantees if it does otherwise. It tooks us six years to get
viable networking solutions. How long will it take for us to get adequate
laser printer/ bubble jet printer drivers, OCR software, etc?
Commodore has long been admired in the industry for its excellence in
engineering. I have much respect for the commitment my friends within the
company show toward the Amiga. Please let your voice be heard in this
forum.
H. W. GRIGDESBY Krannert Graduate School of Management Purdue University at
West Lafayette, Indiana
– via Whap!
I would believe that due to hardware reasons the 8 bit chipset (if there is
such a thing) would be a part of a new Amiga and perhaps the A3000. It is
my understanding that there is not enough bandwidth on an Amiga 500/2000 to
support more bit planes.
Considering the fact that Commodore does not feel it can support its Unix
product with continued development beyond the 2.02 about to be released I
find it extremely difficult to believe that Commodore intends to release a
new machine with a new operating system in the near future.
-mark=
– via Whap!
Mark,
I was stunned, too. But the source is reliable. As to Unix, only CBM-USA
wanted it. They account for 15% of CBM-Int'l Amiga unit-sales. Rather
than bash Commodore over Unix, I think praise is in order for the US
Operation. Amiga Unix can be a strategic market! Commodore USA is showing
great moral courage by looking at a long-term commitment to a market
segment that will enable it to enter industry in grand style.
Aside from the videography niche (yes, it's a growing segment… but it's
not growing as fast as expectations), Unix System V Release 4 Revision 2.0
is the only way to gain entry into industry. Countless X-window
applications exist and will work with a mere recompiling in this new
environment. The price, however, is too high.
Commodore is the bargain basement of computing. It delights in providing
terrific technology at prices any consumer can afford. Enter Unix, the
operating system of the decade and the next century. If Commodore can
deliver powerful color Unix workstations at a price-point below Sun, DEC,
and NeXT… watch out, stockholders.
The problem is: the lessons that Commodore has learned in marketing the
Amiga running AmigaDOS do not necessarily apply to the Amiga 3000UX. The
product manager of the UX must encourage X-window Software Developers to
make their code available for the 68020 version of Unix which runs on the
UX. The price-point of a UX with ULowell board should not exceed $2500 to
the academic end-user. Universities must be evangelized. Business will
follow
What does a Sun workstation cost these days?
From the July 1991 Price List. (Discount below prices by 40% for Academic
Purchasers)
SPARCstation SPARCstation SPARCstation SPARCstation
Product Name ELC IPC IPX 2
When Introduced July 1991 July 1990 July 1991 November '90
Packaging Desktop Desktop Desktop Desktop
SPECmarks 20.1 13.4 24.2 24.7
[ MORE ]
[ continuation ]
Processor SPARC SPARC SPARC SPARC Clock
Speed 33Mhz 25Mhz 40Mhz 40Mhz Main Memory
8-64 MB 8-48 MB 16-64 MB 32-128 MB
Memory Type 4&16 MB SIMMS 1&4 MB SIMMS 4&16MB SIMMS 4 MB SIMMS
Data Interchange/ Software Distr. CD-ROM CD-ROM/3.5"
CD-ROM/3.5" CD-ROM/3.5" Disk Capacity 207MB-5.2GB 207MB-15.6GB
207MB-15.6GB 424MB-20.8GB Graphics Mono Color
GX GX,GX+,GS,GT
Slots NA 2^2 SBus 2^2 SBus 3^2 SBus Tape
Backup ———–>150MB 1/4" tape or 2.3GB 8mm tape<——–Cache
Size 64KB 64KB 64KB 64KB
Bundled Software FOR ALL———> SunOS, ONC/NFS, OpenWindows, Open Look
Base Config: 8MB 8MB 16MB 32MB
17" mono 16" color 17" gray 19" mono
diskless 207MB 207MB 424MB
Base Price (MSRP) $4,995 $6,995 $11,995 $15,495
The Amiga 3000UXG should be positioned against the SPARCstation IPC. Here's
the rub: educational pricing for a UXG with 1950 monitor and tapestore
exceeds the suggested retail price of the SPARCstation IPC. That needs to
be fixed immediately!
– via Whap!
"Enter Unix, the operating system of the next decade and century"
You know… my first instinct was to post in reply the following:
"Har!"
But I overcame that (just barely) and will simply say, there isn't a chance
in heck of that happening. It's a niche operating system with more variants
than protzoa, and even less appeal.
–Ben
Ben,
In my part of the universe, you're totally wrong. Most Manufacturing
Planning and Control Systems (MRP-II), OPT, and "Just-in-Time"
manufacturing philosophies use UNIX based software. These packages run
mostly on SUN and DEC workstations.
A friend of mine obtained his PhD from Purdue in EE several years ago. You
might be interested to know that he wrote the 911 Emergency Software for
Indiana Bell. He has built a successful business around selling turnkey
UNIX systems to a myriad of clients from Fortune 500 companies to state and
local polic stations.
If your criticism of UNIX were applied to OS/2, then I would agree. UNIX,
Ben, is hardly a niche market.
Hal Grigdesby Krannert Graduate School of Management Purdue University
– via Whap!
You know, Unix is the operating system found in the most powerful desktop
computers in the world. I don't know about you, but I would consider that
fact in itself as appealing. One other thing about Unix, it is portable to
multiple architectures and processors. Don't try that with AmigaDos.
Maybe if we're lucky, we will see device independent graphics under
AmigaDos sometime this century. Memory protection would be nice too. Hey,
while we're at it, why don't we put serious networking in the Amiga
operating system. My point is maybe Unix isn't perfect, but its got a
couple of legs up on the Amiga's proprietary operating environment.
Tim
Timothy,
Thank-you for agreeing with me. Jim Dionne has brought AmigaLAN in from
Canada. It is a subset of TCP/IP NFS and runs entirely under AmigaDOS
2.04. It is being beta tested by JAC-CEN-DEL School Corporation in Osgood,
Indiana.
It is, essentially, a superset of AmigaNET… a Workbench 1.3 peer to peer
networking solution by Hydra Systems in the UK. AmigaNET was not Workbench
2.04 compatible, but I understand they are working on it.
AmigaLAN uses Commodore's ethernet board.
Unix is portable?? if so, that's a VERY recent philosophy change. Seems I
have to recompile darn near everything every time they come out with a
'later' version. Sheesh.
Vic,
I believe Unix applications have ALWAYS been specified as SOURCE compatible
which means you MUST recompile to the target machine. However, you should
not have to recode.
wmc
Wayne,
I wasn't complaining about recompiling to move to a different machine.
I was talking about getting a new release on the SAME machine. I will
grant you my experience is dated (ver 6, ver 7, sys III, sys V) all those
happened to me within 3 years, and startins with sys III a 'promise' that
this is the way it's gonna be…we're not gonna change. Yeah, sure.
Vic,
I must admit I don't have that much history with Unix. My total Unix
experience is about 2 years, so I'm still very much a novice. I don't
necessarily like it, but it is an increasing necessity for my livelihood.
So I'll just have to get used to it. I got used to DOS, so Unix should be
a little less stressful – at least it is a real operating system :^}
wmc
Source compatibility is only good for like UNIX versions. For example
source code that was written for Berkley UNIX does not compile without
modification on an UNIX System V system.
I think when we talk UNIX, we ought to specify which flavor. Be it NeXT's
Mach, AT&T System V Release 4, SunOs, ULTRIX, and/or BSD.
Pick your poison! All things with UNIX in the name are not the same.
-mark=
– via the Whapster!
Just so people don't get the wrong impression: It *is* possible to write
source on Berkley UNIX and compile it without modification on a System V
system. Not all source, since obviously code that has Berkley specific
things in it won't work, but a large amount of code will work.
Steve
Sure and if I write a K&R compatible program in C, it will compile on
almost any machine. 🙂
But the majority of programs I have run into; especially those with a BSD
hertiage takes advantage of the better enivronment provided with BSD.
I had the pleasure of porting a program several years ago from a BSD system
to our wonderful (pig!) 3B2 running UNIX System Release 3. Do I need to go
on? 🙂
Needless to say, System V Release 4 comes a long way to making life better.
-mark=
– via the Whapster!
Mark,
Out of curiosity…which 3B2 model?
Just as an aside, and my experience with 3B2's confirms it, they
are relatively slow when compared to other machines under a light user
load…but as you add users…the 3B2's tend to maintain their performance
much better than many other machines. Till they hit swapping of course.
UNIX Review noted this same behavior when they tested the 3B2-1000/80 a
year or so ago.
Don
Mark,
Not necessairly true. It all depends on how complex the code is.
It also depends on the implementation. For example, Pyramid has OSx which
is a 'dual port' UNIX. The kernel is BSD 4.2, but the system comes with
both SVR3 and BSD link libraries that do any necessary conversions. At the
user level you get to select your 'universe' and can switch back and forth
between SV and BSD freely not matter what your default or current universe.
I've recently ported some quite complex code from pure SVR3
machines to the Pyramid with simple re-compiles…no changes at all.
Don
I must admit little knowledge of a Pyramid machine. It does sound decent
from your description. Obviously there are exceptions to what I said.
However, you must admit that trying to move code from BSD to System V
Release 3 machine that uses BSD features is not something you can just sit
down and do 'instantly'.
-mark=
– via the Whapster!
Mark,
I know. I found that out when I tried "ls" on DomainIX from Apollo vs a
unix on a Zilog mini.
wmc
For home computers…you're right. For business usage…you're
wrong, for governments…it's already a fact that UNIX has "taken over".
For business usage, MS-DOS is by far the most common operating environment.
For gov't use, I have no idea as I stay as far away from those involved in
professional inefficiency as I can. I can't say I'm surprised, considering
how poor gov't data processing is, but I guess they need some kind of
standarization. Too bad the field to pick from is so small.
One thing I am sure of – UNIX isn't the operating system of the future, nor
is the operating system of today. That's wishful thinking on the part of
unix users, predictable and completely wrong. Time will bear me out; in the
meantime, I don't feel bad for letting forum readers know that claims like
"OS of the decade and the next century" are seen as balderdash by many
people who don't have a stake in Unix.
You might look to CBM for a hint… all that effort on Unix, and it
wouldn't sell. A good implementation from all reports, yet not anything
that anyone wanted. I think the only people who were suprised were… you
guessed it… Unix people.
–Ben
You know Ben I agree with you on your comments about UNIX. UNIX does not
grab me and say "look at me.. I am in your future". Now I grant that the
work being done on the POSIX standard will do the UNIX world and perhaps
computing in general some good. But operating system of the next century?
I pray not.
I think Unix has a real future, and most definitely into the next century.
There are loads of user-friendly GUIs out there for Unix. The trend is
towards simplicity and power.
User-friendly GUIs for UNIX? Where oh where? XWindows? Open Look? Motif?
Tell me can the average person start a unix box, get into Open Look and do
something as easy as change the font size? Hell even on the Amiga 3000UX
where it is all installed on the hard disk you have to run the oladduser
program, and then start it with 'olinit'. That is user friendly? Please.
My experience with Suns has more than proven to me that the current X
implementations do not make things easy for the user, unless you have a
UNIX god set it up for you. The .xinit file is enough to drive most away.
I am not even going to point out that you the GUIs to date are miserably
slow unless you have a high-powered box. To me that sounds like the
Microsoft approach to the future.
UNIX has a future only in that it happens to be on tons of workstations. It
does not represent "state-of-the-art" by any means.
You speak of trend, trend says that MS/DOS is the operating system of the
future. 🙂 That is a trend that I intend to ignore.
Personally I find DEC's VMS product much more useful and much easier to use
than UNIX. Of course I run a VAX Cluster for a living so I guess I have a
personal bias. Give me DECWindows on VMS anytime.
The future, at least according to the trade magazines is toward POSIX
complient operating systems. To date, I do not know of a single OS that
completely matches this 'standard', including UNIX.
I also wanted to ask this question to those who say UNIX is the future of
computing. Which UNIX? Is the future in AT&T's UNIX System V Release 4?
is it Berkley UNIX? is it Mach? is it Xenix? is it SCO Unix?
So many futures, so little time. 🙂
Mark,
I'm really not on Steve Job's payroll, but the latest iteration of the NeXT
is user-friendly and very powerful. It's hard for normal human beings to
find dealerships that carry NeXT, but look for one. I have mixed feeling
about proprietary UNIX (Mach), but Steve Jobs has certainly made that Unix
implementation attractive and affordable.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Harold, we all have gripes about CBM's marketing plans sometimes, but I do
believe CBM DOES listen if you sit down and write them a personal letter
and express your feelings. I wrote Mehdi Ali a letter expressing my
concerns about the 2.0 non-release scare, and received a nice letter back
from Bill Sydnes, Vice President of Engineering. If you have concerns about
CBM's marketing plans, and if you haven't already done so, write Mehdi Ali
a letter and express those concerns. I believe CBM does listen, even though
we sometimes think otherwise.
John – K7KB (via Whap!)
You assume me to be unwashed! I have fondled a NeXT, and I am quite aware
of the nice GUI that is placed on the NeXT. I am also aware that the NeXT
while cute is quite unaffordable and not likely to succeed in the long run.
It suffers the same problem the A3000UX does, people ask the question "Why
should I buy a NeXT instead of a Sun?"
So because of this, I do not include the NeXT as part of the mainstream. So
my argument still holds. The GUIs on the mainstream UNIX boxes suck wind.
🙂
-mark=
– via the Whapster!
Don't even worry about it. When you hear things like that, look who's
talking. Don (SYSOP) is a Unix system handler; Larry works for a company
that makes powerful Unix systems; Mark is at a university where, unless I
miss my –>guess, they use Unix in the computer dept (proving that schools
are always far behind in the technology arena).
I'd say the Amiga has a better chance in the long run, because the tech is
very close to being up to date, and is likely to become up to date in the
near future. It wouldn't take much. And, as an OS, it just needs memory
protection to be "serious".
But then, I'm an Amiga user – and that's what I'm likely to think. Of
course, I've been around a bit too, and I'd like to think that some of that
contributes to a reasonable evaluation.
Time will tell. But I'd bet dinner on it. :^)
> all it needs is memory protection ..
Yeah, sure… and that little gem will change the entire Amiga gestalt so
much that it would be nearly unrecognizable.
Regarding memory protection, you bet it would. But the OS is nice as is,
otherwise – the OS, not the apps.
As you apparently realized when you said it, MS-DOS runs on far more
machines.
Let me point something out here: I am *not* saying tha MSDOS is superior to
Unix. I think otherwise. What I am saying is the Unix isn't going to "take
over" from MSDOS. Power isn't the issue, if it was, they'd all have Amiga's
and be laughing Unix out of the room for price/performance. :^) People will
NOT buy something they don't understand; and Unix isn't understandable by a
person who doesn't even understand MSDOS.
Ben,
You really suprise me. As a developer you deal with CBM in ways that let
you see how *(&^#-up they are and you think they will some how wake up and
out-market, IBM, Apple, AT&T, HP, UNISYS, Sun, etc before 1999????
Could you send me some of what you're smoking!?!?! ;^}
wmc
Heck no I don't think so. Heck, no.
I *hope* so.
–Ben
Are you always such a CBM pessimist? Get up from behind your computer
and go walk in the sunshine. Smile. Laugh. Enjoy life. Unix is
cool. The ULowell board is sweet. A Nec5D + ULowell Board +
Revision 2 even edges out Ham-E.
If a board priced where the UofL baord is can't edge out a HAM-E, then it's
junk. :^)
yes, I'm always such a CBM pessimist. I've been involved with Amiga
development since before the machine hit the stores in Oct '85. CBM hasn't
let me down yet in my (low) expectations. They have yet to make any
reasonable approach to the marketplace. The technical people are wonderful,
but I'd not shed a tear if the entire marketing group was lined up and
shot.
Unix rots – it's overcomplex, aged beyond belief, cryptic, most
implementations are poor multitaskers, VERY few are what you'd call a
"real-time" operating system (giving nod to Modcomp), and as such, I'd not
want them in my computer. Too many things you can't do without real time
capabilties. It needs a real user interface, instead of the abuser
interface (awk, indeed :^) it has.
And yes, I know about xwindows and etc; fine – let me know when all Unix
systems have xwindows, and I'll be glad to agree it's moved as far forward
as 1985.
Regarding the UofL board in general – an FC24 blows it out of the water,
bodily. Any of the DMI Resolver, the PP&S Rambrandt, or the GVP wowzer
board make it look old tech, which it is; note that while the DMI Resolver
is an 8 bitplane 32010 based display (like the UofL), it has software,
unlike the UofL – although there are good odds that the SAGE lib may work
with it, in which case it will be come useful. In any case, DMI is coming
with a 24 bit display, and that WILL blow the UofL out in a big way (32020
driven).
The HAM-E is a $300 add-on for any Amiga. As such, it's great. We never
said it was anything else, just that it's good at what it does. And it is.
:^) It's better than anything anywhere NEAR it's price class.
In the meantime, Black Belt has support software for the FC24, the stock
Amiga, and the HAM-E; we're coming with software for the DMI resolver and
the PP&S Rambrandt, and some other stuff is cooking beyond that. The only
thing I have to say about the UofL that's positive is that DMI's SAGE
[ MORE ]
Thanks for the nod. I'll pass that along to everyone at work. 🙂
I'm curious as to how much experience you've had using UNIX. What type of
system(s) did you use? What kind of projects? What was it that turned
you off about it, and how long did you take to reach your conclusion?
Steve
I've spent little time using Unix, in the low level sense. I've spent a lot
of time at the desk next to the system administrator listening to his
comments about Unix; I've had to do some simple development things, where I
kept my needs down to assemble, compile, outload to the hardware. I had a
nice list of the cmds next to my desk, a bunch of totally indecipherable
names to do things (I have little idea what names went with what functions
– I called for help when I need to do anything more complex than a file
copy). What turns me off about it? It's not realtime (yourselves excluded,
of course) so there are a multitude of things it can't manage without
extra, very expensive, hardware with it's own processors. In the past (and
possibly now) you couldn't install a new device driver withoubringing down
the system and re-something or othering it, which I've always felt was a
major, major flaw. Especially for a big system! Until the relatively recent
advent of a standard for gfx ops (xwindows), the environment was a mishmash
of user interfaces, or the abuser interface, the command line. :^) Most
implementations are HUGE, and that's a bad sign… I don't need (I don't
think anyone needs) tens of megabytes to hold an operating system. OR
hundreds. OR even just ten. Storage, both ram and archival, is changing,
and this won't be significant for too much longer… but I'd sure like to
see something efficient, like OS9 rather than a hog. I've done C compiling
and assembling on unix machines, it feels the same as anything else,
although for my work, it felt REAL slow… I know that was because it was
one big machine for a bunch (70) of people, most of whoom were active. We
user Harris Vulcan and somebody elses Unix – from massachusets is all I
remember – we saw them a lot (remember I sat next to the system operator).
Steve, part of my personal feeling that Unix is a bad deal is that I sit in
front of a real-time OS machine, with windows, CLI, development tools,
reasonably fast graphics, user installable device drivers, file systems,
[ MORE ]
Odd, most of the OS's I use on Unix machines are between 900K and 1.2
megabytes. When you realize what all this entails, like the device drivers,
networking primitives, and so on, it isn't too bad. I wonder what machine
you were using that had an OS that was 'tens of megabytes in size'.
-larry
[ continuation ]
you name it – with only a few holes in same (protection specifically) and I
see that I'd like things to go "this way". Stepping back from what I want
to what the market wants, they want MSDOS – they want powerful MSDOS I'll
grant you, but that is ALL they want. I have my own reasons, which are (as
usual) different from the average user. Bottom line, no one wants Unix, and
my conclusion is that's a good thing because I don't want it either. :^)
What DO I want? First, I'd like all CBM marketing execs shot… no, we've
gone over this already. :^)
–Ben
Thanks for your perspective. It's unfortunate that your experience with
that UNIX box was so bad.
But hey, what can I say? Some of what you want can be done with hardware
can be done with REAL/IX via shared memory and connected interrupts….all
from user level processes. The only down time involved is to plug that
board it. 😉
And I wouldn't say that no one wants UNIX either…if that were true, SUN
wouldn't be making the big bucks. From my perspective, and from what I've
seen, almost no one wants MS-DOS….not when they can have UNIX instead.
For course, that's only from my perspective, since I don't hang out with
MS-DOS folks, or go to places that use MS-DOS. :*)
And there are millions of UNIX users….heck, just look at the Internet and
USENET…..bazillions of machines and quadzillions of users.
The actual OS itself and the number of utilities you need to get it running
in an embedded application is actually pretty small. Not as small as
others, but pretty small. It's all the utilities that make it take up tons
of disk space. In the case of REAL/IX on a 68K, the OS itself is less than
a MEG.
Now, even I think that's too big for UNIX. I wish it were smaller, but it
isn't.
It'll be interesting to see what The Labs does with Plan 9 and what
Tannenbaum does with Amoeba….at least I know those will be tiny. :*)
Steve
Steve
Steve,
I agree with Ben on the userfreindliness of Unix – it ain't. I disagree
with him on the multitasking, flexibility, even display cards (check out
suns Hi-Res moving picture in picture, freeze frame etc board).
I also disagree about the "nichiness" of it – the damn thing is EVERYWHERE!
wmc
UNIX is user aggressive. 🙂
Steve
[ continuation ]
library may make it useful. As a front for xwindows on the Enuich CBM is
making, I guess I'm glad it's there for you unix types. Me, I'll develop
for the OS which is being supported (IE, there are still employees left).
Last but not least, I'm smiling – I'm one of the most cheerful guys you'll
ever meet. I just don't expect much from CBM (or Unix). It never comes, and
I'm NEVER dissapointed. ;^)))
–Ben
Ben,
I am glad you took my retort in the good humor in which it was intended. I
have sold and do now offer your product, DCTV, and the FireCracker Board in
my retail store. I, and my customers, are aware of each product's pros and
cons.
Since C= owns the right to the ULowell Board, they control the cost of it
to a large extent. If they use the right pricing strategy, they can really
deliver low cost Unix work-stations. That was my point.
In understand your position clearly: You don't like Unix. You don't like
Commodore. You are courageous in your own right, because you are an
entrepreneur. It takes guts to bring a new product to market with little
outside financial backing. I respect you for that.
Please understand that we come from to opposite perspectives of the Unix
market place. As a former Regular Army Captain (now Reservist), I have
some sense of why UNIX is both attractive to the government and in
particular the military. The Army's personnel system (in Europe at least)
is Unix-based software. Logistical support software packages are also Unix
based. In the Purdue University community, UNIX is the most important
operating system on campus. Remember, Purdue is known for advanced
research in the sciences and engineering. The Business School is also
highly regarded nationally and internationally. (Last year, we came in at
#20 amongst all business schools, both public and private, within the US).
Unix will continue to become more powerful and simpler as time progresses.
Where is Unix now? Look at Mach! Spend some time in front of a NeXT
machine and understand that UNIX doesn't have to be implemented in such a
fashion as to baffle the end-user.
In Wednesday's Wall Street Journal, (10/2/91), the IBM-Apple joint venture
was profiled on page A3. I was suprised to read that a major strategy of
that alliance was the merging of A/IX and A/UX, the IBM and Apple versions
of UNIX. They are developing an operationing system (PINK) for IBM's Risc
[ MORE ]
As a Purdue Alum, I can't help but add that Purdue was where the Dual-Vax
(Master/Slave) was invented….and it ran under UNIX.
Steve, you're a Boilermaker too? Add me to the list…though I never
graduated.
Michael
[ continuation ]
6000 processor. Apple intends to make their high-end Mac's around this
chip.
A couple of things are noteworthy about this alliance.
1) It attempts to cut Microsoft's powerbase.
2) It challenges AT&T's Unix standard.
3) Apple is choosing to continue the Macintosh family into a new,
high-end tier utilizing advanced technologies.
4) It allows software developers to write programs for the Risc 6000,
regardless of whether they run on an IBM or Apple labelled machine.
Your thoughts please, Ben.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Just to muddy the water, a DMI rep at AmiEXPO said the Resolver Board *IS*
the A2410 replacement and that C= has turned over the Lowell code for
porting to the Resolver.
Dear Rob,
I hadn't heard that. I was a bit dismayed at the state of packaging of the
ULowell Board. The one which arrived for the Purdue University Computing
Fair was sans instructions and was shipped in a plain white carton. The
UXD which arrived was shipped with draft manuals which didn't cover how to
turn-on the ULowell board.
After a phone call to John Higgins of Commodore Inside Sales, he completely
squared away our knowledge gap in how to use the board. In my book, John
Higgins is one of Commodore's bright shining stars. He definitely
contributed to a successful show at Purdue. (Also notewrthy in the fair's
success was Mr. Larry Noetlich, Central Regional Sales Manager, and Mr.
Steve Johnson, C='s Chicago-based account executive).
I am aware of the DMI Resolver as that company has faxed me specs in hopes
that I will pick-up the product. The price point (retail) is around $2000
at last check. (One of my people provided me that info; I don't recall the
exact price). At any rate, it was significantly higher than the ULowell
Board.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
My thoughts… What IBM and Apple do here isn't going to change a thing.
MS-DOS, under Windows, will continue to be the "thing" for the vast
majority of PC users (who are the vast majority of all computer users…
"large machines" are now far in the minority, probably too far to recover).
As we have seen, IBM, big as it is, flounders every time thry try to change
anything about the status quo. OS/2 flopped (and I'd point out that I said
it would, from the day it was released, right here, in big bold letters :^)
The XGA graphics standard they are *trying* to promote is in the process of
flopping – gfx developers want to create hardware that accellerates
windows, _specifically_. IBM's two approaches at "jr" hardware flopped.
IBM's microchannel flopped. In short, if it's not MSDOS std stuff, it
flops. Unix is next in line – it doesn't meet the base criteria, and it
will flop. My opinion, of course.
For Apple, if they deviate from the main Apple line, I think it will flop.
For CBM, if they deviate from the main Amiga line, I think it will flop.
You mentioned price control of UofL as a key factor in CBM's success (or
non) of the Unix product. I agree. Also for the rest of the Amiga line.
I've maintained for a long time now that what CBm *should* do is include
the best bridgeboard they have in every 2000 series frame they sell. No
option to remove it. That gets them listed as a PC compatible, and that
will sell machines. The board is relatively cheap, it should be in there at
cost, no more. Also, the markup on 2000 series frames is extremely high.
That could, and should, be shaved. The Unix system was pretty high, too,
and I'm sure htat bit them – the UofL board probably wasn't a major factor
in the cost of the system, at the price they set. In my opinion as an
engineer.
In short, I don't think the Apple/IBM thing will do diddly but raise a lot
of smoke. In any case, it will be *years* before an effort like that can
provide results – by then, it'll be old hat anyway. Tech changes too fast.
[ MORE ]
[ continuation ]
I'll probably be running my 68060 at 250 mhz and laughing my way to the
keyboard.
Regarding Purdue being highly regarded, and running Unix, and the former
providing credence to the latter…
I don't buy it. Credit for good work in any particular area should be
awarded only on the basis of that work. If people credit me with being a
quality martial artist because the company's name implies that and we do
other good work (the implication being that Purdue produces quality
graduates, the only measure of a good school that means anything), then
they're reaching. To say that Unix is good, because Purdue is good *and*
they use Unix doesn't follow.
Hey – Unix works. It's an OS, and it's been successfully used to teach many
people many phases of computing. Fine. But it's old, the number of new
coats of paint it's got hasn't made it into a realtime OS, and it's
unlikely to change in the future (again, nodding at Modcomp, who, if you're
not aware, have a realtime Unix).
The premise I'm disagreeing with here is that Unix is significant enough
that the world will want to change to it. I think not. Some big machines,
doing some big things, are running Unix or a close variant. You have to
remember how many machines aren't running Unix – most of them.
And my solid conviction on this point is based upon an inner certainty that
the average user does not, under any circumstances, want to learn anything
new. They dig in their heels and they will not budge. And the threat to a
company when something as major as the OS changes is huge. For that reason,
I think that where Unix is, Unix will stay. Where MSDROSS is, it will stay.
New systems will be run under groups of small machines, netted under some
variant of MSDOS.
That's what I think.
–Ben
Here is some interesting information for you Ben. Commodore has decided
that they would like to sell some Amiga 2000s! Effectively immediately the
A2000 suggested retail price is $999. The A2000HD is $1295. It is about
time Commodore!
I agree with your stance on UNIX mostly. I do think that UNIX will be with
us for the forseeable future because the workstations are built around it
anymore. I think people (John Q. Public) are resistant to change, and will
never accept UNIX as a standard for lots of reasons. The biggest reason is
that it currently requires too much understanding of computers to be able
to use UNIX effectively.
I believe people are willing to abandon the status-quo when they see
something that they can use, that is easy to use and doesn't require them
to toss all of their software in a garbage can. I believe that is why
Windows is successful while OS/2 is not.
I do not think that a UNIX based personal computer with that being its
native operating system with no other will ever sell well to John Q.
Public. AT&T could not sell its UNIX-PC (of course AT&T could not sell a
glass of ice water in the desert).
The bottom line is that no matter what type of GUI is glued on top. UNIX
will still be UNIX. 🙂 Complete with Panics and lost i-nodes.
I think more than likely the future computing world will embrace a
Windows-like GUI that is very much like what we have today on the Amiga
with further refinement.
Just think, in 1990 PC users discovered multitasking. We of course
discovered it in 1985. I suppose by the time 1995 rolls around we will be
experiementing with virtual reality and the PC user will just discover the
value of co-processors and a real multitasking operating system. I have
faith in the common man. Computers and users that follow the mainstream
"trend" will continue to block the path to innovation and revolution.
But hey? What else is new?
-mark=
– via the Whapster!
Ben,
We are definitely talking about two different markets right now. The
market which you are most comfortable with is the personal computer market.
The market which I'm describing is the personal workstation market. The
hardware and software needs today definitely diverge one from the other.
A few years ago, a C64 was sufficient for my computing needs. Later, I got
an Amiga 1000 (1.5MB Ram, Sidecar, 30MB HD). A year ago, I got my Amiga
3000 (105mb HD, 14MB Ram, AMAX-II, XETEC CD-ROM, etc, Framebuffer, ad
nauseum).
I believe a trend exists. Either I am preoccupied with Amiga toys, my
reliance on software tools is growing as life becomes more frantic, or all
of the above.
Here's my thesis: The trend is much more powerful computers are entering
homes every year. This is a fact that is lost on IBM. Their approach to
home computing is providing of subset of office capabilities in a home
machine, aka PS/1 or PS/1.5. I believe consumers are smarter than that.
Most folks who purchase a computer these days have used one at the office.
They want to use the same software at home. It is important for them to
take their floppy (5.25", 3.5", 360K, 720K, 1.2MB, 1.44MB) to and from work
without data translation gyrations.
IBM's principle strategy has been to play off the first time buyer's fear
of computing. In the PS/1 literature, a very strong theme surfaces. Sell
Big Blues image of standing behind the end user. Make service channels
responsive and readily available (many locations over a large geographical
area). With Sears as a vendor, they got half of the strategy right. The
problem is: the first time buyer today is more sophisticated than the
first time buyer of 10 years ago. Chances are, these consumers now know
what criteria (hardware/software) to look for. They often consult friends
who have machines. (The "safety in numbers" mentality is often a
significant entry barrier for Amiga salespeople to overcome).
Your idea of a standard bridgeboard is wonderful. Let's add VGA graphics
[ MORE ]
[ continuation ]
to the BB and a 386SX processor w/2 mb ram. Let's have Commodore provide as
standard equipment a VGA switch cable with every CBM VGA or Multisync
monitor sold.
A certain segment of 2000 purchasers want a plain vanilla machine. More
often then not, it's for one of two reasons: lower initial cost or they
want someone else's accelerator/harddrive/memory expansion. Commodore
knows what the mix is. I understand that 2000 sales have been flat for a
while with the trend shifting to the 3000. Your suggestion may help drive
demand for the 2000.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Well, what OS will be the OS of the future? UNIX certainly will be for
the coming years. It's where Universities are going, where Government is
going, and where business is going.
Your example of CBM's inability to sell UNIX is pretty weak. Look how well
they've handled the Amiga.
The OS of today, and of the future, is and will be some variant of MSDOS
running in a windowed environment.
The machines will get bigger, the nets will get "thicker", and more
sophisticated s/w will expand these systems. There are nearly 100 million
MS-DOS systems installed. Can Unix claim a market even 1% that size? I
don't know, but I would tend to doubt it. MSDOS will change, but not in a
way that threatens the user. Hardware will become more powerful, and the
machines faster, more capable of big jobs; heavy duty file servers will
proliferate to handle the big data chores, but the machines that do the
most work will be right on the desk, running esoteric software like Word
Perfect and Lotus under MSDOS (Yes, I know there are Unix variants).
Again, I see you asserting that Unix is it "for the coming years". And I
step back, and I look, and do I see that you are directly associated with
Unix? Yep. Right? I know of _one_ business that uses Unix to run things,
that's our computer dealer in Denver. I know (mainly because I'm active
here) a few people who work with Unix. They all, without exception, sing
it's praises, and make claims that it's the coming thing. But I don't see
anything in the market that says so. Just look at a magazine rack. Count
the mags. That little piece of research will tell you right there what the
system distribution is.
Business is clearly going with MSDOS, nothing else, period. If accounting
has a Unix based system, or if engineering does, there will likely be
enough PC's in the same location to bury the Unix system physically. Claims
that business is "going to Unix" are hogwash, and promulgated by Unix
devotees. It's like hearing that people are "going to the Amiga". Only an
Amiga user could believe that one. :^)
If universities – ALL of them – go to Unix for ALL machines in the shool
system, and the gov't goes to Unix, for ALL machines in the gov't system,
you've got a lot of Unix machines. I don't beleive for a second that it
will happen (just saw a contract for the local indian res, gov't dept that
[ MORE ]
[ continuation ]
handles their health care – they are going for MSDOS, that's the
requirement. Wasn't a WORD about Unix)… but even if it does, the average
business fella will use his PC, because he's already got it, he's already
learned how to handle it (somewhat) and he's got better things to do than
spend time learning something because a bunch of computer weenies think
it's better than what he's got. Until he has a need that the installed base
of machines (MSDOS clones) can't meet, he won't change. And the thing is,
all the time the Unix people are cooking up new h/w and s/w, so are the
MSDOS people. When he goes back to market in ten years or so, because he
just can't run the s/w he needs on that 386 anymore, there will be a nice,
easy MSDOS solution waiting in the wings. It'll be mass marketed, like all
previous to it, cheaper than anything slow, costs heavy companies like
Modcomp and Sun and IBM and Apple can provide, and it will run all his old
stuff.
That's my opinion. I'm happy to wait and see what happens, which I do NOT
know, I can only guess – but I will say that the things I've put forth here
make a lot more sense to me than repeated cries of Unix! Unix! coming from
a tiny sector of people who have a sourcing stake in such a series of
events. I've heard cries of OS/2, of OS9, of Unix, of AmigaDOS, and of
older, more obscure things like Vulcan and some of IBM's weirder stuff. The
only significant palyer in the market at this time is MSDOS, and it is not
showing any weakness – there is no basis for the assumption that it is on
the way out that I can see. Universities are typically way behind the times
because they have poor budgets for such hankie panky, and the gov't can be
even worse for budget reasons, and gov't inertia reasons.
The cry in engineering (and I am SURE you've heard this) is "Good enough
for gov't work". The reason that cry is sounded is because the gov't is the
very embodiment of mediocrity in it's decisions and it's actions. To say
that something has credence because the gov't, of all people, says
[ MORE ]
[ continuation ]
it's the thing to do, is ludicrous. "Whatever the gov't shall do, thou
shalt run the other way if thou wishest success in life and business".
They're a great market for a bunch of big companies pounding out mediocre
hardware and software (yes, I mean Unix). More power to you – sell em all
you can, I'd rather they bought computers of ANY kind than built more
bombs. But they don't lend credence to Unix as a "good" system.
–Ben
I finally realized that your definition of business is "Joe Business guy
running his business", while mine is "Joe Industry running his plant". For
small business, I agree with you. For large scale business, such as
Alcoa, I do not.
My position here is based upon the majority of machines that are out there,
IE, computers. There are more MSDOS machines than Unix machines. That's all
I'm saying. There are more MSDOS developers than Unix developers. There are
more MSDOS applications than Unix applications. There are more machines
that concern themelves with PC emulation/compatibility than with Unix; all
the "big" machines can pretty well do both. The Amiga can do MSDOS and
Unix. So can the Mac. There are Unix implementations (myriad) for the PC.
There are (myriad) Unix machines that can run MSDOS in a window, more or
less. So it's complex.
But bottom line, I see no compelling evidence that this is the
"decade/century" of Unix. I see things going just the other way. Windows,
on a PC, going faster and faster until it is (at last) usable. A machine on
each desk – running MSDOS (some variant thereof). You don't see many PC's
running a plant – so far, they're a little bit lacking in power. Not much –
ten years ago, you couldn't buy the compute power a 50 mhz 486 puts on the
desk (or a 25 mhz 68040) to run a plant without being Boeing or someone
like that. Now, joe user can do it. But it may not be long before you DO
see it. Fault tolerant machines are there already, powerful machines are
there, big memory machines are there, networks are in place… I see no
need for Unix, other than to support companies that make Unix.
–Ben
Ben,
If you're counting machines…that's stupidity. A few years ago,
using your logic…the world was moving to the C=64! After all, for quite
a number of years, it was the most popular selling computer.
Of course there are more MS-DOS boxes out there than there are UNIX
boxes, if for no other reason…you need more MS-DOS boxes to do the same
amount of work!
BTW…while you're absolutely right…I run the UNIX systems at
work, but that's not why I like UNIX. I got into running the systems
because I liked UNIX, not the other way around.
Don
I know nothing at all about UNIX, but think I have a comment of interest
in this thread.
Reading these messages, I have concocted a analogy between UNIX the OS
and HP RPN (reverse polish notation) calculators. I used to hate HP calcs
because they were difficult to learn how to use and seemed very user
unfriendly. But one day on a whim I bought one. By the end of the week I
was hooked. I now find RPN to be much easier to use than standard
calculators (algebraic ones) and have since purchased 3 other HP's. They
are undoubtable the best.
Unix seems to be in a similar category. Unix users, once they learn it,
love it and can't say enough about it. People who are not unix fluent hate
it because of its complexity and poor interface.
Try as I might to convince people to switch to HP's, I more often than
not fail. Why? As Ben points out, most people do not want to invest the
time to learn the new system. I think he is correct about UNIX. Whatever
it's advantage, real or imagined, it will not catch on at the desktop
level.
<<< Michael Shea >>>>
<<< 76665,3360 >>>>
Michael;
Well we sure agree on the HP calculators. I was a little more fortunate in
that the first 'pocket' calculator that I learned to use was an HP-35. I
have never liked so called 'algebraic' calculators since.
Your analogy sounds pretty good too. -bill – via Whap!
Steve;
BTW, while we're writing things that we hope CBM is listening to…. I
picked <flames> on another network over Unix Reviews' review of the
workstations. I don't know how valid this is but the guy claimed that the
major contributor to the Amy's poor showing was that the swap space was too
small.
Again, I don't know if he is right but if so then his other statement was
also correct in that while one might expect such things from Byte or a PC
rag, Unix review should know better than to handicap a machine in that
fashion.
He did not publish any numbers (which brings his assertions into question)
but he stated that when properly configured, the Amy Unix box gives
performance figures comparible to what one would expect based upon
processor differences. -bill – via Whap!
Not only that, but running the tests under Beta software is inexcusable
too. Unix Review should know better than that.
Steve
It will be interesting to see how things go when the environment is better
known. Although a poor implementation is the obvious "whipping person", it
sure seems funny that a machine as powerful as the 3000 would fare poorly
in those benchies.
Wonder how the 040 would help. :^)
–Ben
Ben,
When I boot up the UXG that sits in my store, it informs me that the
version present is optimized for the 68020. Commodore's draft manuals
state that this is indeed the case.
My educated guess is that an 040 might not work with Unix. To be honest, I
really haven't the foggiest… we need Commodore to answer that one. The
draft manuals also are silent on the issue.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Ben,
>…a good implementation from all reports
Garbage…it tested out VERY poor when compared to other SVR4
machines. It *may* have all the goodies there…but it doesn't have the
performance and that's the bottom line.
At this point in time, according to articles I've read (obviously
not done the research myself)…in Fortune 500 companies…UNIX is the
prefered OS because among other things….it's not hardware dependant.
Don
What *I* read was that it was superior to at least some others. They threw
the Mac implementation out of the running at at least one school on the
basis of benchmarks and reliability – or so they said.
It's not hardware dependant – so, I can take binaries from Larry's machine
and run them on a 486 platform? And the Mac? And the Amiga? And the NeXt?
Without compiling? Without changing disk formats? If so, then the
performance is going to be garbage. If not, it's not "hardware
independant". If by hardware independant, you mean that you can recompile
and run, fine – except that again, the vast majority of computer users (who
are MSDOS users, of course) will NEVER do this in a million years.
Unix is far better than MSDOS. Fine. And it doesn't stand a chance against
it as the most popular / common / used operating system. I've not heard a
word here to make me think otherwise – but time will tell.
–Ben
Ben,
As I've said all along….for home/desktop use…no, I don't
believe that UNIX will "take over". It's simply not the best environment
for simple tasks such as word processing and simple spread-sheets and/or
databases. That's what the vast majority of MS-DOS machines are being used
for.
However, when you get into bigger applications, networked systems,
have a need for a real multi-tasking OS, want to put multiple users on a
single machine, want to run client-server architectures over simple serial
data lines, then you're moving in the UNIX arena.
As to hardware independence…with the other large machine OS's out
there such as VMS, they're proprietary…you're stuck with DEC (or whoever)
equipment.
With UNIX, you don't like AT&T's machines? Fine…buy Pyramid's,
or Sun's, or DEC's (under Ultrix), or Honeywell-Bull's or Data General's
,etc.
Under the older versions of UNIX…a simple recompile was all that
was necessary in most cases. Under the newer versions of UNIX, even that
may not be necessary.
I've moved entire applications from AT&T 3B2's (and 3B4000's) onto
both a Pyramid and an AT&T 386 box (running UNIX, not MS-DOS). These were
large complex applications. None of the processors are compatible.
Recompiles were ALL that was necessary….not one line of code was changed.
Script files moved as is..no change at all. Other than the import of the
data into a database, the move took less than a day, most in an hour or
two.
Yes…these were all applications running under Oracle and those
executables were "new" for each machine…but our code included many
thousands of lines of custom code, programs, forms and so forth.
Don
Ok. Well, what you call "compatible hardware" I'd call a system-superuser
dependant cross-compile, something that is of no use to the average user
with their own machine.
And ALL that I say is that Unix isn't poised to be the OS of the future, in
the sense that it won't be the most common OS.
I find it interesting that you say it's not the best environment for WP and
spread sheets – why? Isn't the Unix OS good enough? Xwindows looks great in
pictures, so what's wrong?
–Ben
Ben,
Because it'd be a pure waste of money and UNIX is complex. While
yes, you can take a box like the 3000UX out of the box and run it under
UNIX for simple tasks like word processing and small spreadsheets…it
makes no sense.
It'd be like buying a 10×10 dump truck so you could take grass
clippings out to the dump once a week. Or ordering out a load of concrete
on a cement mixer truck when you've got one post hole to fill.
All of the above can be done, but there are much easier, cheaper
ways of doing it…and more efficiently.
Don
Unfortunatly, claims aside, your facts are plain wrong, Ben. Coopers and
Lybrand, Exxon Data services, many airlines and air freight companies, all
the telephone companies, most banks I know of, department store chains,
etc. run Unix. I believe if you had number of installations they would be
smaller thatn DOS, but number of users would be greater do to the
multi-user and heavey netowrking found in the Unix world.
Also your contentions about government programming are totally false. As
No commercial venture aside from IBM could take over 2000 software
engineers spread across the country, produce 5 million lines of code,
integrate it and launch space shuttles with such high reliability. Very
few mission delays/scrubs/snaffus are software in nature. Also government
software (much of it developed on Unix, BTW) also seemed to hit the mark –
excuse the pun – during Desert Storm. Note: this is not an endorsement of
such behavior.
Yes, the software is ultimatley developed by private corporations, but to
government standards, by government procedures.
You should really check out your facts before you make such Archie
Bunker-ish statments.
Disclaimer: I am not a Unix booster (or even fan). I have made a living
for 12 years largely on engineering large real-time systems for government
programs. My best freind has made a living for thee past 12 years in the
commercial sector exclusively as a Unix system programmer and designer. I
know whereof I speak.
wmc
Wayne, you should know better. I have my facts fine. You have blinders, is
all. There are over 100 million – that's MILLION, you know, the number
after 999,999, except, a HUNDRED times? Anyway, 100 MILLION pc's out there.
You tell me: How many Unix machines. 1 million? 500 thou? How many?
–Ben
Ben,
I once saw a graph of the $ volumn spent on various OS's. That's
actually a better gauge of who's using what…since you get about the same
number of users per dollar spent. At that point ( a year or two ago) UNIX
was behind MS-DOS, but the trend clearly indicated that (presuming things
keep going at the then current rates) UNIX would bypass MS-DOS within a few
years. Since then, UNIX has accelerated if anything. Contracts such as
the Air Force Computer 3 (2 Billion dollars worth) and the Treasury Dept
contract (3,500 machines at about $500,000 apiece) didn't hurt either.
Looking at a recent ad…Sun claims 9 Billion dollars in sales
(total) in 1/2 a million machines. The ad also claims that SunOS alone is
the 3rd most popular OS around.
Add all the rest of the UNIX boxes sold by other than Sun, and
you've got a pretty good idea that UNIX just may end up being the most
popular OS available.
Don
Uh… don't you think that $$$ volume is only comparable if the OS's are
all priced the same? <scratching head>
–Ben
Ben,
Isn't that what I said? Cost per user was about equal? Also,
those figures I saw weren't simply for the OS. It was for hardware and
software. UNIX itself is relatively cheap. In most cases…it's included
in the price of the machine, so you have to look at the upgrade cost. In
our case, the upgrade cost for a 100+ user machine was only $400.
Oh yea…on our '386 machine…UNIX was a separately priced item,
for the unlimited user version…the cost was in the $600 -$800 price
range. Not too bad
In fact, that machine is a good example. Cost of the machine with
no software other than UNIX was $17,000. That included the 40 Meg of RAM,
300 Meg SCSI HD, 120 Meg tape drive, ethernet card and TCP/IP software,
video, 16 port serial card and math chip. It easily handles 10 users, and
often does.
That's only $1,700 per user which is cheap for a 33 MHz '386 with
math chip if bought as a stand-alone DOS box.
Don
Ben,
When you go into a department store and buy those fancy white shirts and
ties you where to the AmiExpo, and the gal rings all this stuff through her
NCR, chances ar VERY good that she's using Unix without even knowing it!
Also, don't forget, manyof those DOS boxes are tied in to some form of
netwrk that is running on a Unix/Xenix/Zenix/ network server.
You also can't just talk number of machines since the machines you talk do
very little except keep executives making fewer fool hardy decisions 'cause
they're too busy trying to figure out why Windows don't work worth doggy
doo on their super duper 986 wiz bang clone box. Then there is the
DOS/Clone compatibility problems (works on one machine – not someone else's
…)
I don't believe DOS is the future of anything. Besides, in the business
I've been in for the past 12 years I've been around or run into Unix at all
levels. Dos is relegated to the trivial – little light word pro and
spreadsheet stuff.
wmc
GRIN. Unless, of course, you bought that shirt and tie from JC Penney's,
which at least a coupla years ago was using OS/2-based machines.
I mention this only because I knew the poor guy who had to work on it (come
to think of it, he may be reading these messages), and he may not have
recovered yet 😉 Altho OS/2 at the time was awfully beta, and it would
probably be easier to do now.
PS: yah, many MSDOS users now realize that it has little future… it's
just a matter of time.
Actually, I buy them at "Beede's mens store" in Glasgow, and they use PC's.
I'm just bored with the Argument. I hate PC's. I give anything if they'd
all fade into the woodwork. But they won't; they're here, they're staying,
Unix hasn't a chance in the major markets. Ispo Fractured.
–Ben
Kevin,
I got a good chuckle with the new PC Magazine – an article called "power
programming with DOS 5.0". A quick survey of the article showed that the
increased "power" lay in a further segmentation of the memeory map. Lets
see, there's DOS space, ROM BIOS Space, 640 base, 64K EMS Page Frames,
Extended to 1 Meg, Expanded and Dos 5.0 adds two more types!!
Microsoft figure that all those soft-in-the-head DOSites can be fooled into
thinking something is powerful just by making it overly complex.
As for OS/2, the OS stands for a rather frequently used exclamation that
comes out of most people that try to work with it – in fact your freind has
probably used it several times :^}
wmc
Chances are equal that NCR machine was running NCR Basic, with a tape based
OS. 🙂
-mark=
– via the Whapster!
The CBM Unix didn't sell because of the CBM name. CBM has done nothing to
change its game machine image. If it had marketted under a different name
(a la NewTek's Toaster System) and actually advertised, they could have
sold.
Case in point: I proposed a unix based system to my company that was
putting together tech-center learning lab that was to be multi-user,
networked wit X-windows focus. I didn't name the machine. Specs and bucks
were right. COmpany very interested until they found it was a CBM based
system. They laughed, then accepted a 486 server with 386 diskless
worksations that is a dog. Its the only way they could compete with CBM
price I quoted but the system is BAAAAD. But it aint a CBM and in there
mind that is the big selling point. wmc
At least for Governments, UNIX has NOT taken over. POSIX compliant OS's
may take over, but as no UNIX to date is POSIX compliant, and 3 non-UNIX
OS's are, that does not mean UNIX!!!! POSIX gives any OS which has it more
compatablility than UNIX. Now POSIX compliant AmigaDOS sounds like a great
posibility! This can be done by a layered aproach (POSIX calls implemented
in terms of AmigaDOS calls). Statisticaly, in multi-user systems in
government, less than 10% of the computers are using UNIX, hardly "taken
over".
Gary Dunlap : Wyrd Software
: It just works like magic!
Motorola has certified as POSIX compilant for at least one version of their
UNIX OS (I can't remember offhand whether it is 68K or 88K based). And I
know for a fact that Modcomp's REAL/IX will pass the POSIX portion of the
FIPS-151 document. I know because I worked on making it POSIX compilant.
Steve
10% is a big number of machines, though, when you're talking about
something as inefficient and duplicative as the gov't.
And it clearly indicates a trend. Don's point is probably valid.
My point is just that Unix isn't the big deal we keep hearing. The cart is
well before the quadruped at this point.
–Ben
Gary –
I guess I work at a government installation that doesn't reflect your
statistics. For multi-user systems, we are probably 90% Unix, even the DEC
equipment (I know of a microvax running VMS, but the other DEC equipment
I've seen is using Ultrix or 4.3BSD). Even our IBM plug compatible
mainframe (Amdahl) is slated to have Unix installed on it. And there are
PC compatibles and Mac's around running versions of Unix. Unix is now
being used in desktop tactical workstations aboard Navy ships.
While I don't doubt that there are many government facilities using
operating systems other than Unix, Unix is now a major player in
government, and is likely to remain that way.
Jeff Schweiger
Yea! And all software will be written in ADA too. -bill – via Whap!
Bill,
Actually I view the lack of an Ada compiler for the Amiga as a major
deficiency. Guess I'll have to settle for C or C++ in the meantime (he
says, ducking and running 🙂 ).
For large system development, its a good language, and is gaining in
non-DoD use, especially outside the US.
Jeff Schweiger
Jeff,
Likewise. I'd like to have an Ada compiler for the Amiga.
Jeff,
An independent developer was workingon an AMiga Ada compiler. He then got
hired by CBM as a direct. The Ada project was shelved.
CBM shoots the other foot…
wmc
Wayne –
Was that the effort that was to have been published by Oxxi?
Jeff Schweiger
Jeff,
If you are referring to the Ada compiler, yes.
wmc
The computer science department at my university offered Commodore a really
pretty neat deal that would have brought an Ada compiler to the amiga. The
Computer Science department arranged with Meridian software to get their
compiler generation system that allows for the rapid development of an Ada
compiler for any platform. Meridian told NSU that the software package
would be given free if we turned over the work to Meridian at the end so
that they might validate it, package it and sell it. Meridian had been
asked on a number of occassions for an Ada compiler for the amiga, but did
not feel that they could justify the resources at the time they were asked.
Since NSU was an educational institution they felt that the porting would
make excellent study and it might possibly develop into a real product that
they could sell.
Norfolk State needed a bit of financial help from Commodore to pay for
hardware and professors/students to port the compiler. CATS was contacted
as well as Commodore Business Machines (the U.S. Sales Company). CATS did
not feel that the availability of an Ada compiler would do much to improve
overall sales, and the represenatives at Commodore Business Machines never
returned the countless phone calls.
So, there was an effort made to bring an Ada compiler to the Amiga by a
university that did not have a single amiga in their computer science
department. Because of the support Commodore had shown for UNIX and the
fact that a UX was given to the university for a short time the CS
department was willing to do some work with the machine, and was in fact
excited about doing it. The mere fact that UNIX existed (and was
supported!) on the Amiga had changed the overall perception of who and what
Commodore is. Unfortunately, with the recent layoffs and folks leaving the
UNIX development team, the CS department has now abandoned the Amiga 3000UX
as a viable platform.
I believe that Meridian would still entertain the possibly of porting Ada
to the Amiga if someone would do the work. I also grant that the work
[ MORE ]
Mark –
I was aware of the Meridian-NSU effort, both from Meridian and Prof.
Harrison. The was one other change in the situation, the point of contact
on the Meridian end left Meridian, and a new person is on the job there.
What makes it worse is that there are Ada literate people working at
Commodore.
Jeff Schweiger
[ continuation ]
NSU would have been done would not have provided access to any of the
system libraries in its first release. So that would probably negate its
complete usuefulness and may be why Commodore was not interested.
I am not posting this as a 'sour grapes' story. I am posting it so that
the world knows that a serious effort was made. I am also posting this so
that the folks that think UNIX should have never been attempted by
Commodore might realize that UNIX is a viabilty stamp that changes peoples
attitudes rapidly. The UX completely changed the minds of people who
were/are in love with PCs and Suns. Had Commodore continued on the course
that was set, NSU would proably have 20 A3000UXs by now, or at least be in
the process of purchasing them.
-mark=
– via the Whapster!
Bill,
You sound like a Department of Defense type when you write about ADA.
Understand ADA is neat.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Hal;
What I know about ADA could be written on a postage stamp. That was just
an example of my 'more facetious' behaviour. It is not ADA as much as the
attitude, often see in government, that their proclamations against the
face of reality is sufficient to change reality. -bill – via Whap!
Bill,
The decree about ADA went something like this: If you're a software
developer for DOD and you want a contract, you must write in ADA. Now, the
kinds of things that they are talking about are mostly for weapons systems.
My understanding is that the control software in modern aircraft is ADA
based.
(I was Regular Army from May 1982 to November 1989. Got out as a Captain
of Field Artillery to come to school here at ol' Purdue. Entered the
Reserves as a part-time job. Was activated in support of Desert
Shield/Storm. So my understanding of what they do in the Air Force is
naturally quite limited. ADA was however used in AFATADS, a state of the
art, field artillery computer fire control/coordination system for Corps
and echelons below).
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
k6TEOD"1WD"6yTk9MT,mhMYn|BLzw[{Lrod – via Whap!
The present version of the Ada mandate is law, part of Public Law 101-511:
"Sec. 8092. Notwithstanding any other provisions of law, after June 1,
1991, where cost effective, all Department of Defense software shall be
written in the programming language Ada, in the absence of special
exemption by an official designated by the Secretary of Defense."
BTW, those exemptions, and determinations that Ada isn't cost effective,
are not easy to get. The situation is that to not use Ada, one must prove
why Ada can't be used for a particular application in the DoD.
Jeff Schweiger
Harold;
I don't really doubt that ADA has its use. Since its development is
recent it probably takes better advantage of hardware capability than some
other languages. I would imagine that the optimizer in the compiler is
quite advanced.
Even fortran and cobol still have their place (besides a museum), though I
certainly would not want to write anything else in either one. -bill – via
Whap!
I really hate it when people type Ada in all caps. ADA stands for the
American Dental Association, and I really don't think the Department of the
Defense is too concerned with dental status of the military.
-mark=
– via the Whapster!
Just as long as we're picking nits…..
It's X Window, not X windows.
:*)
Steve
X Windows, noted and logged. I did not know. The ADA vs. Ada thing should
be known by most by now I think.
Oh well. I have added X Windows to the list of things to spell right or be
considered a fool. 🙂
-mark=
– via the Whapster!
Considering that all the _major_ players have seen the need, in the past
few years, to jump onto the Unix bandwagon, and that this was done
reluctantly, would seem to indicate that your statement may not be
completely valid.
Niche indeed! You may not like Unix, with or without good reason, but it is
here with a vengeance.
-larry
No… the users are, that's all. Very vocal, very commited. But Unix isn't
"here with a vengance". Jumping on on a bandwagon doesn't inherently make
the wagon roll forward. Look around. There are many who are jumping off,
too. CBM, for one.
If all the major players do make it move forward, like they did for MSDOS,
then we're in for another period of "computer retardation". Hopefully it
won't last long.
–Ben
Well, it has lasted since 1969, so far. More and more applications are
using it, and more and more bids are going out with the requirement to run
Unix. Comparing it to MsDos is a little silly. Unix is portable (source,
and with sVr4, binary) across many platforms, and those platforms don't
have to be YAPCC (Yet Another PC Clone). CBM may or may not be 'jumping off
the bandwagon', but if they are, I doubt it's for anything but the
inability to compete (at this time) effectively in the workstation market
without going to a much faster processor.
-larry
Larry, FORTRAN has lasted even longer. Longevity is no recommendation, all
it says is that there is a large (huge, in this case) investment in old
tech, and replacement is out of the question for software and data reasons.
It's the same reason MS-DOS is still here – Darwin knows it sure isn't
performance of the OS.
And you're right – comparing Unix to MS-DOS is silly – there are so many
more machines out there running MSDOS that Unix looks like the orphan in
the corner.
MS-DOS is running on more machines than Unix is, or (I think) is ever
likely to be. Big machines are dying, small machines are taking over the
desk. I think that's the way things are going. Maybe I'll have a CRAY on my
wrist before I expire (assuming it's old age that gets me, and not severely
annoyed unix types :^).
If you are saying that CBm was aiming at the workstation market, then we've
(at last) identified a market where Unix is very heavy. A very small
market, compared to normal desktop use. I dunno – I don't think they were
aiming there- I think they were aiming at eduactional markets, business,
etc, and found out there is almost no demand in those sectors. Horsepower
wise, they're slow – they should have had an 040 long a ago (before the
Next, for instance) but CBM is CBM, IE slow as molasses and twice as sticky
to do business with. Not that an 040 is killer next to some of the advanced
CPUs out there, but it's no slug, either. HP wise, the university where
they tested it seemed to like it, anyway… so I don't think that was it. I
dunno what WAS it, but I'm guessing it was total lack of market. In
Florida, I *never once* had to look at a Unix machine. Not once. Here in
MT, where we have municipal computers and sherriff's computers and library
computers and small business computers… the only Unix system I know of is
hidden in a back room at the hospital where it mumbles it's way (slowly)
thru the payroll. PC's _everywhere_. Unix, nowhere.
I hear much ado about Unix – but I look at the mag racks in B Daltons, and
[ MORE ]
Be careful before you strap that Cray on your wrist….Cray has a heavy
investment in UNIX. :-)))
Steve
If Unix turns out to be the thing, I'll be using it. No problem. I just
don't think it will happen.
–Ben
[ continuation ]
I see about 10 mac mags, 2-3 Amiga mags, about 10-20 PC mags (more, if you
count rags like Byte), items like Midnite Engineering which are not really
computer mags per se, but entirely adress the PC market when the word
computer is spoken… and then there is one Unix rag, in SOME of the
stores. WHen I look to see where the noise is coming from, it's people
working in, or on, Unix. You, Don, and (I bet) this other fella who made
the posting are all involved in Unix.
Preaching from the cenverted is always suspect. :^) Why do you think the
MS-DOS addicts can't get it thru their thick skulls that an Amiga is
better? :^))
–Ben
Ummm….same reason that some Amiga addicts don't like UNIX? 😉
[running away….FAR away] 😉 😉
One more thing….how do you expect to answer back to you about UNIX???
People who *don't* work on it???
Well, of course. I'd like to hear someone say, "yeah, we're looking at
Unix, because it offers us these advantages over our MSDOS machines and we
really oughta… etc". I'd like to hear someone say, "yeah, not being able
to run all those MSDOS apps is no problem because Unix… ________". I'd
like to hear someone say, "We changed to Unix last year because of _______
and we are very happy we did because ______ and we have great confidence
that we'll be ok in the future because the company that is providing us
that hardware and software is ______" or "we have great confidence we'll be
ok in the future because although the company we bought from might go under
and leave us stranded with their machines, it's ok because we can
__________ and everything will be ok with little extra cost, just like we
could with our MSDOS machines."
I *never* hear stories like this. I'm enough of a pessimist to not wonder
why.
–Ben
Guess it's just cause we all (everyone here) run in different circles. From
my perspective, I've been able to see customers like JPL and Alcoa either
switch from their old systems, or just flat out buy new systems because
they want to go to UNIX.
Ben, you are wrong about MSDOS prevailing in a partly-unix shop.
We bought *four* SPARC stations *solely* to run AutoCAD. Is it unix? Is it
Sun? Or both? I don't know, but we now have 16 SPARC's running
*BRILLIANTLY* with the Loyal Staff begging for more.
Sun/unix is driving MS/DOS out the door at my place. After the Suns
arrived, the L.S. started calling Banyan Vines "Dial a Prayer."
So, to fill in your blanks:
I don't miss any MSDOS aps; the ones we used weren't so hot to start with.
I'm delighted we switched to unix; it's GREAT to have a REAL network. And I
have great confidence in the future of open systems, since any individual
vendor is for the most part irrelevant.
BTW, Sun wins on *price* as well as on performance in our environment.
Rob.
Ben;
The reason that you don't hear that is because most of the people that run
Unix systems would die laughing if anyone even pretended that MSDOS was
even in the running for the magnitude of the work that is done.
I have seen several companies move from 'other' real systems (ie:
multi-user) to Unix. A major consideration was that such a move would
allow for future expansion and reduce the dependency upon any one
particular type/brand of computer.
Indeed may of the 'favorite' DOS box type programs either have Unix
versions or have more 'user friendly' Unix equivalents. -bill – via Whap!
Ben! I am almost shocked at one of your statements.
That sounds about like the "Union of Concerned 'Scientists'" saying that no
one should listen to ANS (American Nuclear Society) personnel or people
working in the Nuclear Industry. What balderdash! I will admit that one
needs to consider possible bias and self-interest but…….
-bill – via Whap!
Bill, sit back and take it in context.
The statement was made that Unix was the operating system of the decade and
the next century.
That was the context I made the remark of taking things at face value. I
don't think that you can take a nuclear scientist's opinion either, if he
says "This is the decade of the fission reactor", which is a lose analogy
to "This is the decade of the Unix OS". Fusion may be just around the
corner, and in the meantime, there is always coal and oil. :^)
Listening is one thing – assuming that the statements are gospel is
another; and people have a long history of enlightened self-interest
guiding their statements (turn on any evangelist and see :^)
I agree that what you said is balderdash. But what you said, isn't what *I*
said. :^)
–Ben
Point accepted 🙂 -bill – via Whap!
Ben,
> Longevity is no recommendation…
True, but I was answering a specific statement you made, and my answer to
it was not aimed at justification through longevity.
> MS-DOS is running on more machines than Unix is…
No it isn't. Well, it is if you count every individual machine. What I
meant (and perhaps didn't make clear) was that Unix is running on more
varied platforms than MsDos (other than in software emulations, etc.). Unix
is a real OS, as compared with the primitive program loader that is MsDos
(and you can say the same thing about Amigados).
> small machines are taking over the desk.
True, and many of them run Unix, and put more power on the desk than any
MsDos machine ever will. While the industry has a fairly specific
definition of what constitutes a 'workstation, you will find that many
vendors are claiming workstationness, and are aiming their software toward
providing the benefits of same. The demand for workstations is tremendous,
whether or not they are really workstations or not. In North America, CBM
has traditionally been a supplier of _HOME_ computers, and Unix does not
fit well into this niche. It requires a fair amount of knowledge to keep it
running smoothly and efficiently; far more knowledge than the average home
user is willing to acquire.
As for magazines… there are lots, whether or not they appear in B
Daltons. This is analogous to the many other special-interest publications
that don't appear in the average bookstores.
[more]
As for market, well, let's just say that you and I live in completely
different worlds. I see businesses switching to Unix, small and large. I
see small, powerful servers and small powerful workstations taking the
place of large mainframes in any number of places, military, educational,
engineering, publishing, manufacturing, and more. Before I started looking
at Unix machines, I didn't know anything about them either. 🙂
-larry
The only thing I need to know for this thesis, is that there are more MSDOS
machines than Unix machines, and things are likely to stay that way. If I
was looking to get into a Unix machine (I'm not, thanks), I'd be all over
them busy as can be. But *my* point was simply that the statement Unix is
the OS of the decade and the coming century is wrong. I've been able to
make the point repeatedly, without any contentions that needed serious
consideration.
I started out by saying I don't like it, but I'm entitled. I also said
(repeatedly) that it's more powerful than MSDOS and even better. None of
which changes my point, which is it's not going to take over the market,
and in fact, hasn't a chance in heck of doing so.
You see things from one of the tops of the heap in the Unix world. I see
them from a small hill of the Amiga world, and the few (say 50 or so)
businesses I know here in town (of which you can count a lot more than 50
PC's, and NO unix machines – I've been 'round asking in the last day or so,
as this has been on my mind). The hospital does have that one doing
payroll, as I mentioned previously. The MSDOS user sees things from
anywhere but Unix. I don't see it chaninging.
If, and it's a big if, Unix takes over, it has far, far to go before anyone
notices it in your average business.
We're not talking the one or two plants per 5k people, we're talking the
several HUNDRED businesses per 5k people.
–Ben
Ben,
I think the number of Unix installations and users would not support your
contentions. It is nearly impossible to find a univerity that is not
running Unix; all major manufacturing/aerospace and architectural design
firms that require high powered CAD and CAM run on Unix. A version or
Unix, POSIX, is the Space station operating system. All space station
software is being developed on Apollo worksations that have both the
berkeley and 5.4 versions of Unix on them. You cannot be a serious
software engineer, network user, or large scale system user without running
into Unix. It is not a niche operating system. Between it and VMS, IBM's
OS/MVS and variants are on the decline. Even IBM supports unix in micro,
mini and mainframe versions now.
BTW – I ain't necessarily a Unix fan, but to survive in the industry I've
run into it too many times in DTP, CAD, CAM, network, Communications, etc.
to be able to maintain ignorance of it and make a living – My observation
is that it is everywhere.
In that light, I think CBM needs to get the Amiga shrink wrap Unix on the
shelfs – NOW.
wmc
CBM will announce the shrink wrapped version at UNIX EXPO.
Steve
Steve-
That is GREAT news. Does that include new ROMS to boot to UNIX or ADos? Do
you think GVP accelerators will support it or just A3000s? (GVP has always
claimed unix compatibility with the correct autoboot roms …)
wmc
I don't have any other details about what's included in the package, and
I'm reallllly curious to see how GVP handles it too. Time will tell!
Steve
According to Digital Equipment Corp., they sell more VMS based machines
than UNIX. UNIX is hot in the workstation arena, but VMS is the choice of
business running DEC VAX machines. I might add that Digital has not been
doing well selling its VMS-based workstations nor UNIX-based but the latter
has been a price/performance ratio problem rather than the lack of
acceptance as is the case of the VMS-based workstation.
UNIX has not provided the best environment for the MIS departments. There
are nice products _now_ for UNIX-based MIS work, but that was not so just
five years ago. There is now a large entrenched group of businesses that
depend on the flexibility of VMS and its redunancy provided by a clustered
environment. Will UNIX take over MIS? I don't think so, as most MIS
departments tend to be less interested in OS's and more interested in
database managers, ease of use, and the availability of COBOL compilers.
Certainly ULTRIX does not get the same support from within Digital as does
VMS. Digital is working hard on producing a POSIX complient version of
VMS. They believe the future to be in POSIX complience with VMS so I don't
think Digital thinks the future is UNIX for them.
I might also add that there are not many people who are running IBM's
IX/System product, and those that I have met that are running it are less
than pleased with the implementation. UNIX is not something IBM seems to
want to get involved with either. I believe that Sun as well as the other
workstation vendors are currently the big (only?) major UNIX players today.
Note, no where in this posting did I say the MIS departments of the world
were intelligent.
-mark= VAX Cluster System Manager representing/but not working for VMS.
🙂
– via the Whapster!
Mark,
Not that it matters, but IBM is the driving force behind the Open
Software Foundation. OSF includes some big players…I believe that DEC is
among them. In fact, some of the players have joined both OSF and UNIX
Int'l. Between those 2, almost every computer manufacturer is
represented….and many software companies too.
Don
I am curioius as to whether they are just represented or are they really a
'driving' force?
-mark=
– via the Whapster!
Mark,
Very definately a driving force…and a founding force. The
standard story behind it is that IBM feared loosing too much market share
in an open market, so decided to get their fingers into that market also.
They didn't want to join AT&T since they wanted the "standard" to be an IBM
"standard"…thus they formed OSF and have been the driving force behind
it.
For example, they "forced" Motif (a pure IBM product) to be the
standard GUI for OSF's version of UNIX and also it appears that OSF will be
almost pure AIX…which is IBM's version of UNIX.
The parts I know to be true…is the part about Motif and the
basing on AIX, other than that…I'm merely repeating what's been widely
reported in the trade magazines.
Don
Mark,
I'm sure that most all of what you say is right BUT….UNIX IS HOT!
Small business installations are often best implemented with a UNIX machine
and a few terminals. This is both from a cost and performance standpoint.
LANs are more expensive and slower at this level. IBM is definitely
interested in UNIX and their RS6000 is very popular for the type of
installations I have mentioned.
….Mark 74335,1677
Mark (great name by the way…)
Are you sure about that? Are these machines PC based? If so, what UNIX is
being run? I see far more systems set up with a Novell network than I do
UNIX based PC clones in offices.
-mark=
– via the Whapster!
MARK,
I'm not saying that there are more UNIX installations out there. I am
saying that the tide is turning. The recession has made businesses take a
HARD look before buying a system and a UNIX system could be far less
expensive than a Novell LAN. The people I have dealt with are using 386
and 486 machines with SCO or Interactive or have gone one step up with
RS6000 IBM machines. Just compare the cost of adding dumb terminals versus
286/386/486 workstations with Ethernet cards. Netware isn't cheap either.
The choice was plain for me when I went to buy a 10 user system and this
was echoed by the vendors I contacted. Networks may be hip but they are
not the only answer. Certainly they are right for plenty of situations but
not for mine.
….Mark 74335,1677
Mark,
I recently had the misfortune to have to run Interleaf on a VMS vax
workstation. It took a full ten minutes to open a dataflow diagram with
maybe 6 bubbles. On an Apollo it took 10 seconds. I've been curious if it
is the VMs vs UNIX Interleaf or just VMS vs Unix for screen drawing in
hi-res monochrome. Anyway, it didn't make me want to buy VAX workstations.
wmc
It is my understanding that most people who buy VMS workstations buy far
too little memory causing the performance problem that you describe. It is
my belief that in the general case VMS can perform better than UNIX.
Obvisouly I am talking about a UNIX on a VAX vs. VMS on the same VAX.
-mark=
– via the Whapster!
Mark,
That would make sense since DEC has more control over tuning VAXStation
hardware and VMS to each other than it does over Unix. However, every time
I've seen VAX's bog down on a package that is available in both
environments, it is a CAD, DTP or heavily graphics oriented program. I was
wondering if graphics is an inherent weakness in VMS or as you mentioned,
everyone underconfigures VAXs. Gee, does that mean Unix Sys Admins are
smarter than Vax Sys Admins – or at least better at getting money out of
the company! :^}
wmc
I know that UNIX System Administrators _think_ they are smarter, but only
in rare cases they actually are. Most UNIX System Administrators have a
ego problem that is well beyond that of the mortal man. I do admit that
UNIX machines require more "hands on" maintenance than a VMS machine does.
They consider that "good" while I consider it "bad design". So are they
"smarter"? I am not sure. Are they working harder then me? Definately.
🙂
On the performance question. Since Digital can control the development of
the hardware and VMS they can build machines to compliment the features of
the OS as well as the hardware. I hold SMP (Symetrical Multi-Processing)
and VAX cluster as my examples.
As far as graphics goes. I have little experience as my environment
requires use of text-based applications, with a little bit of REGIS
graphics thrown in for good measure. I would suspect that graphics would
behave badly under VMS given its nature.
-mark=
– via the Whapster!
You haven't met some of the VMS sysadmins I have. I'd say that *both*
camps have their own hard-headed "rather die than switch" sysadmins.
And just to toss a log onto the flames, having been a sysadmin for both, I
can't say I like VMS very much. 🙂
Steve
I don't buy it – but I'll wait and see.
–Ben
No need to preach to the converted. When I talked to Jeff Stilley, the
Vice President of Commodore US Sales I told him exactly the same thing. He
was busy telling me that UNIX was a "completed" product and that there was
really no need for further development. Now that the last three UNIX
people have quit from Commodore, I am watching eagerly to see if Commodore
rehires UNIX people. If they don't, color that product dead. At least in
my professional opinion. I can not and will not advise my university or
others to purchase a product that there is no active development in. I
might consider it if UNIX was a mere application program, but an operating
systme requires full time support and full time continued development.
I wrote an article two years ago called "Can Commodore reach for the Sun?"
and it was my thoughts about Commodore entering the UNIX marketplace. It
has been my belief all along that Commodore would never grab the UNIX
marketplace, but I felt that if they kept the price vs. power in mind that
they had a shot in capturing the low-end market. I felt that the low-end
market would remain open because folks like Sun want to sell big-mamma UNIX
boxes. Commodore, probably through no fault of their own, priced the UX
too high. The question always came to be "Why should I buy an Amiga
instead of a Sun?".
Anyway, I am hoping that Commodore is not developing a new machine with a
new operating system. I think it would be a terrible mistake, and honestly
I don't think they could afford to make that one.
-mark=
– via Whap!
Mark,
Explain to me why (if Unix is so wonderful) why every company needs to
do active development on it.
Huh?
Who said I said I thought UNIX was "wonderful"? I have said and do say
that UNIX provides a viability stamp for Commodore and Amiga that removes
from most peoples eyes the idea that Commoodre can only produce a game
machine.
-mark=
– via the Whapster!
Mark,
I went back to see what I said, that cause this reply. The line I was
referring to said something about …could not recommend any Unix for which
ongoing support wasn't being done.
What I was saying is that I don't believe in buying machines with an
operating system that the company is not actively supporting. In my
opinion, Commodore provides no support for its UNIX product. To go a step
further I would say that I would also not purchase a machine with an
operating system that is considered "finished" by the manufacturer. I have
yet to see a finished OS.
-mark=
– via the Whapster!
By on-going support, do you mean active development, customer support, or
both?
Steve
Steve,
I don't know….I was questioning Mark. Mayhaps you should direct the
question to him.
I mean both Steve.
A lack of customer support is a primary concern. I believe there may yet
be a single member in CATS that handles UNIX customer support. That is,
if you believe that a single person can handle 'customer support' for the
entire UNIX market.
Active Development helps to insure that the product keeps up with the
times, and further demonstrates the company's commitment to the product. In
Commodore's case it is important to remind people of the commitment on a
daily basis with continued efforts in what may be a non-profit item in the
short term, but may pay off in crediability and dollars later.
I believe that Commodore killed UNIX before it ever started.
-mark=
– via the Whapster!
Mark,
Amiga Unix is supported by Honywell/Bull tech support. C= apperently
has a contract with them. There are more than one tech persons there with
A3000UXs running on their desks when you call. They are both courteous and
knowledgable. They are also responsible for identifying bugs for repair.
Currently SVR4 release 2.0 is in late beta. Yes, there is support. The
failing point is marketing which is almost non-existent. I can't even seem
to get ahold of a unit for evaluation. 50 (no exaggeration) phone calls
later I have nothing more than vague promises. Rats, looks like another
I486 goes to work.
….Mark 74335,1677
Hmmm… I missed that announcement, and indeed I find it comforting if
true.
However, this I have heard, UNIX 2.02 was released to marketing about a
month ago. If you are right and it is still in "late beta" then it is
going to be there a long time I'll wager. Hard to do much UNIX development
without any UNIX programmers. I guess they could pull the three that are
supposed to have been transfered to AmigaDOS work back to "finish the
operating system".
-mark=
– via the Whapster!
At one point, I had heard that Commodore was hiring an outside consulting
firm to handle customer support for UNIX. I have no idea what the state of
that is, but I said at the time (and still believe) that it's a very good
idea. Why burden the engineers with customer support stuff? It's really
great having a "firewall" to stand behind, and only get the tough problems
when they come in. CS can handle most of the other questions.
It's unfortunate that the CBM engineers that worked on UNIX were all put on
to other projects or let go. Very difficult to understand.
I wonder what people's reactions would have been if the word never got out
that Commodore isn't actively working on an upgrade? There are more than
a few Amiga companies that sell Amiga programs that haven't been updated in
a long time (and may never be). We never hear people yelling and screaming
that they won't buy the product because the company isn't working on an
upgrade.
From the things I've seen about the Amiga UNIX 2.0x upgrade, the only major
thing thatR.';issing is ABI support. I'm curious about what other things
are people expecting to be improved or added? Don't say anything like "run
Amiga programs under UNIX", because that would require a *lot* of work, and
it would be a major, major project. I think it would not be in Commodore's
interest to dump money into it….not for the return they would get.
Steve
Steve,
I thought you would give me a bit more crediability! 🙂 I would not
suggest that Commoodre should attempt to make AmigaDOS programs run under
UNIX. To me, that would be a waste of time, and likely to be un-doable.
I do think though there could be some integration between the UNIX and
AmigaDOS environment. I would like to see a AmigaDOS filesystem for UNIX
so that while under UNIX you could read and write AmigaDOS directories and
files, and while under AmigaDOS you could read UNIX filesystems.
I would like to see solid support libraries created for the custom chips.
After what good is an amiga without access to the custom chips? I know
there has been _some_ work done there, but I am willing to wager it is not
complete. I also think that ABI compliance is a _must do_.
I think things might be better today in the UNIX marketplace if the word
had not gotten out. However, to attempt to keep quiet a mass-layoff is to
waste ones time. People who get laid off tend to be "unhappy".
Of course, Amiga people are different in some respects than other computer
users. We are all so concerned that Commodore "makes it" that we keep a
microscope on their every movement, arm chair quarterbacking all the way. I
am of course a part of that crowd as most of us are. In Harold's case
however, I think we can all agree that while we arm-chair quaterback he is
arm-chair coaching. 🙂
Personally I like to quarterback the plays after the play has been made.
Makes my vision all the more clearer. 🙂 🙂 🙂
-mark=
– via the Whapster!
Do you know what the SPECmark rating of the 3000UX is? Just
printed in the magazine UNIX Review (Oct 91 issue): rating is 1.7
In fact, the 3000UX came in a very poor 6th when compared against
the following:
Nextstation, HP9000/425e, DECstation 3100, Sun-4/20 and a no name
clone all running various versions of UNIX.
While I realize (and the review stated) that the version of UNIX
being run was a beta version….the rating still stands as what readers of
UNIX Review will "know" about the 3000UX until they do a new review of it
(not likely…but possible).
Don
Don,
Revision 2.0 really is faster. No one would dispute that a 68040 based
AmigaUX would be less expensive to build and much faster. However, I am
uncertain whether the 68020 code would run without exception on the 68040.
Hal Grigdesby Krannert Graduate School of Management Purdue University
– via Whap!
Hal,
Faster is relative. The ratings given the beta UNIX on the 3000
were as follows in comparison to the other machines:
Machine Dhrystones Whetstones MFLOPS SPECmarks
3000 3144 1969 0.220 1.7
HP9000 58,139 10,858 1.687 10.3
Clone 25,906 5580 0.992 7.9
DEC 3100 26,086 11,601 3.518 11.1
Sun 4/20 19,867 7042 1.757 8.9
Next 24,793 4167 1.755 n/a
They also have what they call a Graphstone and Khornerstone test,
the Graphstone test obviously tests graphics, the Khornerstone is a series
of 21 tests of CPU, FPU, disk activity, etc. and these are their results:
Machine Graphstones Khornerstones
$$/Khornerstone
3000 6183 4671 $1.60
HP9000 26,118 42,753 $0.45
Clone 7120 11,933 $0.58
DEC 3100 19,520 17,500 $0.51
Sun 4/20 9137 11,459 $0.83
Next n/a 17.200 $0.58
As you can see…the 3000UX, even if twice as fast, really isn't
even in the race. It's quite obvious that the hardware capability isn't
being used by UNIX, and whoever "did that" should be ashamed of themselves.
Don
I would further point out (for Ben's benefit) that the 4/20 is the lowest
cost machine produced by Sun.
-larry
Don,
There needs to be an 040 Amiga platform. Time will tell.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Harold,
All else being equal…putting an '040 in the 3000 would about
double those numbers. And they'd still be quite poor.
Don
What I never have understood from the very beginning the decision not to
use the custom chip set to 'help out'. Certainly xwindows could use all
the help it can get, and performance wise it seems to me the more chips
doing things the better.
The thing that seperates the Amiga from the mac is the custom chips in the
hardware. I don't understand why they were left out of the design of the
UNIX product for the Amiga.
-mark=
– via the Whapster!
The reason they didn't use the custom chip set was posted a while back on
the internet by one of the Commodore UNIX guys:
"The answer is simply that the 68030 can do the job faster and better than
the blitter chip. You must remember the blitter chip was made to enhance a
7mhz 68000 with bus accesses of only 16-bits wide."
Now wait 🙂 I know what you're going to say. That it would make more
sense to have the blitter, say, scroll a section of video memory while the
cpu was "off doing something else". And I'd agree. Except…
1. It looks (and is) slower to the user viewing it.
2. You could have scenarios where the cpu is actually idling while the
blitter finishes a blit that the cpu could've done faster itself.
Obviously some smart combination of cpu/blitter usage would be best, but
personally I suspect that it just wasn't worth the effort to implement the
blitter for the uncommon (to X) logic/shift cases where it might be faster.
kevin
I guess you have a point there. Actually I guess the UNIX folks had a
point. Could it be implied by that decision that Commodore does not intend
to release an enhanced blitter? If an enhanced version of the blitter was
going to be released one day, seems to me (hate to use those words now with
Andy hanging around :-)) that it would have been better to accept the
performance problems that using the blitter might give, so that they could
immediately take advantage of the better blitter technology later.
Oh well. I don't think it makes a diff now.
-mark=
– via the Whapster!
"Design"? "Unix"?? Those two words shouldn't appear in the same sentence
without a negative.
Oh, please. Contrary to apparently popular belief, there are companies
that produce UNIX products that are carefully designed. I, and the rest
of my group work d*mn hard on our designs and products. If it weren't for
the *quality* of our high performance realtime UNIX products, places like
NASA and JPL (just to name two) wouldn't have bought our systems.
While I'm sure it wasn't your intention, I find blanket statements that
imply people involved in UNIX development can't design things pretty
insulting. 🙁
Steve
Don,
Given your info on the 040, the UX and the marketplace… what should C= do
with the UX or UNIX?
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Hal,
My opinion…for what it's worth…is that they ought to add the
'040, and recode the kernel for higher efficiency.
I'm not saying their current (2.0) kernel is inefficient, but based
on the figures in the test of the Beta version….it certainly could have
used some tweaking
Don
Don,
Thy OUGHT to get a new compiler. Dhrystones has NOTHING to do with any
kernel that is running….it's pure compute bound stuff.
Don,
Was the NeXT cited an 040? If so, looks like The UX would be more
competitive with an 040 processor.
The most telling thing in your data was that Commodore's truly has the
wrong price for the UX. A slower UX is okay if the price is a bargain.
Still think the price for a UXG should be around $2500 for students. (A
UXG = ULowell Card + 8MB Ram + 200 MB HD).
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Hal,
Yes…the machine tested was the Nextstation…a 25MHz 68040 box.
Don
Don,
If the NeXT in your stats had an 040/882, then dare to postulate why the
Amiga would still not be as fast. Inherent differences between Mach and
AT&T System V Release 4? Swap space? Ram?
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Hal,
No, swap space should have played no part at all…the 3000 UX used
in the test had 9 Meg of RAM. The NeXTstation had 16 Meg of RAM. Both
should have been sufficient. Neither used and external processor cache.
Not sure how much internal cache the '030 has, but the '040 has an 8k (4k
instruction, 4k data) internal cache.
I seriously doubt that the differences between Mach and SVR4 are
enought to account for the differences.
However, I do *suspect* that since the NeXT is about the only
machine using Mach…and there's no ABI compliance requirement…that the
Mach running on the NeXT is specifically tuned to the hardware.
More efficient compilers, kernel, libraries, etc. can all make a
big difference. For example, the HP9000/425…also running a 25MHz 68040
(same as the NeXTStation) about doubled most of the CPU intensive
benchmarks over the NeXTStation.
If Commodore could make their hardware/software as efficient HP's,
they may have looked significantly better.
Don
Don,
Care to theorize why C= wrote AMIX the way they did (ABI compliant, but not
optimized for the hardware)?
Again, I enjoy using the UXG version. Got to find some X-window stuff to
run on it other than XV, a GIF viewer and some other provided demos.
Hal Grigdesby, Krannert Graduate School of Management, Purdue University
– via Whap!
Hal,
Oh…forgot to mention this:
Performance does come at a price though:
The base price of the HP is $5,490 but that's for an 8 Meg RAM,
diskless, monochrome system. Moving to color ups the price to $7,490 and
adding another 8 Meg of RAM and a 400 Meg HD doubles that to $15,490.
The Amiga on the other hand cost $6,398 for a 9 Meg RAM, 200 Meg
HD, color monitor, and Ethernet card.
Don
Don,
I've been tearing around trying to find that issue, but I haven't been able
to so far. Was the entire SPECmark published, or only the final rating?
Steve
Steve,
Just the final rating….but I've left some of the other ratings in
another message to Harold.
Don
>Just the final rating….
Interesting….the folks that put the SPECmark benchmark together don't
really want that done. They want all the parts of the SPECmark published
because it shows the overall story more clearly. It would be interesting
to see how the other tests did. It'd be even better to see how 2.02 does.
I think bashing Commodore over some of their recent decisions et al re
their UNIX offering is inappropriate. Things take time. Plans are in
effect. Let the wave ride.
Hey!
I am _not_ Commodore bashing! I am not rumor mongering. I am stating
opinion on what their actions. That is what message bases are about is it
not?
Those who know me know that I am a hard core Amiga/Commodore flag waver.
Ask Ben! He'll tell ya!
I said in my last message that I am watching to see what Commodore does
with their UNIX programmer shortage.
-mark=
– via the Whapster!
I think that commenting on what CBM has or has not yet decided to do with
technology they haven't annouced in a marketplace that isn't here yet under
economic and social conditions we're not sure of is silly.
I do think CBM is a pretty poor excuse for a computer company. I think the
technical types there are top notch. I hope they do well, because if they
do, we have some chance. Other than that, "No Comment". :^)
–Ben
Dear Ben,
Thank you for your comments. I am not a Commodore basher. The company has
weathered some threats to its existence. My aim is only to help them
examine what to do with this new technology from all angles.
– via Whap!