perceived forced conversion from cp/m to ms-dos in late 80's
belin...@my-deja.com
conversion from cp/m to ms-dos in the late 80's?
In one user's group I was part of in S.F., we welcomed
the dos people. Then, we started a second newsletter FL
in addition to FH (the cp/m newsletter). After a while,
the dos people took away the FH and went to a dos newsletter.
At that time, I was using an Osborne-1 and Xerox 820. Many
great developments were acheived on both machines. But, the
dos people made people who used those machines and their
programs feel inferior... Did anyone else have this happen?
To avoid being called inferior, I finally bought my first dos
machine in 1990 -- not because I liked it, but because it was
the only I could the management committee to listen to what
I have to say...
To me, developing the idea is more important than what machine
you use... What do you think?
I would be happy to share stories and support with others.
Sent via Deja.com
http://www.deja.com/
Mike Spencer
belin...@my-deja.com wrote:
> At that time, I was using an Osborne-1 and Xerox 820. Many
> great developments were acheived on both machines. But, the
> dos people made people who used those machines and their
> programs feel inferior... Did anyone else have this happen?
I used an Osborne 1 from '87 to '94. I collected up nine of them and
supported two other people who wrote their master's theses on them.
They never got any flak about it. I used two. With an external
monitor and a homemade termcap entry, they were fine to dial into
Unix and VMS. Well, as fine as 2400bps can be.
> ...the dos people made people who used those machines and their
> programs feel inferior... Did anyone else have this happen?
I joined the First Osborne Group (FOG) which about that time was
running down. They found it impossible to keep up the really great
newsletter and farmed it out to an outfit in Iowa, which published
somthing called "Computer HELP". What a bummer. The Computer HELP
guys claimed copyright for any submissions and filled their little rag
with PC and MS-DOS promotion for their biz.
> To avoid being called inferior, I finally bought my first dos
> machine in 1990...
Well, mostly people either thought my O1 was boring or that it was
really cool that I was using an antique they'd never heard of. One
guy stopped me in the street when I was lugging the thing to tell me
he'd had one. A friend visiting from MIT had a fit of giggle logging
into his MIT Unix account from this weird little box in a farm house
in Canada.
--
Mike Spencer Nova Scotia, Canada
mspenb...@tallswhistleships.ca
(Remove bells and whistles)
B W Spoor
>
> I was wondering does anyone else feel a perceived
> conversion from cp/m to ms-dos in the late 80's?
>
>
>
>
I felt a bit the same way, but my conversion was forced for commercial
reasons.
In my jobs I worked on various early micro-computer systems from 8-bit
8080/Z80 machines (pre CP/M) until I worked for a company selling ICL
PCs (and ICL DRS-20 8-bit network systems) with custom written software
(I programmed them, initially in MS-Basic, then Micro Focus Cobol).
These ICL machines were solidly built business PCs running CP/M, MP/M
and later CCP/M (which became Concurrent DOS). All PC models had serial
ports for dumb terminal(s) and printer(s), the 8-bit PC specs were:-
8085, 64K, twin 720Kb(?) 5 1/4" floppies with CP/M
8085, 64K, 5Mb HDD and a 5 1/4" floppies with CP/M
8085, 256K, 10Mb HDD and a 5 1/4" floppies with MP/M
(this supported up to 3 terminals and 1 printer and could be
upgraded to 512K and 8 ports - IIRC max 6 users)
The 16-bit PC models ran CCP/M and had 6 serial ports as standard, 4
users & 2 printers and were concurrent with the IBM XT and AT.
We had to move to the 'PC' and MS-DOS because of commercial pressure -
customers wanted IBM style PCs that wouldn't talk to each other despite
the fact that you could run multi-user applications on the ICL machines
- most offices do need to share information!!
We didn't ditch the ICL machines, but ran then alongside the IBM
machines and as we programmed in Cobol (these were business
applications), the same source code worked on both systems, we just
needed both versions of the compiler and porting one of the applications
to an IBM AIX system was a doddle as the MF compiler was also available.
Try doing that today with VBasic or similar proprietory languages -
maybe JAVA will solve this.
Performance, even on the 8-bit MP/M machine, wasn't bad for the time; to
do better they would have had to spend many times the cost for a 'mini'
system - out of the reach of most small businesses both financially and
technically.
I still have a (later) model ICL PC in working condition recovered from
a client's attic, after cleaning several years dust from it, it fired up
first time - they were well engineered (spec 80286, 2Mb, 50Mb HDD, 5 1/4
floppy, 6 serial ports, 2 screens, a printer and software).
I also recovered a working Wyse PC+ (IBM clone) of a similar age (spec
8088, 640Kb, 20Mb HDD, 5 1/4" 360 Kb floppy, green MDA screen, MS-DOS
3.2 and software that will run on it) from another client.
My main memory is that everything (hardware AND software) was that much
more reliable, no falling over with silly untraceable/repeatable errors
that we now get with Windoze.
-----------------------------------------------------------
Brian W Spoor MBCS
Chartered Information Systems Practitioner
Friday Computer Services Phone: +44-(0)1803 852625
bws...@fcs.eu.com Fax: +44-(0)1803 854926
-----------------------------------------------------------
Allodoxaphobia
>
>Try doing that today with VBasic or similar proprietory languages -
>maybe JAVA will solve this.
Well, there's REXX and Pascal -- sort of from both ends
of the 'spectrum'. You *don't* have to jump when Wee Willy Gates
says "Jump!".
We ran 3 Sanyo MBC-1000's in our consulting business here.
They were even strapped together with (I forget the name) the
precursor to the DOS version of The $25 LAN.
I still have the 3 Sanyos collecting dust in my Spider-Infested
Museum -- but, since they were so proprietary, I'm sure they are
not worth saving.
Jonesy
--
| Marvin L Jones | jonz | W3DHJ | OS/2
| Gunnison, Colorado | @ | Jonesy | linux __
| 7,703' -- 2,345m | frontier.net | DM68mn SK
B W Spoor
>
> On Thu, 01 Feb 2001 10:22:07 GMT, B W Spoor scribbled:
> >
> >Try doing that today with VBasic or similar proprietory languages -
> >maybe JAVA will solve this.
>
> Well, there's REXX and Pascal -- sort of from both ends
> of the 'spectrum'. You *don't* have to jump when Wee Willy Gates
> says "Jump!".
>
>
on small systems.
Never came across REXX, but I have used Pascal in the past although I
never really liked it.
Steve O'Hara-Smith
> I was wondering does anyone else feel a perceived
> conversion from cp/m to ms-dos in the late 80's?
About that time I was taking most clients from CPM and MP/M to Xenix
and UNIX route on the grounds that they were used to a multi user environment
and found MSDOS too crude.
Anne & Lynn Wheeler
> That's why I _still_ program in Cobol, despite it not being fashionable
> on small systems.
>
> Never came across REXX, but I have used Pascal in the past although I
> never really liked it.
in the past decade I had a job to port a 50,000 line pascal
application from one platform to another. my impression was that
majority of pascal implementations haven't been targeted at large
complex applications; majority of them much more targeted as teaching
tool &/or simple PC applications (i.e. I had done quite a bit of
programming in the early '80s with turbo pascal on ibm/pc and got into
some mainframe pascal in the mid-80s). For one platform I was
constantly submitting bug reports and talking to the pascal support
team ... in what I thot was going to be a simple & straight-forward
port.
REXX i first used in the very late '70s when it was still "REX" and
before the name change and release as a product.
randoms refs:
http://www.garlic.com/~lynn/94.html#11
http://www.garlic.com/~lynn/94.html#22
http://www.garlic.com/~lynn/2000b.html#29
http://www.garlic.com/~lynn/2000c.html#41
--
Anne & Lynn Wheeler | ly...@garlic.com - http://www.garlic.com/~lynn/
Eric Chomko
: I was wondering does anyone else feel a perceived
: conversion from cp/m to ms-dos in the late 80's?
CP/M was pretty much dead by the mid-80s as I recall. It seemed
like the time when the Mac OS (v. 6.0.x) was out that the DOS
people that saw a Mac realized that DOS and command line OS were
old fashioned. In any event, Gates realized it too, and DOS
strived to then be Windows. It was at THAT time that CP/M died,
IMO.
: In one user's group I was part of in S.F., we welcomed
: the dos people. Then, we started a second newsletter FL
: in addition to FH (the cp/m newsletter). After a while,
: the dos people took away the FH and went to a dos newsletter.
Dos people/shmos people, let them have it!
: At that time, I was using an Osborne-1 and Xerox 820. Many
: great developments were acheived on both machines. But, the
: dos people made people who used those machines and their
: programs feel inferior... Did anyone else have this happen?
Nope! I used a microcomputer OS that was NOT CP/M (DOS68 by
Smoke Signal Broadcasting), and fought with the CP/M people that
felt CP/M was superior. Actually, the debate focused more on the
Motorola 6800/6809 vs. the Intel 8080/Zilog Z-80 (which used CP/M).
In any event, I looked at the thrashing the CP/M people got when Dos
took over and then felt pity for them. All the while I became a Mac
person of sorts while, like yourself, used Dos out of necessity.
Today, I'm more interested in CP/M and OS/9 from a retrocomputing
perspective, and still use Mac OS and DOS out of necessity.
: To avoid being called inferior, I finally bought my first dos
: machine in 1990 -- not because I liked it, but because it was
: the only I could the management committee to listen to what
: I have to say...
You should have gone with a Mac.At least that way you would have been
superior for at least 5 years.
: To me, developing the idea is more important than what machine
: you use... What do you think?
Can be either/or or both. Certainly, an Apple I computer is more
inetresting than any application that you would develop on it. How far
would you go to develop on it in the first place? And surely generating
a text document that you need is more important than which machine you
did it on.
: I would be happy to share stories and support with others.
Eric
: Sent via Deja.com
: http://www.deja.com/
Charlie Gibbs
writes:
>We had to move to the 'PC' and MS-DOS because of commercial pressure -
>customers wanted IBM style PCs that wouldn't talk to each other despite
>the fact that you could run multi-user applications on the ICL machines
>- most offices do need to share information!!
Our telephone management software was born on a two-processor
TurboDOS box, had a brief flirtation with a borrowed MP/M box,
then finally settled (sank?) into MS-DOS (although I eagerly
did a Unix port when the opportunity arose).
>We didn't ditch the ICL machines, but ran then alongside the
>IBM machines and as we programmed in Cobol (these were business
>applications), the same source code worked on both systems, we
>just needed both versions of the compiler and porting one of the
>applications to an IBM AIX system was a doddle as the MF compiler
>was also available.
>
>Try doing that today with VBasic or similar proprietory languages -
>maybe JAVA will solve this.
Not if Microsoft can help it. Thank $DEITY Sun had the foresight
to nail down the Java licencing agreement to the point where they
could get a judge to force M$ to back down on their attempts to
turn Java into another proprietary language.
>My main memory is that everything (hardware AND software) was that much
>more reliable, no falling over with silly untraceable/repeatable errors
>that we now get with Windoze.
It took a while for Microsoft to groom a generation of users who
would consider frequent crashes acceptable.
--
cgi...@sky.bus.com (Charlie Gibbs)
Remove the first period after the "at" sign to reply.
Michael Wojcik
> > On Thu, 01 Feb 2001 10:22:07 GMT, B W Spoor scribbled:
> > > [re portability of Micro Focus COBOL programs]
> > >Try doing that today with VBasic or similar proprietory languages -
> > >maybe JAVA will solve this.
> > Well, there's REXX and Pascal -- sort of from both ends
> > of the 'spectrum'. You *don't* have to jump when Wee Willy Gates
> > says "Jump!".
> That's why I _still_ program in Cobol, despite it not being fashionable
> on small systems.
With MF COBOL, you can also compile source into a p-code format that
will work with the various MF COBOL runtimes, like UCSD Pascal or
Java. No recompilation necessary. And if you want better
performance, you can compile the p-code modules into machine code,
and either run those under the runtime or link them with the runtime
into standalone executables. Newer versions of MF COBOL for Unix
also support compiling COBOL modules into shared objects. I'm no
fan of the language itself, but the available tools (and language
standardization) can really help with porting.
> Never came across REXX, but I have used Pascal in the past although I
> never really liked it.
Rexx is a scripting language invented by Mike Cowlishaw at IBM some
years ago. See http://www2.hursley.ibm.com/rexx/ for more info.
It's now standardized by ANSI, I think. Available on a wide range of
(more recent, by a.f.c standards) platforms. It was *the* scripting
language for OS/2, for example. It's also used as an embedded macro
language for applications.
I haven't used Rexx very much, and it's been some time since I last
did anything with it, but as I recall it's a fairly verbose language,
much as COBOL is. It uses words rather than punctuation for most
control structures. Also like COBOL (or Perl, though Rexx is not
like Perl in most ways), Rexx was designed to do things that its
users would be likely to want to do, rather than from, say, a set of
ideas about what makes programming languages elegant or efficient.
(Personally I tend to prefer the elegant sort, but I recognize this
as a valid trade-off.)
--
Michael Wojcik michael...@merant.com
Comms Development, MERANT (block capitals are a company mandate)
Department of English, Miami University
The movie culminated with a bit of everything. -- Jeremy Stephens
Allodoxaphobia
>
>It took a while for Microsoft to groom a generation of users who
>would consider frequent crashes acceptable.
My perception was/is that while the rest of the computer world
was striving for Fault Tolerant Software, Micro$oft was working
on Fault Tolerant Users.
Charlie Gibbs
(Allodoxaphobia) writes:
>On 01 Feb 01 11:02:26 -0800, Charlie Gibbs scribbled:
>
>>It took a while for Microsoft to groom a generation of users who
>>would consider frequent crashes acceptable.
>
>My perception was/is that while the rest of the computer world
>was striving for Fault Tolerant Software, Micro$oft was working
>on Fault Tolerant Users.
ROTFL! That's a keeper. Prepare to be quoted a lot.
Marco S Hyman
> CP/M was pretty much dead by the mid-80s as I recall. It seemed
Killed by Lotus 1-2-3.
> old fashioned. In any event, Gates realized it too, and DOS
> strived to then be Windows. It was at THAT time that CP/M died,
> IMO.
Long before that. Windows (version 3, the almost usable version) wasn't
available until the second half of the 80s (as I remember it, anyhow).
Before that there was always IBM's top view. And OS/2.... eventually
with the presentation manager.
// marc
Steve O'Hara-Smith
> belin...@my-deja.com wrote:
> : I was wondering does anyone else feel a perceived
> : conversion from cp/m to ms-dos in the late 80's?
> CP/M was pretty much dead by the mid-80s as I recall. It seemed
For single user desktop stuff certainly.
Things like MP/M and MMMOST - I think I got the number of Ms right :)
kept it going for a while in places that had come to depend on a multi
user environment and couldn't see what messydos had going for it. Xenix
on fast AT boxes pretty much killed off that segment around 1990.
Paul Repacholi
> I still have the 3 Sanyos collecting dust in my Spider-Infested
> Museum -- but, since they were so proprietary, I'm sure they are
> not worth saving.
That WHY you should save them...
--
Paul Repacholi 1 Crescent Rd.,
+61 (08) 9257-1001 Kalamunda.
West Australia 6076
Raw, Cooked or Well-done, it's all half baked.
Eric Smith
> Our telephone management software was born on a two-processor
> TurboDOS box, had a brief flirtation with a borrowed MP/M box,
> then finally settled (sank?) into MS-DOS (although I eagerly
> did a Unix port when the opportunity arose).
I've been trying to track down a copy of TurboDOS for some years now.
I used it once on a consulting gig. It seemed pretty nice, as CP/M
clones go.
Anyone got the sw and/or docs?
Gene Wirchenko
>In article <slrn97jn2...@animas.frontier.net> Q...@QRM-QRN.net
>(Allodoxaphobia) writes:
>
>>On 01 Feb 01 11:02:26 -0800, Charlie Gibbs scribbled:
>>
>>>It took a while for Microsoft to groom a generation of users who
>>>would consider frequent crashes acceptable.
>>
>>My perception was/is that while the rest of the computer world
>>was striving for Fault Tolerant Software, Micro$oft was working
>>on Fault Tolerant Users.
>
>ROTFL! That's a keeper. Prepare to be quoted a lot.
I have this quote in my sig collection:
"Why put fault tolerance in the OS, when it's already built into the
User?"
- Steve Shaw, comp.os.linux.advocacy, on the apparent lack of
fault tolerance in MS Windows-series of OS's.
Sincerely,
Gene Wirchenko
Computerese Irregular Verb Conjugation:
I have preferences.
You have biases.
He/She has prejudices.
mb.sympatico.ca
>I was wondering does anyone else feel a perceived
>conversion from cp/m to ms-dos in the late 80's?
>
could agree on how to address a serial port, let alone what serial chip to use -
in my CP/M days I ran across just about every single serial chip that was made
at the time ( well, barring an AY 3 1013 or something *that* peculiar) as the serial
I/O. Some were port-mapped. On the Osborne, not only was the chip a Motorola
6850, but it was memory -mapped - and in bank-switched memory. Meanwhile
the port I/O of the Osborne was crippled and completely useless ( except for bank
switching). Some chips had baud rate generators, some did not. A very
few CP/M machines properly supported interrupts for the serial port - now, with
a 2.5 MHZ processor, you really needed interrupts if you wanted to go much
faster than 2400 baud.
Consider disk formats. Again, no two CP/M machines had the same disk format.
Single density, double density ( and on one notably perverse machine in my
collection, the system tracks are single-density and the rest is double density). If
you had an Ibex model 7150 you couldn't swap disks with your buddy who had an
Ibex model 7102 - one was 80-track, the other 40 track.
Consider printer cables. The Osborne had the scummiest printer cable interface
imaginable - a card-edge connector! Again, no two CP/M machines used the
same printer cable. Although I really liked some of the Kaypros that used the same
36-pin "Centronics" style receptacle as most parallel printers used - this made
the cable double-ended and perfectly symmetrical. Would that the PC had chosen
that route, though I suppose they didn't have room for the connector on the edge of
of the cards.
Consider keyboards - if the keyboard in my Osborne siezes up, I'll never be
able to replace it. Custom, proprietary, and unique to that machine - and if I
don't like the layout, I'm out of luck.
While the IBM PC design has all kinds of quirks and limits that have hobbled the
industry for the last 20 years, at least they are *standard* quirks and limits - you
don't have to re-invent the wheel every time. Which was mostly the state that CP/M
had got to; making the simplest things work meant that you had to deal with
each maker's own unique take on the *best* way to do it; when all you really
cared about was getting the job done and not playing about with the hardware.
Since it would cost exactly as much to build a PC clone as a unique custom CP/M
machine, and since the PC clone would have had so much peripheral
hardware available out there - there was no pressing advantage to keeping to the
limits of the 8080 and 64K of RAM any more.
Bill
Tim Shoppa
>
> In <95ar7f$r1q$1...@nnrp1.deja.com>, belin...@my-deja.com writes:
> >I was wondering does anyone else feel a perceived
> >conversion from cp/m to ms-dos in the late 80's?
> >
> It was time. Consider CP/M and serial ports, for instance. No two CP/M machines
> could agree on how to address a serial port, let alone what serial chip to use -
> in my CP/M days I ran across just about every single serial chip that was made
> at the time
And the problem with this is...? The purpose of an OS is to abstract
the concept of input and output to the point where programs don't have
to know about the hardware that implements a port. CP/M did a
wonderful job with this, with 4 logical devices to which various
physical devices could be assigned, via the IOBYTE.
> Consider disk formats. Again, no two CP/M machines had the same disk format.
You're forgetting the granddaddy of them all, the IBM 3740 format
which was (and still is) the de facto standard for CP/M interchange.
> Single density, double density ( and on one notably perverse machine in my
> collection, the system tracks are single-density and the rest is double density).
Again, I think you're holding the flexibility of CP/M up as a
flaw, when it was really a feature.
> While the IBM PC design has all kinds of quirks and limits that have hobbled the
> industry for the last 20 years, at least they are *standard* quirks and limits - you
> don't have to re-invent the wheel every time.
And in the process, we've lost the incredibly rich cornucopia of
microcomputers that flourished under CP/M. Open up an issue of
_BYTE_ from the late 70's and you'll see thousands of ads from
thousands of outfits, all doing their own thing their own way.
Compare that to what we have today, where there are only 4 companies
making PC-clone motherboards. We've lost a lot!
Tim.
Ralph Wade Phillips
<@mb.sympatico.ca> wrote in message
news:nT_e6.4888$wb7.1...@news1.mts.net...
> In <95ar7f$r1q$1...@nnrp1.deja.com>, belin...@my-deja.com writes:
> >I was wondering does anyone else feel a perceived
> >conversion from cp/m to ms-dos in the late 80's?
> >
> It was time. Consider CP/M and serial ports, for instance. No two CP/M
machines
> could agree on how to address a serial port, let alone what serial chip to
use -
> in my CP/M days I ran across just about every single serial chip that was
made
> at the time ( well, barring an AY 3 1013 or something *that* peculiar) as
the serial
> I/O.
<raises hand> I've seen that - well, the AY5-1013 and the AY3-1015
(dunno about an AY3-1013, never heard of it.)
RwP
Douglas Beattie Jr.
This is a different thread now. You're bashing computer architectures,
and it has very little to do with CP/M -- the IBM/ISA architecture was
a very successful marketing push, but CP/M and later DR-DOS were superior
in their functionality and portability -- very useful, and just as useful
today for many things. The serious professional or home-office could
still be using a CP/M machine (and I'm sure many still are), and have
just as much -- if not more -- reliability, and see all the necessary
functionality and productivity; results needed.
Part of the problem is upgrade and replacement, but this is a marketing
thing -- and an ego trip. We want better, faster, fancier... And much
of it is engineered to break, or burn out, or become obsolete -- you
will need to buy more, real soon.
>
> In <95ar7f$r1q$1...@nnrp1.deja.com>, belin...@my-deja.com writes:
> >I was wondering does anyone else feel a perceived
> >conversion from cp/m to ms-dos in the late 80's?
(the original post, and see my comments above).
Does the operating system facilitate interoperability of peripherals and media
compatibility, or does the architecture, or both, and to what degree, respectively?
And what's wrong with re-inventing the wheel? If we didn't, where would we be?
Where are the fundamentals of computer architecture and design today (see my www
site for a few).
Bill <wtsh...@mb.sympatico.ca> wrote
> It was time. Consider CP/M and serial ports, for instance. No two CP/M machines
> could agree on how to address a serial port, let alone what serial chip to use -
> in my CP/M days I ran across just about every single serial chip that was made
> at the time ( well, barring an AY 3 1013 or something *that* peculiar) as the serial
> I/O. Some were port-mapped. On the Osborne, not only was the chip a Motorola
> 6850, but it was memory -mapped - and in bank-switched memory. Meanwhile
> the port I/O of the Osborne was crippled and completely useless ( except for bank
> switching). Some chips had baud rate generators, some did not. A very
> few CP/M machines properly supported interrupts for the serial port - now, with
> a 2.5 MHZ processor, you really needed interrupts if you wanted to go much
> faster than 2400 baud.
Ideally they *would* be port mapped, but that's what a BIOS is for. The end-user
doesn't deal with these issues at all. One machine running a communications program
can exchange data with another, using the same or different chip; same or different
communications program. Such a diversity of architectures and design decisions did
much to teach, and explore design theories. We should be thankful for it.
>
> Consider disk formats. Again, no two CP/M machines had the same disk format.
> Single density, double density ( and on one notably perverse machine in my
> collection, the system tracks are single-density and the rest is double density).
> If you had an Ibex model 7150 you couldn't swap disks with your buddy who had an
> Ibex model 7102 - one was 80-track, the other 40 track.
These are different media. I can't read a 1.44MB disk in a 720K drive either.
That's what media conversion is for. There are media conversion utilities for
DOS, CP/M , Unix and others. Granted: today, media interchange is state-of-the-art
but it would have been anyway -- but you're right... time and advances in technology
and a capitalist incentive factor have facilitated all this.
>
> Consider printer cables. The Osborne had the scummiest printer cable interface
> imaginable - a card-edge connector! Again, no two CP/M machines used the
> same printer cable. Although I really liked some of the Kaypros that used the same
> 36-pin "Centronics" style receptacle as most parallel printers used - this made
> the cable double-ended and perfectly symmetrical. Would that the PC had chosen
> that route, though I suppose they didn't have room for the connector on the edge of
> of the cards.
So, people moved to MS-DOS from CP/M because of the printer connector on the back
of their CP/M machine... That might explain everything -- it's a pretty solid theory!
I have often considered the printer to be the most important part of my computer...
{ but there's more... }
>
> Consider keyboards - if the keyboard in my Osborne siezes up, I'll never be
> able to replace it. Custom, proprietary, and unique to that machine - and if I
> don't like the layout, I'm out of luck.
So you mean it's still working? I guess they don't build keyboards like they used
to. Today I sometimes wonder if a keyboard isn't designed to lock up at least twice
within the first 8 months I've had it (so my company will be forced to buy another
one, from the same vendor). I sympathize, though... it's a very real fear... your
keyboard is becoming harder to find a replacement for. (BTW: I hear that regular
exercise of the keys on your Osborne will prolong its life. ;)
> While the IBM PC design has all kinds of quirks and limits that have hobbled the
> industry for the last 20 years, at least they are *standard* quirks and limits - you
> don't have to re-invent the wheel every time. Which was mostly the state that CP/M
> had got to; making the simplest things work meant that you had to deal with
> each maker's own unique take on the *best* way to do it; when all you really
> cared about was getting the job done and not playing about with the hardware.
This is not CP/M -- it's computer architecture and design. CP/M is not a hardware
architecture. CP/M proved that hardware abstraction was effective, while Computer
engineers proved that there was more that one way to build a computer. And lately,
for years now, one manufacturer has virtually copied the architecture of another
clone, and branded it as their own. A hardware standard has done its fair share to
cripple innovation -- only now are competitors building momentum, and exploiting
diverse architectures; re-inventing the wheel. Unique computer architectures today
still have media interoperability, particularly through communication protocols and
networked filesystems, and the end-user doesn't usually deal with these things at all.
And perhaps just as Windows has been a standard, stifling creativity, and crippling
innovation, a hardware standard has done the same in some respects. Why the memory-
mapped I/O? And further (though perhaps less-relevant), why the segmented CPU?
We discovered alot in the CP/M era, because CP/M would run on almost anything.
Write only the BIOS, and it runs on even the most arcane architecture. You can't
do that any more {with a few exceptions}.
Perhaps Computer Science graduates know less about the fundamentals today than they
did 15 years ago. Is it a trend in education, or have we swept too many fundamentals
under the rug because they're just not "high-tech" enough to appeal to University
curriculum and subsidized funding...? And don't they still use CP/M in education?
(Incidentally, there's alot of good information at http://www.maxframe.com/DR.HTM )
>
> Since it would cost exactly as much to build a PC clone as a unique custom CP/M
> machine, and since the PC clone would have had so much peripheral hardware available
> out there - there was no pressing advantage to keeping to the limits of the 8080
> and 64K of RAM any more.
Cost would be considerably less. Basic functionality for typical applications and
the home/professional user would be comparable. The computer industry in the U.S.
isn't allowing this, though. Manufacturing and funding and engineering costs (and
other factors) won't allow a CP/M-based architecture to maintain a presence in the
market today. However, it doesn't mean that it couldn't happen, or isn't happening
elsewhere in the world. Is CP/M still used in other countries? Are these diverse
architectures still being designed, manufactured and deployed elsewhere in the world?
I'd like to know -- please tell me.
>
> Bill
--
Douglas Beattie Jr. http://www2.whidbey.net/~beattidp/
Steve O'Hara-Smith
> In <95ar7f$r1q$1...@nnrp1.deja.com>, belin...@my-deja.com writes:
>>I was wondering does anyone else feel a perceived
>>conversion from cp/m to ms-dos in the late 80's?
>>
> switching). Some chips had baud rate generators, some did not. A very
> few CP/M machines properly supported interrupts for the serial port - now, with
Indeed CP/M wasn't even supposed to support interrupts really. Then again
it did many things it wasn't originally intended for :)
I fondly remember the fun of the 'extra' registers on the Z80. Some apps
(a certain BCPL compiler springs to mind) would assume that since this was
an 8080 OS they didn't need to take care of them. Some MP/M implementations
assumed they were running 8080 apps and didn't need to be careful with them
...
.. lots of snippage ..
I tend to agree, hardware standardisation pushed by big marketing muscles
has pretty well killed off the variety. But think of the lost markets in
adaptors and cross platform floppy reading software (after all the drives
were usually compatable).
Rickety
news:3A7C70BB...@trailing-edge.com...
..and the compatibility of PCs didn't come immediately either. I recall
that everyone had their own favourite way of deciding how "compatible" a
PC-compatible system was. We used to think if it could run "Sopwith" it was
pretty good!
I also concur with your point Tim that it wasn't a fault of CP/M that there
were differences, it was the wonder of CP/M that all those different systems
could run the same OS, utilities and applications. The variations of floppy
disc formats certainly was the most inconvenient characteristic of systems
at the time as we needed different copies of programs residing on the target
system disc formats. Later on I had a program "Cross-something" that would
copy to and from a wide variety of disc formats including Osborne 1 and PC.
Cheers
--
Rick Lugg
--
Rick Lugg
mb.sympatico.ca
>@mb.sympatico.ca wrote:
>>
>> In <95ar7f$r1q$1...@nnrp1.deja.com>, belin...@my-deja.com writes:
>> >I was wondering does anyone else feel a perceived
>> >conversion from cp/m to ms-dos in the late 80's?
>> >
>> It was time. Consider CP/M and serial ports, for instance. No two CP/M machines
>> could agree on how to address a serial port, let alone what serial chip to use -
>> in my CP/M days I ran across just about every single serial chip that was made
>> at the time
>
>And the problem with this is...? The purpose of an OS is to abstract
>the concept of input and output to the point where programs don't have
>to know about the hardware that implements a port. CP/M did a
>wonderful job with this, with 4 logical devices to which various
>physical devices could be assigned, via the IOBYTE.
>
in an implementation-independant way. I'm not aware of any popular CP/M communications
program that relied on BDOS or BIOS services to access the serial port. This meant
that every program to talk to a serial port had to be hand-hacked to work. In this
sense CP/M was *not* hiding the underlying complexity of the hardware from the user.
>> Consider disk formats. Again, no two CP/M machines had the same disk format.
>
>You're forgetting the granddaddy of them all, the IBM 3740 format
>which was (and still is) the de facto standard for CP/M interchange.
I've never used or seen a machine that supported the original 8 inch single
density disk format. I came to the world of CP/M too late, in 1981 - by this
time the PC impact was already well under way. If you bought an Osborne
and didn't have access to hacker buddies who could transcribe files for you,
you were out of luck- no way to hook up an 8 inch SD drive to an Osborne
( at least, not without spending more than any conceivable program was worth).
>
>> Single density, double density ( and on one notably perverse machine in my
>> collection, the system tracks are single-density and the rest is double density).
>
>Again, I think you're holding the flexibility of CP/M up as a
>flaw, when it was really a feature.
"That's not a bug, that's a feature!". It did make commercial support for CP/M
programs more difficult because you either had to have one of every single model
under the sun or else have some expensive custom machinery to produce all the
concievable formats.
(I don't know if there were any GCR CP/M formats aside from the obscure Commodore 64
CP/M pack - wasn't GCR disk format one of the things that killed off the Victor 9000
8086 machine? This ran DR CP/M 86.)
>
>> While the IBM PC design has all kinds of quirks and limits that have hobbled the
>> industry for the last 20 years, at least they are *standard* quirks and limits - you
>> don't have to re-invent the wheel every time.
>
>And in the process, we've lost the incredibly rich cornucopia of
>microcomputers that flourished under CP/M. Open up an issue of
>_BYTE_ from the late 70's and you'll see thousands of ads from
>thousands of outfits, all doing their own thing their own way.
>Compare that to what we have today, where there are only 4 companies
>making PC-clone motherboards. We've lost a lot!
We've lost a lot of the trilobites and other fantastically-armoured and -shaped
creatures out of the personal computing Jurassic era. But I really don't think
the attitude that said " Hey, here's a new serial port chip! Let's build a CP/M
computer around it and make it totally unique!" is much to be nostalgic about.
The world was full of 4 MHZ Z80 systems that all did pretty much the same thing
from the user's point of view, but that, for baffling technical reasons incapable of
being explained to the people who authorized purchase orders, were incapable of
using each other's periperhals, software, or cables without paying even *more* money
to someone who was a champion of the particular obscure platform.
The fragementation of the user platform meant that worthwhile things like
usable sound or graphics was impossible in the CP/M world ( yes, I *know* there
were CP/M machines with spiffy sound or with spiffy near-VGA-level graphics - but
there was no way to USE these features without rolling your own software, by and
large.)
We wouldn't have Linux today if the attitude of the CP/M hardware developers
had persisted into the 32-bit era. Yes, the DMA controller and interrupts and a
whole *lot* of things are now done very awkwardly *but* there's a hundred
million machines out there that all do these things in the *same* bad way - which
means you only have to solve the problem once.
Bill
>
>Tim.
mb.sympatico.ca
Whoops. It's been a while since I've had my hands on one and I never did
memorize the different models. I stand corrected. Come to think of it, my
TRS 80 Model 100 uses something like an AY 3 1015 or similar style UART for
the serial port, though obviously not running CP/M.
Bill Shymanski
mb.sympatico.ca
>Hi Bill (and all in news:comp.os.cpm );
>
>This is a different thread now. You're bashing computer architectures,
>and it has very little to do with CP/M -- the IBM/ISA architecture was
>a very successful marketing push, but CP/M and later DR-DOS were superior
>in their functionality and portability -- very useful, and just as useful
>today for many things. The serious professional or home-office could
>still be using a CP/M machine (and I'm sure many still are), and have
>just as much -- if not more -- reliability, and see all the necessary
>functionality and productivity; results needed.
Theoretically, you could still use a CP/M Z80 4 MZ 64K machine with two
floppies to solve any computable problem. ( Or a Universal Turing Machine
and enough paper tape, for that matter). Practically, 99.997% of all computer
users today don't write their own programs and getting the most of a
CP/M machine required far more user time and attention than anyone would
reasonably expect to invest today.
>
>Part of the problem is upgrade and replacement, but this is a marketing
>thing -- and an ego trip. We want better, faster, fancier... And much
>of it is engineered to break, or burn out, or become obsolete -- you
>will need to buy more, real soon.
But if a thousand-megahertz 32 bits machine with 128Megs of RAM and a
googolbyte of hard disks costs the exact *same* as the CP/M machine,
what's the point of restricting yourself to the less capable platform? Any
application you could run on a Z80 machine you can run on a Pentium and not
only will it be faster and cheaper, it will come in a blister pack at your local
software store ( not mail-order in 8 to 12 weeks).
Dead wrong. No CP/M communication program of my experience used the BDOS services
for anything. There was no way to write a functional serial communications program
that was implementation -independant on any CP/M compatible system. This is also
true of the Intel machines descended from the PC with the exception that the
implementations are all the *same*.
>
>>
>> Consider disk formats. Again, no two CP/M machines had the same disk format.
>> Single density, double density ( and on one notably perverse machine in my
>> collection, the system tracks are single-density and the rest is double density).
>> If you had an Ibex model 7150 you couldn't swap disks with your buddy who had an
>> Ibex model 7102 - one was 80-track, the other 40 track.
>
>These are different media. I can't read a 1.44MB disk in a 720K drive either.
>That's what media conversion is for.
You miss my point. Aside from the legendary 8 inch distribution format, there was
*no* standard for CP/M floppies. It's not different media - the same
5 1/4 disk could be formatted on two *different* CP/M implementations, and
neither could read the other's diskettes. This verged on gibbering
insanity - the whole point of diskettes is to interchange files and programs with
other systems.
>
>>
>> Consider printer cables. The Osborne had the scummiest printer cable interface
>> imaginable - a card-edge connector! Again, no two CP/M machines used the
>> same printer cable.
>So, people moved to MS-DOS from CP/M because of the printer connector on the back
>of their CP/M machine... That might explain everything -- it's a pretty solid theory!
>I have often considered the printer to be the most important part of my computer...
>{ but there's more... }
Depends on what you're doing of course, but if it was 1979 and someone had
just sold you a system that cost as much as a small car and that was claimed to
do word-processing, you'd be exceptionally ticked off if if could *not* print
anything.
No-one remembers how fantastically *expensive* computers were only 20 years
ago, especially considering what very little they could actually do. And as a
result (by today's standards) no-one *had* computers n the home in those days.
>
>
>> While the IBM PC design has all kinds of quirks and limits that have hobbled the
>This is not CP/M -- it's computer architecture and design. CP/M is not a hardware
>architecture. CP/M proved that hardware abstraction was effective, while Computer
>engineers proved that there was more that one way to build a computer.
I am aware, though I'm not expressing it well, that there is a distinction between
the operating system and the hardware platforms it comes on. (MS DOS has done
surprisingly poorly on the Macintosh, for example!). The point I was trying to
make is that the reason CP/M vanished circa 1981 or so is that no matter
what you thought about MS DOS, the hardware it ran on was so much less
hassle to get working in usable fashion that there was no incentive to
stick with the 8080 world any more.
> A hardware standard has done its fair share to
>cripple innovation -- only now are competitors building momentum, and exploiting
>diverse architectures; re-inventing the wheel.
Name one CP/M "innovation" that still persists today. Thankfully we've at last got
rid of 8.3 filenames (well, we can use them, but we're not obliged to). Most of
the great ingenuity of the CP/M era was expended in getting useful things done
in *spite* of the hardware/software platform.
> Unique computer architectures today
>still have media interoperability, particularly through communication protocols and
>networked filesystems, and the end-user doesn't usually deal with these things at all.
>
I heard about CP/Net - the network capable version of CP/M. It would have been
a real challenge to fit useable network functionality into the limited address space
of the 8-bit era. People do build ships in bottles, but no-one uses the results to
ship freight.
>And perhaps just as Windows has been a standard, stifling creativity, and crippling
>innovation, a hardware standard has done the same in some respects.
Windows has "stifled" creativity? You mean I no longer have to guess which way
the author of whatever program I'm using has decided to implement the "print"
command? I don't think *that* kind of creativity is missed at all.
>We discovered alot in the CP/M era, because CP/M would run on almost anything.
>Write only the BIOS, and it runs on even the most arcane architecture.
True, and it will only support character-mode I/O through the default console and
can't talk to the outside world. We should be trying to discourage the "arcane"
in architecture - computers are meant to be Useful Things, not vehicles to
express the ineffable yearning of Man's questing spirit for union with the overcosmos,
or whatever they were thinking when they designed the Osborne and forgot to
decode any more than the two bottom IO port address lines.
Maybe there's a whole esthetic here that I've missed out...maybe *that's* why
the goofy proliferation of keyboard, I/O and video methodolgies?
>>
>> Since it would cost exactly as much to build a PC clone as a unique custom CP/M
>> machine, and since the PC clone would have had so much peripheral hardware available
>> out there - there was no pressing advantage to keeping to the limits of the 8080
>> and 64K of RAM any more.
>
>Cost would be considerably less.
You need a CP/U. Yes, a Z80 costs $1 vs. a Pentium of a few bucks.
You need a printed circuit board. This cost by the square inch - arguably
an 8 bit board would be a bit smaller (because it can do so less). You
could probably get all of the esstential to run a CP/M environment on
a single chip that would fit in a wrist watch.
You need a power supply. These cost the same.
You need a case to put it all in. Same cost.
You need a keyboard. You could use a clone keyboard or roll your own in
true 1979 "Hey, Ma! I built me a computer!" style at 100 times the cost.
You need FCC certification, manuals, sales/support, all the rest - these are
either the same cost or much pricier for a 21st century CP/M machine owing
to its vastly smaller volume.
You need DVD ROM, 32-bit colour, USB, Ethernet, 56 K modems, stereo sound...
whups - can't do that on a CP/M machine without a *lot* of R&D, yet the meanest
clone bought in the supermarket between the cat food and diapers can
do all these things.
You need to persuade someone to buy the thing....well, I suppose you could
put something in the water supply.
Actually, you don't need any of these things because you can emulate everything
that CP/M did in software on any MS DOS machine and it will run 20 times
faster than any CP/M computer ever sold. So, CP/M machines are "free".
Basic functionality for typical applications and
>the home/professional user would be comparable. The computer industry in the U.S.
>isn't allowing this, though. Manufacturing and funding and engineering costs (and
>other factors) won't allow a CP/M-based architecture to maintain a presence in the
>market today. However, it doesn't mean that it couldn't happen, or isn't happening
>elsewhere in the world. Is CP/M still used in other countries? Are these diverse
>architectures still being designed, manufactured and deployed elsewhere in the world?
>I'd like to know -- please tell me.
Just because someone lives in a Third WOrld site doesn't mean they will put up
with 1979-era computers, even if they are being given away. I doubt that
anywhere in the world that there is any CP/M being sold commercially.
Bill
bill
> You miss my point. Aside from the legendary 8 inch distribution format, there was
> *no* standard for CP/M floppies. It's not different media - the same
> 5 1/4 disk could be formatted on two *different* CP/M implementations, and
> neither could read the other's diskettes. This verged on gibbering
> insanity - the whole point of diskettes is to interchange files and programs with
> other systems.
Well your ''point'' is just plain WRONG. You might need a program like
Uniform
(there were others, as well) but you COULD set up one of your drives to
read
'foreign' formats, usually via another added drive letter. It was no
where near
as hard as many folks who weren't there, and didn't do it, are now
making out.
> >> Consider printer cables. The Osborne had the scummiest printer cable interface
> >> imaginable - a card-edge connector! Again, no two CP/M machines used the
> >> same printer cable.
>
> >So, people moved to MS-DOS from CP/M because of the printer connector on the back
> >of their CP/M machine... That might explain everything -- it's a pretty solid theory!
> >I have often considered the printer to be the most important part of my computer...
> >{ but there's more... }
>
> Depends on what you're doing of course, but if it was 1979 and someone had
> just sold you a system that cost as much as a small car and that was claimed to
> do word-processing, you'd be exceptionally ticked off if if could *not* print
> anything.
I suspect many people moved to the IBM and/or compatibles because of the
Laser
Printer. You could now produce type-set quality output, soon followed by
bit
mapped graphics, that previously had been virtually impossible for under
maybe
$100,000 worth of equipment.
Businesses and offices did not like dot matrix printers. They were
noisy, they
were slow, and they produced crappy output. And you could only get a few
hundred
pages out of a ribbon before it began to fail.
I did some fund raising letters, and Daisy Wheel printed output usually
out-pulled dot matrix printed letters by anywhere from 3x to 10x. I
would
NOT want to listen to a Dasiy Wheel running a several hour job ever
again.
That's reserved for a special corner somewhere in Hell.
Laser printing changed all that. But with a laser printer costing $3,000
to
$5,000 not many showed up in the home. But there were some.
I suspect that without the laser printer, the MAC would never have
gained much
of a following, either.
> No-one remembers how fantastically *expensive* computers were only 20 years
> ago, especially considering what very little they could actually do. And as a
> result (by today's standards) no-one *had* computers n the home in those days.
I had several. Being able to access a BBS was also a good reason.
> >> While the IBM PC design has all kinds of quirks and limits that have hobbled the
>
> >This is not CP/M -- it's computer architecture and design. CP/M is not a hardware
> >architecture. CP/M proved that hardware abstraction was effective, while Computer
> >engineers proved that there was more that one way to build a computer.
CP/M certainly WAS a hardware limited architecture. It didn't do
graphics
> Name one CP/M "innovation" that still persists today. Thankfully we've at last got
> rid of 8.3 filenames (well, we can use them, but we're not obliged to). Most of
> the great ingenuity of the CP/M era was expended in getting useful things done
> in *spite* of the hardware/software platform.
....
> Just because someone lives in a Third WOrld site doesn't mean they will put up
> with 1979-era computers, even if they are being given away. I doubt that
> anywhere in the world that there is any CP/M being sold commercially.
It would take a book to enlighten this much ignorance. In fact, it would
take
the writer who left it the reading of several books, and understanding
them.
As far as I know MS is not giving Windows away. It's based on CP/M. As
you
would know if you'd been reading here for any length of time.
Bill
Tucson, AZ
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
Kelsang Dawa
> You miss my point. Aside from the legendary 8 inch distribution format, there was
> *no* standard for CP/M floppies. It's not different media - the same
> 5 1/4 disk could be formatted on two *different* CP/M implementations, and
> neither could read the other's diskettes. This verged on gibbering
> insanity - the whole point of diskettes is to interchange files and programs with
> other systems.
Whoa, whoa. Back up a few steps. The whole point of diskettes *has
become* to interchange files and programs with other systems. Around
the time of CP/M systems floppy disks did the much more important job of
storing files for your local machine, a role that has now passed to hard
drives.
But mainly I would like to say thank you for writing your piece. It
made me pause and think a little about my obsession with CP/M, which can
only ever help. I also appreciate the words being light and fun.
dawa
B W Spoor
>
>[SNIP]
>
>
> No-one remembers how fantastically *expensive* computers were only 20
> years ago, especially considering what very little they could actually
> the home in those days.
>
>
Wordstar 4 in a DOS box under NT4) for a system that I installed around
1984/5 for an 8-bit network system (ICL DRS-20).
ISTR that a basic workstation was around £2,000 while a server unit was
around £5,000. An OKI Microline 84a (dot matrix printer - still have a
working one and you can still get the ribbons!) was around £1,000 with
the cable costing around £100 - ICL list prices in ukp
The whole system, hardware and software, was around £60,000 of which
about £20,000 was for the custom software. This was a fraction of what
it would have cost to implement on their ME29 mainframe. The finance
director was happy, as the staff savings would pay for the system in a
little over a year.
>
>
>[SNIP]
>
>
> I heard about CP/Net - the network capable version of CP/M. It would
> have been a real challenge to fit useable network functionality into
> the limited address space of the 8-bit era. People do build ships in > bottles, but no-one uses the results to ship freight.
>
>
The ICL DRX operating system for the DRS20 machines was an 8-bit
network operating system, and ISTR that it ran quite well for it's time.
The DRS-20 used multiple 8085 processors in a box running on a
synchronised bus, one processor per card, so a basic workstation would
have a keyboard/display processor (controlling the machine) and an
applications processor running the actual program (in 64Kb). There was a
built in network adaptor, AFAIR controlled via the k/d processor.
A file server would have a keyboard/display processor (controlling the
machine), a file processor, a network processor and an applications
processor running the actual program as it was also a workstation.
There were slots for additional cards, such as a second AP to run a
print spooler, or a communications processor (I used them for WAN
connections, driving a Hassler Telex box for automated generation &
sending of telexes, synchronous comms to an ICL ME29 mainframe for FTP
and MAC - multi access computing, ie using a workstation as a terminal
on the ME29 for data entry).
I designed the system with 2 file servers, primary and secondary for
reliability, with the data being updated simultaneously on both
machines. Updates used the primary files and also wrote back to the
secondary, while enquiries and reports used the secondary fles. This
helped to load share and improve response times - not brilliant, but
probably the best that could be done without spending a far greater
amount of money on a mini-computer.
The central LAN had a telex interface, 2 WAN links (server to server) to
other buildings which had their own LANs running the same system (the
local server had it's own copy of the programs, but the data was over
the WAN), and an ICL C03 (FTP/MAC) link to the ME29 mainframe.
This system ran reliably for around 7 years with very few hardware or
system failures, before being replaced with Netware 2.2 and DOS PCs. So
8-bit networking was around and did work.
Unfortunately, when the DRS20s were replaced, I didn't have space to
take them away - they languished in a back storeroom for quite a while,
but when I looked (now having room for them) they had been scrapped. I
was also asked if I was interested in the ME29 when it was replaced by
an AS/400, but again had no space or means of shifting it. I regret this
now.
>
>
>[SNIP]
jchausler
@mb.sympatico.ca wrote:
> No-one remembers how fantastically *expensive* computers were only 20 years
> ago, especially considering what very little they could actually do. And as a
> result (by today's standards) no-one *had* computers n the home in those days.
Huh!? I had several computers in my home 20 years ago and I don't recall them
being THAT expensive. Yes I assembled them myself (that was half the fun) and
yes the biggest one had only 56K "RAM" memory (the smallest less than 1K :-)
and storage was 5 1/4 inch floppies or even audio cassette tape for some but I
did useful things with them. Now, 30 years ago..............
Maybe cause I wasn't using CP/M I had things easier. Although there were
several different OS's for the 6800/6809 most ran FLEX and although there
was a minor difference between two different vendors for double density
floppy disks I recall exchanging single density, single or double sided,
5 1/4 inch or 8 inch floppies with others with no problem at all and then
of course there were "flippies" as well. I had no problems with them
either. Of course, "no one" had hard drives, they were expensive. As
someone else pointed out, floppies weren't so much for exchange,
they WERE your primary disk storage so as long as you could read
you're own floppies and single sided single density floppies for
purchased software, you were usually OK.
As to printers, most folks I know used serial interface printers with
"buffers" and at least the parallel port for the SS-50 world was
standard if not intended for heavy use. I just built a short cable
from it to a DB-15 which became MY standard parallel printer
interface, centronics connectors being rather expensive.
> Name one CP/M "innovation" that still persists today. Thankfully we've at last got
> rid of 8.3 filenames (well, we can use them, but we're not obliged to).
This seriously predates CP/M.
> Most of
> the great ingenuity of the CP/M era was expended in getting useful things done
> in *spite* of the hardware/software platform.
Frequently that's how its been from the dawn of the computer age and I'm
at least frequently still having to do it in spite of all the "wonderful" technology
we have today. The technology is opening up ever bigger horizons which we
try and get done using today's technology. Try running "today's" programs on
a PC of just 8 to 10 years ago. It won't. I know, I've tried.
> I heard about CP/Net - the network capable version of CP/M. It would have been
> a real challenge to fit useable network functionality into the limited address space
> of the 8-bit era. People do build ships in bottles, but no-one uses the results to
> ship freight.
What do you expect, the technology has gotten better. Back then the ARPANET
ran on IMP's. These were 16 bit minicomputers more or less doing what that
plug in card in your current machine for your ethernet connection does. And
speaking of ethernet, have you ever looked at some of the early ethernet
controllers. I used a number of them from Interlan which were bigger than
a laptop and so crowded with SSI chips that they had thermal problems.
Chris
AN GETTO$;DUMP;RUN,ALGOL,TAPE
$$
James Korman
jchausler wrote:
>
> @mb.sympatico.ca wrote:
>
> > No-one remembers how fantastically *expensive* computers were only 20 years
> > ago, especially considering what very little they could actually do. And as a
> > result (by today's standards) no-one *had* computers n the home in those days.
>
> Huh!? I had several computers in my home 20 years ago and I don't recall them
> being THAT expensive. Yes I assembled them myself (that was half the fun) and
> yes the biggest one had only 56K "RAM" memory (the smallest less than 1K :-)
> and storage was 5 1/4 inch floppies or even audio cassette tape for some but I
> did useful things with them. Now, 30 years ago..............
????????????
I've still got my Tandy Model 4P, runs CP/M well. But...It cost me
$1799.95
in 1983 dollars! That was 2 SSDD drives, 64K, 80X24 display. I just
purchased
an HP 5195 laptop for just a little more. 700MHz PIII, 128Meg of RAM,
etc, need
I go on. The lap top will emulate my 4P at a clock speed of about 50MHz
or so.
That old stuff was quite expensive...And also the greatest learning
experience
of my life.
Jim Korman
Ralph Wade Phillips
<@mb.sympatico.ca> wrote in message
news:9fAf6.6438$wb7.1...@news1.mts.net...
>
> (I don't know if there were any GCR CP/M formats aside from the obscure
Commodore 64
> CP/M pack - wasn't GCR disk format one of the things that killed off the
Victor 9000
> 8086 machine? This ran DR CP/M 86.)
Apple ][ SoftCard, AppliCard, or any Z80 coprocessor for the
Apple ][ family.
Also the "clone"s of the Apple ][ (BASIS 108 comes to my mind here.)
RwP
Roger Johnstone
In article <3a8105ed$0$30003$140e...@news.tdin.com>, Scott
<scot...@bigfoot.com> wrote:
> In article <slrn981sj...@dorothy.msas.net>, Jay R. Ashworth
> <j...@dorothy.msas.net> wrote:
>
>> Absolutely my favorite dead computing idea of the last 20 years was
>> Apple's OpenFirmware: they wrote the on-card bios in *FORTH*, and put a
>> FORTH interpreter on the motherboard. This allows you to make the
>> cards processor independent while still allowing intelligence on each
>> card.
>>
>> It was an absolutely *elegant* idea, probably the most elegant one ever
>> to come out of Cupertino IMHO, and I think it's a damned shame it never
>> caught on.
>>
> It did catch on after a fashion. I believe Sun uses Opem Firmware too
> It's great fun to hold down Command-Option-o-f and drop into the
> command line! :-)
Yikes, it was Sun who started with Open Firmware. Apple hopped on board
later. It is a brilliant idea though. Being able to plug a card into any
type of computer with any type of processor and have it be able to use it
without loading drivers. Very important for video, storage and network cards
which have to work before the OS loads.
The big problem for Open Firmware is the PC clones. Since they all use the
same processor/architecture/ROM they don't really need Open Firmware, and
since they're the majority of the market hardly any cards are made with OF.
--
Roger Johnstone, Invercargill, New Zealand
Apple II - Future Cop:LAPD - Warcraft II
http://homepage.mac.com/rojaws
______________________________________________________________________
My software never has bugs. It just develops random features.
Ben Harris
Jay R. Ashworth <j...@baylink.com> wrote:
>On Tue, 06 Feb 2001 18:34:42 -0400,
>I think a very good case can be made that Linux runs on more processor
>types than any other single program now or before.
That's blatantly untrue. The "hello, world" program from K&R has probably
been run on every system with a C compiler (which includes plenty of
systems Linux won't run on).
As far as real-world programs go, C-Kermit and Info-Zip are probably about
as portable (and ported) as you'll get. Even GCC supports more CPU types
than Linux.
If you insist on comparing OSes, then I think NetBSD gives Linux a run for
its money (last I counted they each did ten CPUs as standard).
--
Ben Harris
Unix Support, University of Cambridge Computing Service.
If I wanted to speak for the University, I'd be in ucam.comp-serv.announce.
gla...@glass2.lexington.ibm.com
>@mb.sympatico.ca wrote:
>> >> Consider disk formats. Again, no two CP/M machines had the same disk format.
>> >
>> >You're forgetting the granddaddy of them all, the IBM 3740 format
>> >which was (and still is) the de facto standard for CP/M interchange.
>> I've never used or seen a machine that supported the original 8 inch single
>> density disk format.
>
>based microcomputers with 5.25-inch disk drives in the early 80's",
>that's not the same CP/M that I grew up with. The CP/M that I grew
>up with came on an 8" floppy, had a license signed by Gary Kildall,
>and came with a "CP/M Customization Guide" that let you install it
>on any hardware you wanted to hook up to your 8080, by toggling
>in a bootstrap, loading the boot sector, and eventually writing
>your own BIOS.
>
>If you got a "packaged" microcomputer system with the BIOS already
>written and installed, IMHO you got screwed. Having the source to
>your CP/M BIOS is absolutely essential to understanding how
>versatile the thing is.
>
>(Comment for the newbies: The CP/M BIOS was code that you actually
>wrote to do the device-dependent I/O functions. Don't confuse it
>with "IBM-PC-clone-BIOS-settings" which is what it means to probably
>99% of today's population.)
>
>> I came to the world of CP/M too late, in 1981
>
>> >microcomputers that flourished under CP/M. Open up an issue of
>> >_BYTE_ from the late 70's and you'll see thousands of ads from
>> >thousands of outfits, all doing their own thing their own way.
>> >Compare that to what we have today, where there are only 4 companies
>> >making PC-clone motherboards. We've lost a lot!
>>
>> We've lost a lot of the trilobites and other fantastically-armoured and -shaped
>> creatures out of the personal computing Jurassic era. But I really don't think
>> the attitude that said " Hey, here's a new serial port chip! Let's build a CP/M
>> computer around it and make it totally unique!" is much to be nostalgic about.
>
>Putting together an Altair or IMSAI with boards from random ads in
>the back of BYTE and a few of your own design *was* a very happening
>sector of microcomputing back then.
There were some of us who even designed our own motherboards,
usually based around available microprocessors, although I know one
guy who designed his own processor from discrete TTL (Sheeze, what a
mess!). Of course, we usually had to write our own monitor code,
and the systems were usually pretty simplistic. Most of us lusted
after systems with floppy drives, and would have given someone's
right arm for a system that would run CP/M. Tapes were pretty
awful, but they beat having to key the application code in by hand
(Usually we burned the monitor into a PROM or EPROM.). However,
after using one for a while, we became pretty proficient at using a
keypad or a bank of switches to key in a program, and we could
immediately tell what was happening by examining the status lights.
Umm, does anyone need any 2114 static RAM chips? I think I cornered
the market on them about 20+ years ago, and still have some left.
:*)
>If you think that buying a PC-clone motherboard
>and plugging in a couple of boards so that you can boot MS-DOS
>is "building a computer", us folks in alt.folklore.computers have
>a few things to teach you :-).
>
>> The world was full of 4 MHZ Z80 systems that all did pretty much the same thing
>> from the user's point of view, but that, for baffling technical reasons incapable of
>> being explained to the people who authorized purchase orders, were incapable of
>> using each other's periperhals, software, or cables without paying even *more* money
>> to someone who was a champion of the particular obscure platform.
>
>Again, I think you got in too late - or with a seriously lame
>attitude - to know what was fun about, for example, S-100 computing
>with CP/M in the late 70's. The experiences I gained weren't ones
>that I'd wish on J Random Luser, but they were very valuable to me.
In an way, they were 'fun', but I sure don't wish to repeat those
experiences. Still, we developed an understanding of the system.
(Plus, it's so much easier developing systems around things like the
Microchip PIC series of chips.)
>> We wouldn't have Linux today if the attitude of the CP/M hardware developers
>> had persisted into the 32-bit era. Yes, the DMA controller and interrupts and a
>> whole *lot* of things are now done very awkwardly *but* there's a hundred
>> million machines out there that all do these things in the *same* bad way - which
>> means you only have to solve the problem once.
>
>I think you miss one of the fundamental points about Linux, that
>IT IS PORTABLE AMONG MULTIPLE PLATFORMS. (OK, it wasn't portable
>at first, but now it's pretty good.) It runs on 80x86, it runs
>on Alpha, it runs on SPARC, it runs on lots of other CPU's,
>and it runs with lots of different peripherals that hook up to
>these CPU's, and if you want to write a driver for your own custom
>device it's all documented very clearly.
(Shameless plug warning:)
Hey, you left out the IBM ESA/390 (or, should I say IBM z/Architecture)
mainframes! (Ok, so I work for them.).
>Note that the original MS-DOS (actually, QDOS) was not tied to
>any particular peripheral layout either. It ran on a wide variety
>of 8086/8088 machines, many of them S-100 machines with periperhals
>that were also usable under the 8086 version of CP/M. It's the
>PC-clone market that cemented all the bad things about the original
>IBM PC into stone.
>
>Tim.
Dave - architect of the HG-51 microcomputer.
P.S. Standard Disclaimer: I work for them, but I don't speak for them.
jchausler
gla...@glass2.lexington.ibm.com wrote:
> Umm, does anyone need any 2114 static RAM chips? I think I cornered
> the market on them about 20+ years ago, and still have some left.
> :*)
I could use a few as spares, they're the main memory component (ahead of
TMS4404's and 2102's of which there are also quite a few :-) of all of my
8 bit 68XX stuff. I don't have anymore spares of them.
jchausler
@mb.sympatico.ca wrote:
> Hmm. 1981. I bought an Osborne 1. It had a 5 1/4 inch monochrome text screen, a
> 4 MHZ Z80 processor, 64K of RAM, two single-density 5 1/4 icnh diskette drives
> storing an awesome 93 kilobytes each. It came bundled with a word-processor,
> a spread sheet, interpreted BASIC, and a BASIC "compiler". This cost at the
> time $2700 Canadian, which in those days was several months take-home pay
> for a newly-graduated engineering student. I could process words on it - provided
> I didn't want to check spelling or use footnotes. I could do smallish spreadsheets -
> you could do a whole year month-by-month, but it would have been hard to
> do a year week-by-week or day-by-day.
>
> In 2000, I bought a Toshiba laptop. It came with 64 meg of RAM, a 4 gig hard drive,
> a 1.44 meg floppy drive, a DVD ROM drive, colour 800*600 display...yadda, yadda...
> a nice laptop, quite unexceptional by today's standards. 430 MHZ processor.
> This cost me around $2300 Canadian. It is a far more useful machine, as well as
> being more powerful - I can carry everything I need around in it, without the
> endless floppy shuffle that the Osborne required. Oh, and I can use it on an
> airplane.
Again, the technology has gotten better along with the expectations. Further,
at that time, it was usually a lot cheaper to buy "kits" and "bare boards" and
more or less roll your own than buy an off the shelf system. Of course, many
folks didn't have the interest or knowledge to roll their own. They just wanted
the "appliance" computer.
> You know, I've always wondered why the Motorola 8-bit processors didn't
> take over the 8-bit desktop market of their day? The 6809 was a
> fine bit of work, but almost the only mass-market machine I can think of that
> used it was the RadioShack Color Computer (aside from those
> little ads in the back of BYTE advertising Gimix systems). And even a CP/M
> enthusiast of the time would have recognized that FLEX had many advantages.
Too little too late. I don't know why I'm attracted so frequently to the "losing
side" (see my collection of Beta video recorders ;-) Although FLEX had some
good applications for spread sheet and word processing (the "killer aps" of their
day), they came after they were already available for the CP/M world, not to
mention the early appliance machines like the Apple II and the RS Model 1.
Although there was (re GIMIX) a growing "industrial base" for FLEX/68XX,
the arrival of the IBM PC rapidly killed it off just like it got the CP/M world.
Jay R. Ashworth
Ben Harris <bj...@cus.cam.ac.uk> wrote:
> In article <slrn981se...@dorothy.msas.net>,
> Jay R. Ashworth <j...@baylink.com> wrote:
> >I think a very good case can be made that Linux runs on more processor
> >types than any other single program now or before.
>
> That's blatantly untrue. The "hello, world" program from K&R has probably
> been run on every system with a C compiler (which includes plenty of
> systems Linux won't run on).
Ok.
Forgive me. "Progam" wasn't sufficiently restrictive, though I *did*,
in the very next paragraph (which you've so kindly elided) specify
further that I was referring to "the Linux kernel", in its class.
> As far as real-world programs go, C-Kermit and Info-Zip are probably about
> as portable (and ported) as you'll get. Even GCC supports more CPU types
> than Linux.
To target? Or to run on? I'd forgotten Kermit too, of course, but, as
noted, I *was* talking about OS's.
> If you insist on comparing OSes, then I think NetBSD gives Linux a run for
> its money (last I counted they each did ten CPUs as standard).
Are they tied? Good. Diversity Is Good. I'm not a bigot.
And, as I think I already noted, FORTH blows them both out of the
water, possibly by as much as an order of magnitude, but there's not
enough commonality for the point I was making.
Excellent observations, though. How many known ports of C-Kermit are
there?
Cheers,
-- jra
--
Jay R. Ashworth j...@baylink.com
Member of the Technical Staff Baylink
The Suncoast Freenet The Things I Think
Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
Jay R. Ashworth
gla...@glass2.lexington.ibm.com <gla...@glass2.lexington.ibm.com> wrote:
> >I think you miss one of the fundamental points about Linux, that
> >IT IS PORTABLE AMONG MULTIPLE PLATFORMS. (OK, it wasn't portable
> >at first, but now it's pretty good.) It runs on 80x86, it runs
> >on Alpha, it runs on SPARC, it runs on lots of other CPU's,
> >and it runs with lots of different peripherals that hook up to
> >these CPU's, and if you want to write a driver for your own custom
> >device it's all documented very clearly.
>
> (Shameless plug warning:)
> Hey, you left out the IBM ESA/390 (or, should I say IBM z/Architecture)
> mainframes! (Ok, so I work for them.).
He did, but I didn't.
Let me ask you a quesiton, sirrah.
About 3 months ago, I had to pass on a small (1 rack) 390, that
eventually went for about $1500 (no, that is *not* a typo).
If I actually bought something like that, and I wanted to run a bunch
of Linux/390 instances on it, just exactly how high up Sisyphus' hill
am I gonna have to roll rocks?
Jay R. Ashworth
Roger Johnstone <roj...@mac.com> wrote:
> > In article <slrn981sj...@dorothy.msas.net>, Jay R. Ashworth
> >> Apple's OpenFirmware: they wrote the on-card bios in *FORTH*, and put a
> >> FORTH interpreter on the motherboard. This allows you to make the
> >> cards processor independent while still allowing intelligence on each
> >> card.
>
> later. It is a brilliant idea though. Being able to plug a card into any
> type of computer with any type of processor and have it be able to use it
> without loading drivers. Very important for video, storage and network cards
> which have to work before the OS loads.
Ah. My apologies.
> The big problem for Open Firmware is the PC clones. Since they all use the
> same processor/architecture/ROM they don't really need Open Firmware, and
> since they're the majority of the market hardly any cards are made with OF.
Perhaps the movement of Linux into alternative processors will make it
commercially feasible to sell motherboards with those chips, and thus
pull OF along with it. We can hope, no?
> http://www.openfirmware.org/
Where it belongs; I should ahve looked. You live and learn. Thanks.
Cheers
Jay Maynard
>About 3 months ago, I had to pass on a small (1 rack) 390, that
>eventually went for about $1500 (no, that is *not* a typo).
That's not surprising, either.
>If I actually bought something like that, and I wanted to run a bunch
>of Linux/390 instances on it, just exactly how high up Sisyphus' hill
>am I gonna have to roll rocks?
You'll need VM/ESA on it, as well; beyond that, it's a matter of disk space
and connectivity.
One gotcha: Linux/390 requires a G5 or better processor (IIRC). That means
relatively recent 390.
Charlie Gibbs
writes:
>@mb.sympatico.ca wrote:
>
>> >IIRC Kermit could use the serial ports through BDOS and I don't
>> >remember doing any hacking to Kermit.
>>
>> Neat. I never used Kermit under CP/M though I'd seen it. I don't
>> quite see how it could have worked with *no* hacking. Certainly
>> you would have had to use an external program to set up the port
>> baud rate, etc.
>
>I have used Kermit under CP/M to communicate with both DOS and CCP/M
>machines - it works a treat and no 'tweaking' was required. I didn't
>need an external program to set the speed - it's a function within
>Kermit.
After having a lot of fun with MODEM7 I moved up to MEX, which
was quite sophisticated for its day. The big difference between
programs then and now is that back then you were given enough
technical information to do your own customizing, rather than
depending on hardware vendors to supply you with a driver disk
because the specs are either unavailable, outrageously expensive,
or provided only under a non-disclosure agreement.
I hacked a relay into the line side of my little 300-baud manual
modem, driven by DTR. A friend modified the MEX overlay that
was driving our 8251 USARTs so that it could toggle DTR to make
the modified modem pulse-dial. MEX allowed you to configure the
pulse rate - our central office was upgrading at the time, and
afterwards I was able to get away with 20 pulses per second, which
was starting to approach Touch-Tone speeds. Common MEX code took
it from there - I was able to turn it loose with a list of BBSes
and it would hammer away at each one in turn - assuming that a
lack of carrier meant that the line was busy - until it finally
got through to one, at which point it would start beeping to
attract my attention from whatever else I was doing. This was
great for getting into busy BBSes, and few terminal emulators
before or since have offered such flexibility.
--
cgi...@sky.bus.com (Charlie Gibbs)
Remove the first period after the "at" sign to reply.
Ben Harris
Jay R. Ashworth <j...@baylink.com> wrote:
>On 7 Feb 2001 12:16:58 GMT,
> Ben Harris <bj...@cus.cam.ac.uk> wrote:
>> In article <slrn981se...@dorothy.msas.net>,
>> Jay R. Ashworth <j...@baylink.com> wrote:
>> >I think a very good case can be made that Linux runs on more processor
>> >types than any other single program now or before.
>>
>> That's blatantly untrue. The "hello, world" program from K&R has probably
>> been run on every system with a C compiler (which includes plenty of
>> systems Linux won't run on).
>
>Ok.
>
>Forgive me. "Progam" wasn't sufficiently restrictive, though I *did*,
>in the very next paragraph (which you've so kindly elided) specify
>further that I was referring to "the Linux kernel", in its class.
>
>> As far as real-world programs go, C-Kermit and Info-Zip are probably about
>> as portable (and ported) as you'll get. Even GCC supports more CPU types
>> than Linux.
>
>To target? Or to run on?
Both. I believe it will run on, and target, every CPU supported by either
Linux or NetBSD (both rather depend on it) as well as several that neither
supports.
>Excellent observations, though. How many known ports of C-Kermit are
>there?
<URL:http://www.columbia.edu/kermit/ckermit.html> claims:
* Every known variety of UNIX
* (Open)VMS on VAX and Alpha
* Data General AOS/VS II
* Stratus VOS
* Microware OS-9
* Plan 9 from Bell Labs
* QNX4 and Neutrino
* The Commodore Amiga
<URL:http://www.freesoftware.com/pub/infozip/> claims for Info-ZIP:
Info-ZIP supports hardware from microcomputers all the way up to Cray
supercomputers, running on almost all versions of Unix, VMS, OS/2, Windows
NT/ 9x (a.k.a. Win32), Windows 3.x, Windows CE, MS-DOS, AmigaDOS, Atari
TOS, Acorn RISC OS, BeOS, Mac OS, SMS/QDOS, MVS and OS/390 OE, VM/CMS,
FlexOS, Tandem NSK and Human68K (Japanese). There is also some (old)
support for LynxOS, TOPS-20, AOS/VS and Novell NLMs.
Steve O'Hara-Smith
> If you think that buying a PC-clone motherboard
> and plugging in a couple of boards so that you can boot MS-DOS
> is "building a computer", us folks in alt.folklore.computers have
> a few things to teach you :-).
Hmm, solder burns, how much wire-wrap radiates and my favorite the
insertion force required for an original ceramic pack 68000 into a new
socket :)
PS: I never did find the extraction force - thankfully.
Steve O'Hara-Smith
> I think a very good case can be made that Linux runs on more processor
> types than any other single program now or before. The Linux kernel
Gcc is an obvious counter example :)
mb.sympatico.ca
>@mb.sympatico.ca writes:
>> Acck! It's been a long time. The ironic thing is that I actually
>> have a 5 1/4 GCR diskette for an Apple II CP/M card ( Applesoft
>> FORTRAN) and I'd completely forgotten about this, likely the most
>> common GCR implementation.
>
>If it's Apple Fortran, it doesn't run on the CP/M card.
>
>If it's Microsoft Fortran, it does.
>
>AFAIK, there was not any product called "Applesoft FORTRAN".
I ran down to the basement and dug up the rather nice green 3-ring binder.
The product that I have is called "Microsoft Fortran-80 Compiler for
Apple II and IIe with SoftCard(TM) and CP/M (R)-80".
I did look at the disks and they are 5 1/4 inch, so at least I got that part right.
Bill
mb.sympatico.ca
two different points of view. I may be playing up the commercial, appliance-operator
type of personal computer user. I had a toe in the waters of the computer
hobbyist for several years, though - I'd customized a modem overlay for
a new CPU, I built a serial port for a machine that didn't have one from the
factory ( Coleco Adam), and even struggled with updating the Osborne
1 to read double-sided disks ( this meant re-generating source for the BIOS
from the EPROMS, guided by the technical manual, then burning new EPROMS - the
standard "assemble-burn-boot-test-curse-bleach" cycle got to be very
familiar to me).
I wouldn't dispute that CP/M was the better vehicle for learning about the machines.
For actually getting anything done, you wanted a more standardized and more
ecnomical hardware base. The same guy who restores '57 Chevies on weekends may
also drive a leased Corolla to get to work!
In <3A8043C2...@trailing-edge.com>, Tim Shoppa <sho...@trailing-edge.com> writes:
>@mb.sympatico.ca wrote:
>> >> Consider disk formats. Again, no two CP/M machines had the same disk format.
>
>If you got a "packaged" microcomputer system with the BIOS already
>written and installed, IMHO you got screwed. Having the source to
>your CP/M BIOS is absolutely essential to understanding how
>versatile the thing is.
>
>
>> The world was full of 4 MHZ Z80 systems that all did pretty much the same thing
>> from the user's point of view, but that, for baffling technical reasons incapable of
>> being explained to the people who authorized purchase orders, were incapable of
>> using each other's periperhals, software, or cables without paying even *more* money
>> to someone who was a champion of the particular obscure platform.
>
>Again, I think you got in too late - or with a seriously lame
>attitude - to know what was fun about, for example, S-100 computing
>with CP/M in the late 70's. The experiences I gained weren't ones
>that I'd wish on J Random Luser, but they were very valuable to me.
I never owned an S100 machine, but I did have the chance to use one for a
short time - I had to fix it, as a matter of fact, since the memory refresh wasn't
working. I had manuals, and a scope, and could tell from the ROM monitor
testing that if I tested, say, 1 K of memory at a time, it refreshed OK and passed
the test - but if I tested 32 K, by the time the test routine walked around to the
end of the area, the first few bytes had gone random again. I was rather
pleased that I was able to get the machine stable again, although I never
did figure out how to get the 8 inch hard drive running ( this was the user
group's ex-BBS machine and I undertook to salvage all the backed up software
and move them to a more modern machine via serial port - nothing else we owned
would read the 8 inch hard-sectored disks. I think the machine was called a Vector
Graphic though I no longer recall - it didn't have vector graphics, but could do
bit-mapped raster graphics!).
This whole experience is completely impossible in the world of PC descendants...
there's no point in fixing motherboards when you can buy a new MB for less
than what a few hours troubleshooting time would cost. Again, computers
were fantastically expensive in those days.
The spirit of the CP/M enthusiast lives on, of course, in terms of the Linux
hordes. I don't think we're in any serious danger of all becoming appliance
operators...there will always be those who hack ( in the true positive sense).
Thankfully we no longer have to teach secretaries hexidecimal in order to
install WordStar.
Bill
Richard Plinston
>
> > As I said before, CP/M provided a valuable service at the time - that time was
> > between 1977 and 1981. After that, DRI did not bring it up to date and did
> not
> > price it competitively with MS DOS for the 8088/8086 world,
You are incorrect about the 'DRI ... did not price it competitively'.
When DRI took action against SCP, MS and IBM over copyright infringement
by PC-DOS 1.x part of the out of court settlement was that IBM would
offer CP/M-86 alongside PC-DOS. However, no discussion about pricing
occurred and so _IBM_ set the PC-DOS price at $70 and CP/M-86 at $250
(or so). When DRI tried to get IBM to set the price at a competitive
level they refused and also refused to offer the 1.1 version keeping the
1.0 version at the original price in their pricebook. This may be seen
as payback from IBM for losing the case.
DRI brought out version 1.1 but could not sell it through IBM retailers
and had to distribute via the add on resellers where it was priced at
$70.00 or so. However, by then most IBM PC buyers bought their OS with
the machine and IBM only offered the obsolete version at extortionate
prices.
Within the same timescale DRI had released Concurrent-CP/M-86 (version 1
in Sept 1981) with mult-tasking and (later) multi-user. Note also that
CP/M-86 could support hard drives while this was not available with
PC-DOS/MS-DOS until version 2 a couple of years later.
They had DR-NET peer networking, Star-NET serial remote access. While
PC-DOS sites were passing floppies around DRI sites used hard drives,
shared files on multi-user and networked systems.
> > Though, come to think of it, maybe the most valuable service CP/M was to
> > provide was the illustration that commands like
>> A> pip b:foo.bar=a:file.txt[Zu100]
> >
> > was NOT going to be the future of mass-market computing.
Do you think that EDLIN was going to 'the future' ?
Bill Marcum
Richard Plinston wrote in message <3A819922...@Azonic.co.nz>...
>As a comparison to CP/M Plus, MS introduced MSX-DOS as an 8bit
>competitor to CP/M and this failed to inspire anyone. The answer really
>is that those who wanted to keep 8 bit systems were mainly quite happy
>to stay with CP/M 2.2. They did not want an 'up to date' system.
MSX was also competing with the C-64 and other 8-bit computer systems
(or trying to compete, with absolutely no success in the US market).
Richard C. Steiner
>I think a very good case can be made that Linux runs on more processor
>types than any other single program now or before.
NetBSD? :-)
--
-Rich Steiner >>>---> rste...@visi.com >>>---> Eden Prairie, MN
Written online using slrn 0.9.5.4!
The Theorem Theorem: If If, Then Then.
Bill Pechter
Charlie Gibbs <cgi...@sky.bus.com> wrote:
>In article <3A804535...@fcs.eu.com> bws...@fcs.eu.com (B W Spoor)
>writes:
>
>was quite sophisticated for its day. The big difference between
>programs then and now is that back then you were given enough
>technical information to do your own customizing, rather than
>depending on hardware vendors to supply you with a driver disk
>because the specs are either unavailable, outrageously expensive,
>or provided only under a non-disclosure agreement.
>
Let's hear it for Ron Fowler, Author of Mex and MexPlus (and MexPC).
The program was configurable, documented and worked without being
bloated.
MexPlus gave VT100 emulation to those of us without VT100 compatible
tubes, added long kermit capabilities and Ymodem...
I still have the PC version and CP/M version. It was amazing to do menu
driven screens with cursor control under CP/M which was ProcommPlus users
of the same period could do.
The VT100 emulation worked pretty well with DEC Vaxen... I tested it
against the stuff at DEC Princeton and used it daily on my Zorba
which served as my download from the DEC20 market machine.
And both the PC version and CP/M version were in assembler!!!
Ron -- if you're lurking -- great job.
Bill
--
--
bpec...@monmouth.com | FreeBSD since 1.0.2, Linux since 0.99.10
Brainbench MVP | Unix Sys Admin since Sys V/BSD 4.2
Unix Sys.Admin. | Windows System Administration: "Magical Misery Tour"
Marco S Hyman
> I think a very good case can be made that Linux runs on more processor
> types than any other single program now or before. The Linux kernel
More than NetBSD? There were over 30 distinct platforms it ran on
last time I looked, even on a dreamcast. Some of the older versions
of UNIX ran on a lot of platforms, too.
// marc
Charlie Gibbs
gla...@glass2.lexington.ibm.com (glass2) writes:
>Most of us lusted after systems with floppy drives, and would
>have given someone's right arm for a system that would run CP/M.
It was a long time before I could afford to hang a couple of
8-inch floppy drives on my IMSAI, but I wanted to run CP/M
programs anyway. So I wrote an emulator that intercepted
BIOS and BDOS calls and simulated them on my cassette-based
system. It actually worked pretty well for programs that
didn't use random-access files. I could load the emulator,
use it to load the standard CP/M MODEM7 program, and I'd be
online (fifteen minutes later :-). Mind you, downloads were
a hassle - MODEM7 would download data in 16K chunks, and it
took so long to write 16K to tape that the other end would
time out. (Fortunately a retry would work.)
Jay R. Ashworth
Jay Maynard <jmay...@thebrain.conmicro.cx> wrote:
> On Wed, 07 Feb 2001 18:01:04 GMT, Jay R. Ashworth <j...@dorothy.msas.net> wrote:
> >About 3 months ago, I had to pass on a small (1 rack) 390, that
> >eventually went for about $1500 (no, that is *not* a typo).
>
> That's not surprising, either.
Really? That's not an unreasonable price? Would a machine like that
be provisioned to bring back up?
> >If I actually bought something like that, and I wanted to run a bunch
> >of Linux/390 instances on it, just exactly how high up Sisyphus' hill
> >am I gonna have to roll rocks?
>
> You'll need VM/ESA on it, as well; beyond that, it's a matter of disk space
> and connectivity.
Well, I could run, like, 8 or 16 instances on the bare metal, right/
How much would IBM nick me for VM, *only* to run Linux? I know they
have a deal on extra CPU's...
> One gotcha: Linux/390 requires a G5 or better processor (IIRC). That means
> relatively recent 390.
Hmmm...
The underlying question here, of course, is: is this worth it?
Richard Drushel
: I built a serial port for a machine that didn't have one from the
: factory ( Coleco Adam)
Hi Bill, any details, please? The ADAM is my hobby...and my users
group is organizing ADAMcon 13 for Cleveland this July (official announcement
and details to be posted soon). There are 2 non-Coleco serial port types that
I know about; I'd love to hear more.
*Rich*
--
Richard F. Drushel, Ph.D. | "Aplysia californica" is your taxonomic
Department of Biology, Slug Division | nomenclature. / A slug, by any other
Case Western Reserve University | name, is still a slug by nature.
Cleveland, Ohio 44106-7080 U.S.A. | -- apologies to Data, "Ode to Spot"
jmfb...@aol.com
/BAH
Subtract a hundred and four for e-mail.
Brian Inglis
Ashworth) wrote:
>On Wed, 07 Feb 2001 22:41:36 +1300,
> Roger Johnstone <roj...@mac.com> wrote:
>> > In article <slrn981sj...@dorothy.msas.net>, Jay R. Ashworth
>> >> Absolutely my favorite dead computing idea of the last 20 years was
>> >> Apple's OpenFirmware: they wrote the on-card bios in *FORTH*, and put a
>> >> FORTH interpreter on the motherboard. This allows you to make the
>> >> cards processor independent while still allowing intelligence on each
>> >> card.
>>
>> Yikes, it was Sun who started with Open Firmware. Apple hopped on board
>> later. It is a brilliant idea though. Being able to plug a card into any
>> type of computer with any type of processor and have it be able to use it
>> without loading drivers. Very important for video, storage and network cards
>> which have to work before the OS loads.
>
>Ah. My apologies.
>
>> The big problem for Open Firmware is the PC clones. Since they all use the
>> same processor/architecture/ROM they don't really need Open Firmware, and
>> since they're the majority of the market hardly any cards are made with OF.
>
>Perhaps the movement of Linux into alternative processors will make it
>commercially feasible to sell motherboards with those chips, and thus
>pull OF along with it. We can hope, no?
DEC's solution to building Alpha desktops with PCI cards was to
build a 386 emulator into the Linux initialization code -- see
the Linux Journal articles on the Alpha port at
<URL:http://www.linuxjournal.com>.
>> http://www.openfirmware.org/
>
>Where it belongs; I should ahve looked. You live and learn. Thanks.
>
>Cheers
>-- jra
Thanks. Take care, Brian Inglis Calgary, Alberta, Canada
--
Brian_...@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca)
use address above to reply
Jay Maynard
>On 7 Feb 2001 18:42:28 GMT,
> Jay Maynard <jmay...@thebrain.conmicro.cx> wrote:
>> On Wed, 07 Feb 2001 18:01:04 GMT, Jay R. Ashworth <j...@dorothy.msas.net> wrote:
>> >About 3 months ago, I had to pass on a small (1 rack) 390, that
>> >eventually went for about $1500 (no, that is *not* a typo).
>> That's not surprising, either.
>Really? That's not an unreasonable price? Would a machine like that
>be provisioned to bring back up?
Depends on what all was in that rack. IBM has made 390s that would fit in a
rack with room for tape and DASD left over. Not the fastest boxes in the
line, but not bad for a home box.
>Well, I could run, like, 8 or 16 instances on the bare metal, right/
That assumes the machine has LPAR functionality. I dunno at what level that
became standard across the line.
>How much would IBM nick me for VM, *only* to run Linux? I know they
>have a deal on extra CPU's...
IBM has a new product called VIF (Virtual Image Facility) that will run just
Linux guests. I dunno how much it costs, but I assume it's not trivial.
VM/ESA, the full version, runs somewhere in the $20-30K range.
>The underlying question here, of course, is: is this worth it?
Dunno...I paid $1300 for a P/390 board set, for pure hack value...
Eric Sosman
>
> If you got a "packaged" microcomputer system with the BIOS already
> written and installed, IMHO you got screwed. Having the source to
> your CP/M BIOS is absolutely essential to understanding how
> versatile the thing is.
De gustibus non disputandum est, but I think I got far more
delight from disassembling my Kaypro BIOS (both the RAM- and
ROM-resident parts) than any amount of reading the source would
have provided.
By the way, I remember wondering while disassembling at the
strange "feel" of the code. Some of it was careful, dense, and
idiosyncratic: exactly the sort of thing a human turns out when
writing assembly language. But some of the code was patterned,
a little obese, and repetitious: compiler output, almost surely.
And the BIOS itself seemed to have two distinct chunks (I'm
speaking of two chunks in the RAM-resident portion, not the
division between RAM and ROM), with a jump table sitting between
them to vector calls back and forth between the chunks.
My best guess at the time was that the BIOS must have been
written in a higher-level language (PL/M?) which allowed embedded
assembly code (or else somebody massaged the compiler's output);
this would explain the curious mix of one-of-a-kind and
cookie-cutter coding. The presence of the jump table suggests
that there may have been two separately-compiled modules which
were then glued together without benefit of anything we'd call
a "linker" nowadays.
Can anyone who actually Knows confirm or refute or amplify
any of these suppositions?
Eric Smith
>> of Linux/390 instances on it, just exactly how high up Sisyphus' hill
>> am I gonna have to roll rocks?
>
> You'll need VM/ESA on it, as well; beyond that, it's a matter of disk space
> and connectivity.
>
> One gotcha: Linux/390 requires a G5 or better processor (IIRC). That means
> relatively recent 390.
Are you saying that on a G5 you can run Linux on the bare metal (no VM/ESA),
or are you saying that you need to have a G5 *AND* VM/ESA?
If Linux can run on the bare metal on a G5, is there anything fundamentally
difficult about making it run on the bare metal on earlier processors?
Floyd Davidson
>Tim Shoppa wrote:
>>
>> If you got a "packaged" microcomputer system with the BIOS already
>> written and installed, IMHO you got screwed. Having the source to
>> your CP/M BIOS is absolutely essential to understanding how
>> versatile the thing is.
>
> De gustibus non disputandum est, but I think I got far more
>delight from disassembling my Kaypro BIOS (both the RAM- and
>ROM-resident parts) than any amount of reading the source would
>have provided.
That is true. I had an '83 Kaypro-4 that eventually had a 68000
co-processor inside the box too. The RAMDISK, using the 68000
and 768K of RAM, was really neat once I put the interface code
for it into the ROM. That and code for four quad-density half
height floppy drives. Eventually I had code in ROM to read
several different floppy formats too. And my keyboard had
several extra keys. The last thing I did was put in 256K RAM
chips and set it up for bank switching, but all I ever did with
that was verify that it worked. (AT&T dumped the 3B1 UnixPC's
on the market and like a lot of others, my dreams were realized
with a UNIX box at home. The Kaypro went to the dump.)
> By the way, I remember wondering while disassembling at the
>strange "feel" of the code. Some of it was careful, dense, and
>idiosyncratic: exactly the sort of thing a human turns out when
>writing assembly language. But some of the code was patterned,
>a little obese, and repetitious: compiler output, almost surely.
>And the BIOS itself seemed to have two distinct chunks (I'm
>speaking of two chunks in the RAM-resident portion, not the
>division between RAM and ROM), with a jump table sitting between
>them to vector calls back and forth between the chunks.
>
> My best guess at the time was that the BIOS must have been
>written in a higher-level language (PL/M?) which allowed embedded
>assembly code (or else somebody massaged the compiler's output);
>this would explain the curious mix of one-of-a-kind and
>cookie-cutter coding. The presence of the jump table suggests
>that there may have been two separately-compiled modules which
>were then glued together without benefit of anything we'd call
>a "linker" nowadays.
>
> Can anyone who actually Knows confirm or refute or amplify
>any of these suppositions?
Eventually the source files were released, so I've seen the
original code, though by that time I wasn't really very
interested in the original. It appeared not to be what you are
suspecting though, if I remember right. But they did have a
great deal of the example BIOS from DRI's book in there, even
when it was not being used.
The oddest part was the video. Eubanks (I can't remember his
first name) supposedly was in a rush to get it out the door,
and couldn't quickly figure out how to scroll the display using
a circular buffer and pointers. Instead he had it actually
shifting the entire screen full of data by one line to scroll.
That caused it to be excessively slow, making it impossible to
keep up with a modem going faster than 1200 bps (unless the
clock speed was hacked too, and then it could do 2400 bps).
Another mistake was one of the video attributes, which was
supposed to emulate an ADM3A, used a decimal 11 instead of
hex 11 (or else it was the other way around).
--
Floyd L. Davidson <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska) fl...@barrow.com
Jay Maynard
wrote:
>Are you saying that on a G5 you can run Linux on the bare metal (no VM/ESA),
>or are you saying that you need to have a G5 *AND* VM/ESA?
You can run it on the bare metal, if you only want to run a small number of
instances and dedicate entire DASD volumes and such to each instance. VM/ESA
removes both those restrictions.
>If Linux can run on the bare metal on a G5, is there anything fundamentally
>difficult about making it run on the bare metal on earlier processors?
Yes. It'd take a bunch of work. The gcc used to build Linux for the 390 uses
relative branch instructions that were introduced on the G5 (again, IIRC,
but it's definitely not the first 390 definition). You'd have to make gcc
allocate and initialize base registers, and then make code deep in Linux do
the right thing.
Paul Repacholi
> <URL:http://www.columbia.edu/kermit/ckermit.html> claims:
>
> * Every known variety of UNIX
> * (Open)VMS on VAX and Alpha
> * Data General AOS/VS II
> * Stratus VOS
> * Microware OS-9
> * Plan 9 from Bell Labs
> * QNX4 and Neutrino
> * The Commodore Amiga
And the PDP-11 OSs, unix RSX, RT, RSTS, many IBM sytems and foo knows
what else. Oh, and Tops-10 and TOPS-20. New 20 version just released
a few weeks ago. It is THE canonical KERMIT.
--
Paul Repacholi 1 Crescent Rd.,
+61 (08) 9257-1001 Kalamunda.
West Australia 6076
Raw, Cooked or Well-done, it's all half baked.
Eric Chomko
: Don Stokes <d...@news.daedalus.co.nz> wrote:
: > In some ways, it makes sense to put the BIOS in ROM; that way the stuff
: > that could be different across platforms is supplied with the hardware
: > and the OS doesn't need to worry about it.
: Absolutely my favorite dead computing idea of the last 20 years was
: Apple's OpenFirmware: they wrote the on-card bios in *FORTH*, and put a
: FORTH interpreter on the motherboard. This allows you to make the
: cards processor independent while still allowing intelligence on each
: card.
: It was an absolutely *elegant* idea, probably the most elegant one ever
: to come out of Cupertino IMHO, and I think it's a damned shame it never
: caught on.
No worse than the DEC Alpha having PALs (ROMs) that were customized to the
OS of your choice: Win NT, UNIX (OSF) or VMS. And we all know what
happened to DEC.
Eric
Eric Smith
> Dunno...I paid $1300 for a P/390 board set, for pure hack value...
Is it documented well enough that something could be done with it
*without* using the OS/2-based host software? In other words, is
there any chance of getting it to run in an Linux/x86 host system?
Jay Maynard
>No worse than the DEC Alpha having PALs (ROMs) that were customized to the
>OS of your choice: Win NT, UNIX (OSF) or VMS. And we all know what
>happened to DEC.
You're confusing two different things here. PALcode, on the Alpha, has
nothing to do with PALs or ROMs. It's the Privileged Applications Layer,
where some operations must be executed in order to preserve the integrity of
the machine. PALcode for OpenVMS, Unix, and NT are different because
different low-level operations are needed for each OS; the PALcode is loaded
by the boot loader before the OS itself is loaded.
Nopw, it is true that some Alphas - but not all, by any means - had ROM
BIOSes that would only work with NT and Linux, but that has nothing to do
with the PALcode involved.
Jay Maynard
wrote:
I doubt it. I know of no low-level docs. Mine's not running under OS/2,
though; I have it installed in a Micro Channel RS/6000 and run it under AIX.
The RS/6000 7009-C10 it's installed in only cost me $100, so it's not a big
deal.
Eric Sosman
> [concerning the 1983 Kaypro-4]
> The oddest part was the video. Eubanks (I can't remember his
> first name) supposedly was in a rush to get it out the door,
> and couldn't quickly figure out how to scroll the display using
> a circular buffer and pointers. Instead he had it actually
> shifting the entire screen full of data by one line to scroll.
> That caused it to be excessively slow, making it impossible to
> keep up with a modem going faster than 1200 bps (unless the
> clock speed was hacked too, and then it could do 2400 bps).
I had the 1984 model, by which time they'd figured out how
to scroll the screen with pointers and offsets. The screen was
still slow, though, because the ROM BIOS code insisted on writing
each character in "random access" mode: send the address, wait,
send the character data. (In fact, it did this twice because
it also wrote an attribute byte to a non-nearby address.)
I hadn't the equipment (or knowledge) to change the ROM BIOS,
but even so I managed to get my Kaypro to talk with an external
modem at 19200 bps. Had to use RTS/CTS a lot if the incoming
characters were going to the screen or whenever I had to fire up
the floppy drive, but it ran Original Xmodem at a pretty good clip.
Many happy hours of hacking went into that highly-customized and
completely non-portable hunk of code.
Jay R. Ashworth
Jay Maynard <jmay...@thebrain.conmicro.cx> wrote:
> >> >About 3 months ago, I had to pass on a small (1 rack) 390, that
> >> >eventually went for about $1500 (no, that is *not* a typo).
> >> That's not surprising, either.
> >Really? That's not an unreasonable price? Would a machine like that
> >be provisioned to bring back up?
>
> Depends on what all was in that rack. IBM has made 390s that would fit in a
> rack with room for tape and DASD left over. Not the fastest boxes in the
> line, but not bad for a home box.
Indeed.
> >Well, I could run, like, 8 or 16 instances on the bare metal, right/
>
> That assumes the machine has LPAR functionality. I dunno at what level that
> became standard across the line.
Oh, ok. Yeah, I knew about LPAR, but I didn't know it was new.
> >How much would IBM nick me for VM, *only* to run Linux? I know they
> >have a deal on extra CPU's...
>
> IBM has a new product called VIF (Virtual Image Facility) that will run just
> Linux guests. I dunno how much it costs, but I assume it's not trivial.
> VM/ESA, the full version, runs somewhere in the $20-30K range.
I'll try and find an IBM sales droid to ask.
If I can han a couple hundred people on a $3000 or so 390, I can get my
boss to buy the next one that flies by, is the object of this inquiry.
> >The underlying question here, of course, is: is this worth it?
>
> Dunno...I paid $1300 for a P/390 board set, for pure hack value...
I love that guy who's selling CPUs with the nasty notice to IBM
security... :-)
Lee Hart
>> But mainly I would like to say thank you for writing your piece. It
>> made me pause and think a little about my obsession with CP/M, which
>> can only ever help.
@mb.sympatico.ca wrote:
> I, too, was a big CP/M buff probably till long after it was no longer
> a commerically viable software platform. It still bothers me a little
> bit that I have nearly *no* idea what is going on inside the Wind 98
> box I use at work, or the Win 98 box I carry around, and only a little
> more about what is going on inside this OS/2 box I'm using to type
> this. I will never again have such intimate mastery of the inner
> details of the machine...
...and that is the answer for me. CP/M is important because it showed us
all that "you can do it." You can have your own computer. You can
understand it, change it, program it to do what YOU want it to do, the
way YOU want it done. It truly allowed us to have a PERSONAL computer.
This attitude enabled a wonderful burst of creativity and enthusiasm,
the results of which we all benefit from today.
I feel the IBM "PC" phenomena stole the dream. It's not my personal
computer; it's only what IBM / Intel / Microsoft et.al. want me to have.
I can't understand it, nor program it, nor make it do what I want. It's
an appliance, whose inner workings are deliberately arcane and
inscrutible.
Despite all the latest PC's speed and power and memory capacity, they
don't really do anything I want to do any faster than my old CP/M
machines. They don't make answering my email any easier, or print any
faster, or balance my checkbook any more accurately, or remember my
Christmas card mailing list any better. It's all glitz and glamour,
which I am forced to take whether I want it or not.
--
Lee A. Hart Ring the bells that still can ring
814 8th Ave. N. Forget your perfect offering
Sartell, MN 56377 USA There is a crack in everything
leeahart_at_earthlink.net That's how the light gets in - Leonard Cohen
Chris Baird
> "feel" of the code. Some of it was careful, dense, and idiosyncratic:
> exactly the sort of thing a human turns out when writing assembly
> language. But some of the code was patterned, a little obese, and
> repetitious: compiler output, almost surely.
Yep, parts were written in PLM. The unofficial CP/M sites
like http://www.cpm.z80.de/ have all the gritty details,
sources for PLM compilers, etc.
--
Chris,,
Eric Chomko
: On 9 Feb 2001 18:48:22 GMT, Eric Chomko <cho...@IDT.NET> wrote:
: >No worse than the DEC Alpha having PALs (ROMs) that were customized to the
: >OS of your choice: Win NT, UNIX (OSF) or VMS. And we all know what
: >happened to DEC.
: You're confusing two different things here. PALcode, on the Alpha, has
: nothing to do with PALs or ROMs. It's the Privileged Applications Layer,
: where some operations must be executed in order to preserve the integrity of
: the machine. PALcode for OpenVMS, Unix, and NT are different because
: different low-level operations are needed for each OS; the PALcode is loaded
: by the boot loader before the OS itself is loaded.
Thanks, for the dictinction. I never worked on a DEC ALPHA but did attend
a seminar.
: Nopw, it is true that some Alphas - but not all, by any means - had ROM
: BIOSes that would only work with NT and Linux, but that has nothing to do
: with the PALcode involved.
Yes, the local surplus electronics guy sells that exact system. If I had
more room I'd get one.
Thanks, for clearing up PAL and ROMs for me.
Eric
Paul Repacholi
> No worse than the DEC Alpha having PALs (ROMs) that were customized to the
> OS of your choice: Win NT, UNIX (OSF) or VMS. And we all know what
> happened to DEC.
WHAT! Is this a reference to pal *CODE* by any chance?
PLZI
"Richard Plinston" <rip...@Azonic.co.nz> wrote in message
news:3A819922...@Azonic.co.nz...
> DRI also had GEM being delivered while MS was still talking about its
> vapourware Windows 1, many say that MS disn't catch up to the original
> GEM until Windows 2.
This is something I really can agree with. I remember using (and installing
and supporting) a Xerox Ventura DTP system, which was built on GEM. It was a
"ms-dos" -program, ms-dos -part being that the dos provided the program
loading, but after that the GEM with single-mined task of running Ventura
took over.
It was blindingly fast on 286. It could draw the whole page of text, couple
of autodesk dxf line drawings and bitmapped graphics on the screen in a
fraction of a second.
A few years later I saw the first windows versions. On a 3.0, 386SX/25MHz.
You could go for a coffee inbetween the page redraws. Well, not really, but
it was slow.
And yes, I am a Windows advocate. These days. :)
- PLZI
Brian {Hamilton Kelly}
> I suspect many people moved to the IBM and/or compatibles because of the
> Laser
> Printer. You could now produce type-set quality output, soon followed by
> bit
> mapped graphics, that previously had been virtually impossible for under
> maybe
> $100,000 worth of equipment.
I cannot perceive laser printer availability as being a force for the
growth of PCs. I bought my first laser printer, a DEC LN03, in around
1984. This was not a personal purchase, but for work, where it was
attached to a VAX-11/730, replacing the Versatec electrostatic printer/
plotter that had hithertoo been our HQ output device.
The price was almost GBP5000 + VAT.
At that time at work we would never have countenanced the concept of
connecting a laser printer to one of our (almost equally expensive
PC/ATs); that's what Epson printers were for. Anyway, no one expected to
do any *real* work on a PC; for that one ran Kermit and used it as a
terminal to the VAX.
> Laser printing changed all that. But with a laser printer costing $3,000
> to
> $5,000 not many showed up in the home. But there were some.
I bought my first laser printer for personal use in 1989. It's a HP
LaserJet IIP, and cost me over a thousand pounds, and that was after
getting it at a discount through a university. I'm pleased to say that
it is still going strong, having needed the replacement of two parts in
the past year (readily obtained from <http://www.fixyourownprinter.com>
to give Moe a little plug).
I still saw this as an extravagance at the time, and very few small
office setups were using lasers even then (except for the Mac
fraternity).
> > No-one remembers how fantastically *expensive* computers were only 20 years
> > ago, especially considering what very little they could actually do. And as a
> > result (by today's standards) no-one *had* computers n the home in those days.>
> I had several. Being able to access a BBS was also a good reason.
They were still very expensive less than ten years ago. I replaced an
old 286 system at home by a 486/33 in 1992. This came with 16MB RAM,
5-1/4 & 3-1/2 floppy drives, an unprecedentedly huge 540MB HDD, 250kB/s
QIC streamer tape, 17" monitor and full-tower case. The price was
GBP4,200 + VAT.
--
Brian {Hamilton Kelly} b...@dsl.co.uk
"We have gone from a world of concentrated knowledge and wisdom to one of
distributed ignorance. And we know and understand less while being incr-
easingly capable." Prof. Peter Cochrane, BT Labs
John Francis
Brian {Hamilton Kelly} <b...@dsl.co.uk> wrote:
>
>They were still very expensive less than ten years ago. I replaced an
>old 286 system at home by a 486/33 in 1992. This came with 16MB RAM,
>5-1/4 & 3-1/2 floppy drives, an unprecedentedly huge 540MB HDD, 250kB/s
>QIC streamer tape, 17" monitor and full-tower case. The price was
>GBP4,200 + VAT.
You pay for cutting-edge technology. About that time my full tower
Gateway 486-DX2/66, 8MB, dual floppies, QIC tape, twin 340MB drives,
and a mere 15" monitor (albeit with a VESA Mach-32 video card) was
$4,500. What with the LaserJet 4M, Scanjet IIc, and DeskJet DJ500
I spent over $8,500 on that system (including thinwire ethernet cards
to network it to the 386/25 my wife had decided was hers, not mine!).
That's about the only time I've been near the sharp end of the
technology curve - there weren't too many people running home
ethernet networks in the early 90s. Of course I'm *still* using
thin wire ethernet in our home office today, although I'll have
to upgrade to twisted pair when I install a DSL routing hub.
Charles Richmond
>
> In article <3A7EFB...@SunSouthWest.com> bi...@SunSouthWest.com "bill" writes:
>
> > I suspect many people moved to the IBM and/or compatibles because of the
> > Laser
> > Printer. You could now produce type-set quality output, soon followed by
> > bit
> > mapped graphics, that previously had been virtually impossible for under
> > maybe
> > $100,000 worth of equipment.
>
> I cannot perceive laser printer availability as being a force for the
> growth of PCs. I bought my first laser printer, a DEC LN03, in around
> 1984. This was not a personal purchase, but for work, where it was
> attached to a VAX-11/730, replacing the Versatec electrostatic printer/
> plotter that had hithertoo been our HQ output device.
>
for the Macintosh. And Desktop Publishing was fueled by the availability of
the Laser Printer.
--
+-------------------------------------------------------------+
| Charles and Francis Richmond <rich...@plano.net> |
+-------------------------------------------------------------+
Steve O'Hara-Smith
> I cannot perceive laser printer availability as being a force for the
> growth of PCs. I bought my first laser printer, a DEC LN03, in around
> 1984. This was not a personal purchase, but for work, where it was
> attached to a VAX-11/730, replacing the Versatec electrostatic printer/
> plotter that had hithertoo been our HQ output device.
You had a VAX available :)
The early growth in PCs was not home use but small office use where
previously there had been *no* computer. The laser printer certainly put
PCs in places where daisy wheels were too noisy (and slow) and dot matrix
too tacky.
Brian {Hamilton Kelly}
st...@ams-gw.sohara.org "Steve O'Hara-Smith" writes:
> Hmm, solder burns, how much wire-wrap radiates and my favorite the
> insertion force required for an original ceramic pack 68000 into a new
> socket :)
>
> PS: I never did find the extraction force - thankfully.
DIL packages? That's not "building a computer". Now when I were a
lad[1], we wired up these 10,000 Mullard OC200 transistors into a rake of
half-adders, shift registers, etc.
[1] Truly; circa 1961 at Warwick School
Steve O'Hara-Smith
> [1] Truly; circa 1961 at Warwick School
Warwick - weren't you the ones with the radio telescope back in the 60's ?
Or am I thinking of somewhere else.
PS: Now just how far off thread can we drag this :)
Andreas Långström
>
>PS: Now just how far off thread can we drag this :)
This is a.f.c, home of topic drift. There is _no_ limit of how far
we can go. :-)
/Andreas
Brian {Hamilton Kelly}
st...@ams-gw.sohara.org "Steve O'Hara-Smith" writes:
> Brian {Hamilton Kelly} <b...@dsl.co.uk> wrote:
>
> > [1] Truly; circa 1961 at Warwick School
>
> Warwick - weren't you the ones with the radio telescope back in the
> 60's ?
Not us.
> Or am I thinking of somewhere else.
IIRC, that was Kettering Grammar School. I think they picked up Sputnik
before Jodrell Bank, even.
> PS: Now just how far off thread can we drag this :)
This /is/ afc, after all :-)
Steve O'Hara-Smith
>> Warwick - weren't you the ones with the radio telescope back in the
>> 60's ?
> Not us.
Pity, I could have dragged us all the way to the veracity of the moon
landings thread :)
> IIRC, that was Kettering Grammar School. I think they picked up Sputnik
> before Jodrell Bank, even.
Both of those sound likely. I remember wishing I'd been there at the time
(well Apollo not Sputnik - I'm just a youngster).
> This /is/ afc, after all :-)
So we have to keep the tradition up :)
Jack Crenshaw
tag on my replies. Hope that's OK with everyone.
"Douglas Beattie Jr." wrote:
> Bill <wtsh...@mb.sympatico.ca> wrote
> > It was time. Consider CP/M and serial ports, for instance. No two CP/M machines
> > could agree on how to address a serial port, let alone what serial chip to use -
> > in my CP/M days I ran across just about every single serial chip that was made
> > at the time ( well, barring an AY 3 1013 or something *that* peculiar) as the serial
> > I/O. Some were port-mapped. On the Osborne, not only was the chip a Motorola
> > 6850, but it was memory -mapped - and in bank-switched memory. Meanwhile
> > the port I/O of the Osborne was crippled and completely useless ( except for bank
> > switching). Some chips had baud rate generators, some did not. A very
> > few CP/M machines properly supported interrupts for the serial port - now, with
> > a 2.5 MHZ processor, you really needed interrupts if you wanted to go much
> > faster than 2400 baud.
Then, on the other hand, I know of many CP/M computers that had four or more serial
ports. Many were, in fact, being used by high schools and colleges as experimental
time-share systems. With the PC, even in the year 2001, you get as many serial ports as
they want, as long as you want two, with one often dedicated to a mouse.
> > Consider disk formats. Again, no two CP/M machines had the same disk format.
> > Single density, double density ( and on one notably perverse machine in my
> > collection, the system tracks are single-density and the rest is double density).
> > If you had an Ibex model 7150 you couldn't swap disks with your buddy who had an
> > Ibex model 7102 - one was 80-track, the other 40 track.
Yeah, let's _DO_ consider disk formats. It's true that at one time, between 8" and 5.25"
floppies, hard-sectored vs. soft-sectored disks, and the transitions between DS and SS,
and DD and SD formats, there were many different formats to consider. However, by the
time the PC came along, CP/M machines had already pretty much standardized on DSDD 5.25",
soft-sectored disks. By the time MSDOS was dominant, the Kaypro format had become a de
facto standard. These disks held 400K of software, _MORE_ than the 360.
Now, consider the PC formats. We had 360k, 720k, and 1.2M 5.25" disks, and 1.44M and
2.88M (remember them?) 3.5" disks. At one time we had the unique situation where one
could write data on one floppy (the HD floppy) that couldn't be read by another computer.
Some improvement.
Or perhaps you'd like to talk about MDA, CGA, EGA, VGA, and SVGA monitor conventions. Or
PC vs. AT keyboards. Or PC vs PS2 keyboard plug conventions. Or DB-25 vs. DB-9 serial
connectors. How many of us have bought how many null modems, male-female adapters, or
keyboard adapters for our "compatible" PC's.
Do you realize how much of the BIOS of modern PC's is taken up with testing the hardware
to find out what it is, and how it's connected? It's a jungle out there. This doesn't
even count the Plug & Play devices that don't play, or the device drivers that don't
drive.
> > Consider printer cables. The Osborne had the scummiest printer cable interface
> > imaginable - a card-edge connector! Again, no two CP/M machines used the
> > same printer cable. Although I really liked some of the Kaypros that used the same
> > 36-pin "Centronics" style receptacle as most parallel printers used - this made
> > the cable double-ended and perfectly symmetrical.
I think that's a red herring. The Centronics printer standard has been a standard for
years. I moved my Epson printer from a TRS-80 to an S-100 system to a Kaypro simply by
plugging it in.
Just because Adam Osborne chose to be non-standard is no reason to blame CP/M.
> > Consider keyboards - if the keyboard in my Osborne siezes up, I'll never be
> > able to replace it. Custom, proprietary, and unique to that machine - and if I
> > don't like the layout, I'm out of luck.
>
> So you mean it's still working? I guess they don't build keyboards like they used
> to. Today I sometimes wonder if a keyboard isn't designed to lock up at least twice
> within the first 8 months I've had it (so my company will be forced to buy another
> one, from the same vendor).
ROFL!! I went through three of Microsoft's notorious "ergonomic" keyboards before I gave
up
and accepted the fact that they simply weren't intended to be actually used for anything.
Jack
Jack Crenshaw
> But if a thousand-megahertz 32 bits machine with 128Megs of RAM and a
> googolbyte of hard disks costs the exact *same* as the CP/M machine,
> what's the point of restricting yourself to the less capable platform? Any
> application you could run on a Z80 machine you can run on a Pentium and not
> only will it be faster and cheaper, it will come in a blister pack at your local
> software store ( not mail-order in 8 to 12 weeks).
I'm not convinced that the PC/Pentium 3/googolbyte system _IS_ faster. If my new (but
already obsolete) 600MHz Dell is so bloomin' fast, maybe someone can explain to me
why it takes 70 seconds to boot. If my old Kaypro had done that (or, for that matter,
my old TRS-80), I'd have thrown it out the Windows.
Maybe someone can explain to me why the first time I use a one-key Word macro
(which does something really complicated like change the font of a word), it takes seven
seconds. Don't bother, I know why: Word is loading Visual Basic.
Maybe someone can tell me why some applications (Matlab is one) take so long to load
that I think the machine has hung up.
For that matter, maybe someone can tell my why my machine hangs up!
Not too long ago, I left my home PC on (running through its UPS), with nothing running but
the
screen saver. Next morning, there was a dialogue box that said, "This program has executed
an illegal instruction, and has been terminated." Two days ago, I pressed "Shut down" on my
office computer, and went home. Next morning, I came to work to find the message,
"Operation failed."
Would you like to ask me how many times, in my five years with CP/M, the OS crashed?
(zero). Would you like to ask me how many times Wordstar crashed (zero)? Would you like to
ask me
how many times I needed to run Scandisk (zero)?
> I am aware, though I'm not expressing it well, that there is a distinction between
> the operating system and the hardware platforms it comes on. (MS DOS has done
> surprisingly poorly on the Macintosh, for example!). The point I was trying to
> make is that the reason CP/M vanished circa 1981 or so is that no matter
> what you thought about MS DOS, the hardware it ran on was so much less
> hassle to get working in usable fashion that there was no incentive to
> stick with the 8080 world any more.
I think it was more a case of lemmingism. The earliest, 4.77MHz PC's had nowhere near the
performance of a 4Mhz Kaypro. It took another five or six years, when AT clones became
commonly available, before they had clock speeds to outperform a Kaypro. And even then,
the software was utterly _LAME_ (anyone remember MultiMate?). I had a 10MHz AT clone
with 2Mb of RAM. I could use all of that RAM I wanted to, as long as it didn't exceed
640Mb. The rest just went to waste.
> Name one CP/M "innovation" that still persists today.
Batch files.
Floppy disks.
Hard disks.
Interrupt-driven I/O.
Stretchable disk directories.
Turbo Pascal (still in Delphi).
Of course, if you want to include CP/M replacements, we'd have to consider ZCPR3, in which
case we're still waiting for the PC to catch up.
> Windows has "stifled" creativity? You mean I no longer have to guess which way
> the author of whatever program I'm using has decided to implement the "print"
> command? I don't think *that* kind of creativity is missed at all.
What? You think all Windows programs use the same standards? Try typing any key,
F1 through F12, in different apps and see what happens.
Jack
G Swaine
> @mb.sympatico.ca wrote:
>
> > But if a thousand-megahertz 32 bits machine with 128Megs of RAM and
a
> > googolbyte of hard disks costs the exact *same* as the CP/M machine,
> > what's the point of restricting yourself to the less capable
platform? Any
> > application you could run on a Z80 machine you can run on a Pentium
and not
> > only will it be faster and cheaper, it will come in a blister pack
at your local
> > software store ( not mail-order in 8 to 12 weeks).
>
> I'm not convinced that the PC/Pentium 3/googolbyte system _IS_ faster.
If my new (but
> already obsolete) 600MHz Dell is so bloomin' fast, maybe someone can
explain to me
> why it takes 70 seconds to boot. If my old Kaypro had done that (or,
for that matter,
> my old TRS-80), I'd have thrown it out the Windows.
I can recall someone mentioning to me an old law of computing (I don't
recall if it had a name, or what that name may have been)
It goes something like this: "For every speed/capacity increase of
ratio X, there will be X-Squared asked of it"
So, for a computer 1.5 times the speed of its predecessor, there will be
2.25 times the workload.
Unfortunately, I've found that this ratio doesn't seem to work in
reverse (a computer half the speed will have one-quarter the workload,
and a lot of free processor cycles to do other useful stuff)
--
Quote for the month:
Wally Batty: "She's sending me off to an auction"
Norman Clegg: "Why doesn't she just divorce you like ordinary folk?"
("Last Of The Summer Wine" [circa 1982])
--
g.sw...@the.gluepot.com is a bogus address. I can be contacted via an
intermediary : gem at gem win co nz. I would like to apologise to the
genuine respondents that this may inconvenience.
Harold F. Bower
>
> @mb.sympatico.ca wrote:
>
<snip Jack's experience with MS which I have also experienced;)>
> Would you like to ask me how many times, in my five years with CP/M, the OS crashed?
> (zero). Would you like to ask me how many times Wordstar crashed (zero)? Would you like to
> ask me
> how many times I needed to run Scandisk (zero)?
Ditto, but did run the 'FINDBAD' of my own accord with known bad media.
<snip>
> I think it was more a case of lemmingism. The earliest, 4.77MHz PC's had nowhere near the
> performance of a 4Mhz Kaypro. It took another five or six years, when AT clones became
> commonly available, before they had clock speeds to outperform a Kaypro. And even then,
> the software was utterly _LAME_ (anyone remember MultiMate?). I had a 10MHz AT clone
> with 2Mb of RAM. I could use all of that RAM I wanted to, as long as it didn't exceed
> 640Mb. The rest just went to waste.
There was an even more insidious effect. Because of the poor
performance of the early PCs, it seems that people (anyone writing S/W)
were encouraged to bypass the architectural structure of the system and
directly manipulate hardware from the application. Later when
portability problems became very acute, they still persisted in
bypassing the Operating System core and going directly to the 'BIOS'
without necessarily updating OS structures when done.
>
> > Name one CP/M "innovation" that still persists today.
>
> Batch files.
> Floppy disks.
> Hard disks.
> Interrupt-driven I/O.
> Stretchable disk directories.
> Turbo Pascal (still in Delphi).
>
> Of course, if you want to include CP/M replacements, we'd have to consider ZCPR3, in which
> case we're still waiting for the PC to catch up.
>
Totally agree with many elements adapted from Unix:
multiple commands on a single line
Termcaps for standardizing and abstracting text-based graphics
Tree-structured Help files (that HELP)
Ability to emulate different directory abstractions (including trees)
Furthermore, A CP/M program called BackGrounder(tm) allowed
user-controlled task-swapping with virtual computers in 1986, oddly
enough patching the OS kernel with a mechanism seemingly identical to
the 'inuse' flag MSDOS implemented to allow 'multitasking' in Windows
3.1.
File Date and Time Stamping existed in the early 1980's generically in
CP/M Plus (3) with two times per file, and DateStamper(tm) with three
times (Create, Access and Modify) per file.
Hal
mb.sympatico.ca
>Since I missed Bill's original post re CP/M vs. PC computers, I'm using this message to
>tag on my replies. Hope that's OK with everyone.
>
>Then, on the other hand, I know of many CP/M computers that had four or more serial
>ports. Many were, in fact, being used by high schools and colleges as experimental
>time-share systems. With the PC, even in the year 2001, you get as many serial ports as
>they want, as long as you want two, with one often dedicated to a mouse.
>
Oh, true, there was at least 1 CP/M computer built with any conceivable combination
of options. S100 machiens were particularly good for this, if you needed a
particular function you wire-wrapped a board for it. But the volumes of machines made
were tiny, and you needed a disproportionate amount of expertise to get the machine
going.
In 2001, you put your mouse on a USB port or at least a PS/2 mouse port.
>> > Consider disk formats. Again, no two CP/M machines had the same disk format.
>Yeah, let's _DO_ consider disk formats. It's true that at one time, between 8" and 5.25"
>floppies, hard-sectored vs. soft-sectored disks, and the transitions between DS and SS,
>and DD and SD formats, there were many different formats to consider. However, by the
>time the PC came along, CP/M machines had already pretty much standardized on DSDD 5.25",
>soft-sectored disks. By the time MSDOS was dominant, the Kaypro format had become a de
>facto standard. These disks held 400K of software, _MORE_ than the 360.
I had at one time 3 Kaypros, an Osborne, an Ibex and a Megatel Quark ( now
*that* was a hot machine- 8 MHZ Z80 on the tightest-packed through-hole
board I've ever seen before or since). Without taking extra steps that I'd
hate to explain to a secretary over the phone, I would be *unable* to exchange
disk data between them. I was a hobbyist/hacker at the time, I didn't mind
running Uniform or tweaking the DPB to get them to interoperate on some
level. The advent of PCs made this go away...true, if you still had 5 1/4
360 K drives, you were a little left out till you picked up a 1.44 M drive, but at
least you didn't need to do anything special to recognize the new format once
the drive was installed.
>
>Now, consider the PC formats. We had 360k, 720k, and 1.2M 5.25" disks, and 1.44M and
>2.88M (remember them?) 3.5" disks. At one time we had the unique situation where one
>could write data on one floppy (the HD floppy) that couldn't be read by another computer.
>Some improvement.
I think you've listed *all* the MS DOS formats and most of these related to different
sizes and types of media. In the CP/M world the *exact same* 5 1/4 inch floppy
could be formatted in ( recalling Uniform's setup screens....5 or 6 pages of 20
choices each, perhaps?) at least 100 different mutually incompatible ways.
The advantage of the 1.2Meg/360K incompatibility is that it was a *standard*
bug that *everyone* knew about and could work around in the *same* way.
>
>Or perhaps you'd like to talk about MDA, CGA, EGA, VGA, and SVGA monitor conventions. Or
>PC vs. AT keyboards. Or PC vs PS2 keyboard plug conventions. Or DB-25 vs. DB-9 serial
>connectors. How many of us have bought how many null modems, male-female adapters, or
>keyboard adapters for our "compatible" PC's.
But you can buy the adapters shrink-wrapped for a song at your local Canadian Tire
( or your local equivalent). Whereas in the glory days of CP/M you didn't expect
to BUY a $10 cable or adapter, you expected to solder it up yourself or pay $50
for a cable. The IBM Clone world wasn't perfect but it was much easier to manage
from the hardware point of view than the CP/M world.
>
>Do you realize how much of the BIOS of modern PC's is taken up with testing the hardware
>to find out what it is, and how it's connected? It's a jungle out there. This doesn't
>even count the Plug & Play devices that don't play, or the device drivers that don't
>drive.
But we don't *care* about things the machine handles itself transparently and
reliably!
We were talking about the '80s here - Plug and Play in the MS DOS world wasn't
really prevalent then ( well, Microchannel machines, perhaps). Plug and Play
is definitely the right idea compared with the way it was done with the technology
available in the CP/M era.
<printer cables>
Steve O'Hara-Smith
> It is amazing that a single platform could spawn that many formats.
All from one manufacturer too. At least there was diversity of
manufacturer as an excuse for the CP/M formats, which did settle
down to 400k and 800k about the time MS-DOS hit 360k - hmm!
Ah well CD-ROMs are all the same and everyone knows you need a few
hundred megs to do anything useful :)
Tom Ruben
<jcr...@earthlink.net> writes
>I'm not convinced that the PC/Pentium 3/googolbyte system _IS_ faster.
For those who may not have come across the term before, perhaps it
should be explained that googol is 10 to the power 100. And googolplex
is 10 to the power googol.
--
Tom
Erno Palonheimo
> There are more formats. The original IBM PC used single sided drives,
> only later it was delivered with double sided drives.
> For a vanilla IBM-PC (or clone) the following formats for 5 1/4"
> diskettes have been in use:
>
> MS-DOS 1.00 160K (1 side, 40 tracks, 8 sectors)
> MS-DOS 1.10 320K (2 sides, 40 tracks, 9 sectors)
> MS-DOS 2.00 180K (1 side, 40 tracks, 8 sectors)
> MS-DOS 2.00 260K (2 sides, 40 tracks, 9 sectors)
^^^
Shouldn't this be 360K?
> MS-DOS 3.00 720K (2 sides, 80 tracks, 9 sectors)
> MS-DOS 3.00 1200K (2 sides, 80 tracks, 15 sectors)
--
Erno Palonheimo ; e...@iki.fi ; http://iki.fi/esp/ ; +358505604765
Bill Marcum
G Swaine wrote in message <96lu46$h1l$1...@news.comnet.co.nz>...
>Unfortunately, I've found that this ratio doesn't seem to work in
>reverse (a computer half the speed will have one-quarter the workload,
>and a lot of free processor cycles to do other useful stuff)
>
they'd have a lot of free cycles if they were plugged in.
Arargh!
wrote:
<snip>
>
> MS-DOS 1.00 160K (1 side, 40 tracks, 8 sectors)
> MS-DOS 1.10 320K (2 sides, 40 tracks, 9 sectors)
> MS-DOS 2.00 180K (1 side, 40 tracks, 8 sectors)
> MS-DOS 2.00 260K (2 sides, 40 tracks, 9 sectors)
> MS-DOS 3.00 720K (2 sides, 80 tracks, 9 sectors)
> MS-DOS 3.00 1200K (2 sides, 80 tracks, 15 sectors)
>
>It is amazing that a single platform could spawn that many formats.
Don't forget that goofy 1.7 something meg format for software
distribution :-)
--
Arargh (at enteract dot com) http://www.arargh.com
Jay R. Ashworth
Jack Crenshaw <jcr...@earthlink.net> wrote:
> For that matter, maybe someone can tell my why my machine hangs up!
"Every other computer company has spent the last 20 years working on
improving fault tolerance in their computers. Microsoft, OTOH, has
spent the time working on improving fault tolerance in their
customers."
(Adapted from someone else's epigram.)
Charles Richmond
in, they may *not* be getting any electric power!!!
Ian Stirling
For machines much beyond 3-4 years old, it's cheaper to buy new computers
and use their compute power, than to pay for the electricity to run the
old ones.
--
http://inquisitor.i.am/ | mailto:inqui...@i.am | Ian Stirling.
---------------------------+-------------------------+--------------------------
He who lives in a glass house should not invite he who is without sin.
Robert Hansen
have to put up with these real " Modern, Fast" PC's.
Bob
Harold F. Bower
>
> Bill Marcum <bma...@iglou.com> wrote:
>
> >G Swaine wrote in message <96lu46$h1l$1...@news.comnet.co.nz>...
> >>Unfortunately, I've found that this ratio doesn't seem to work in
> >>reverse (a computer half the speed will have one-quarter the workload,
> >>and a lot of free processor cycles to do other useful stuff)
> >>
> >Look at all the "obsolete" computers that aren't being used at all...
> >they'd have a lot of free cycles if they were plugged in.
>
> For machines much beyond 3-4 years old, it's cheaper to buy new computers
> and use their compute power, than to pay for the electricity to run the
> old ones.
Depends...I run a couple of YASBECs (1991/2) and P112s (1995) without
paying anything for electricity, so the only way a new one would run
cheaper is if they GENERATED it! A couple of photovoltaic arrays and
NiCd wet cell banks from an old Auxiliary Power Unit keep them humming
right along.....
Hal
Tom Ruben
<h...@invalid.invalid> writes
>In article <v4yj6.319$vh7....@news1.mts.net>,
> <wtsh...@mb.sympatico.ca> wrote:
>
>> >Now, consider the PC formats. We had 360k, 720k, and 1.2M 5.25" disks,
>> and 1.44M and
>> >2.88M (remember them?) 3.5" disks.
>
>> to different
>> sizes and types of media.
>
>only later it was delivered with double sided drives.
>For a vanilla IBM-PC (or clone) the following formats for 5 1/4"
>diskettes have been in use:
>
> MS-DOS 1.10 320K (2 sides, 40 tracks, 9 sectors)
> MS-DOS 2.00 180K (1 side, 40 tracks, 8 sectors)
> MS-DOS 2.00 260K (2 sides, 40 tracks, 9 sectors)
> MS-DOS 3.00 720K (2 sides, 80 tracks, 9 sectors)
> MS-DOS 3.00 1200K (2 sides, 80 tracks, 15 sectors)
>
>It is amazing that a single platform could spawn that many formats.
>
--
Tom
Eric Fischer
> The original IBM PC did not have a disk drive, it had a tape interface.
The disk drive was initially optional, but it was available from the
start, and no sane person would have bought an IBM PC without one.
eric