IBM MIcrochannel??
Andrew Gore
Long before I bought my first "IBM-compatible", as they were
known back then, I recall IBM began boasting that their new line of
business PC's had "microchannel" architecture. Thiis was, maybe 1988?
IBM was beginning to get competition from 'clones' like Compaq (and
Osbourne?), and needed to have some 'edge' over the clones. I recall
ata press conference announcing the new line, a reporter naturally
asked the marketing guy just what 'microchannel' meant. He hemmed and
hawed, and clearly couldn't come up with a convincing answer. The
general feeling was, that it was basicaly a marketing device, a
selling point, that really didn't have any real-world advantages.
Remember when one brand of gasoline came with the additive "F-310"
(F three ten)? Same kind of thing. Can anybody confirm this?
Steve O'Hara-Smith
Andrew Gore <dic...@earthlink.net> wrote:
AG> Long before I bought my first "IBM-compatible", as they were
AG> known back then, I recall IBM began boasting that their new line of
AG> business PC's had "microchannel" architecture. Thiis was, maybe 1988?
AG> IBM was beginning to get competition from 'clones' like Compaq (and
AG> Osbourne?), and needed to have some 'edge' over the clones. I recall
I recall it being rather more aggressive than that - IBM had some
trouble with some major manufacturers, but not much. They all had a
nightmare on their hands by the name of Taiwan, from where PC clones
poured out at really low prices. At the time it was noticed that while
the market was awash with PC clones there were no Apple clones, the catch
was to clone an Apple you had to license some bits. Microchannel was IBMs
attempt to kill off the clones, if you wanted to biuld an MCA product you
had to license MCA from IBM. It worked (sort of) almost no maker of cheap
clones licensed MCA, unfortunately for IBM almost nobody bought it either.
IBM was left with the least compatible PC clones on the market and a face
full of egg. Meanwhile the rest of the world went VLB (a hack) and then PCI.
MCA was technically much better than ISA, it may even have prevented
PCI if it had been licensed differently.
--
C:>WIN | Directable Mirrors
The computer obeys and wins. |A Better Way To Focus The Sun
You lose and Bill collects. | licenses available - see:
| http://www.sohara.org/
CBFalconer
>
... snip ...
>
> MCA was technically much better than ISA, it may even have prevented
> PCI if it had been licensed differently.
The thing that killed MCA was that, to get a license, the clone
makers had to license their previous and future usage of the ISA
bus. Not being overly inclined to hand money over to IBM, most
clone manufacturers declined that munificent offer. I believe
licensed MCA usage also involved buying special interface chips
from IBM.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@worldnet.att.net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net> USE worldnet address!
Tim Shoppa
> MCA was technically much better than ISA, it may even have prevented
> PCI if it had been licensed differently.
I have the gut feeling that all pre-PCI bus improvements failed
because the buyer of your typical consumer PC didn't see any need
for a better bus until the mid-90's.
EISA came along a few years after MCA, when there certainly was more
of a need for a better bus, yet it was a flop as far as your typical
consumer-grade PC was concerned. It was backwards-
compatible with the ISA bus, yet it was a flop as far as your
typical consumer-grade PC was concerned. It was a relatively
open standard (not proprietary like MCA), yet it was a flop as far as
your typical consumer-grade PC was concerned.
It's worth pointing out that MCA lived into the mid-90's on
several IBM machines, as did EISA on Compaq and other clone
maker's high-end machines, so they weren't complete failures:
they both had markets where they were relevant and (presumably) useful.
The real thing that will stop EISA living into the future in
hand-me-down machines is that you need that EISA Configuration
Utility floppy that came with your motherboard to add or remove
a card. And that floppy rarely gets handed down with the machine...
Tim.
Jay Maynard
wrote:
>It's worth pointing out that MCA lived into the mid-90's on
>several IBM machines, as did EISA on Compaq and other clone
>maker's high-end machines, so they weren't complete failures:
>they both had markets where they were relevant and (presumably) useful.
I have a couple of IBM MCA RS/6000s. They make great platforms for the P/390
personal mainframe system, much more so than the PCI versions that followed;
you can stuff an MCA P/390 board in any MCA box, and it'll work, while the
PCI version is much, much pickier.
Not all EISA boxen were PC-class. My SGI Indigo 2 has an EISA 10/100
Ethernet card. No, it neither has nor needs a configuration floppy, as the
configuration is done in the driver.
Rupert Pigott
news:3CCE69E4...@yahoo.com...
> Steve O'Hara-Smith wrote:
> >
> ... snip ...
> >
> > MCA was technically much better than ISA, it may even have prevented
> > PCI if it had been licensed differently.
>
> The thing that killed MCA was that, to get a license, the clone
> makers had to license their previous and future usage of the ISA
> bus. Not being overly inclined to hand money over to IBM, most
> clone manufacturers declined that munificent offer. I believe
> licensed MCA usage also involved buying special interface chips
> from IBM.
LOL, this was brought up before in this group - not long ago. I
Googled for info. As it turns out a lot of those "MCA Licensing Was
Evil" stories were exactly that, stories.
Apparently the license was a one off payment, and a modest one at
that (few tens of thousands). To be honest I think the real reason
why MCA flopped was that it required extra R&D and marketing
effort in a well established market.
Cheers,
Rupert
Rupert Pigott
news:slrnact3hc....@thebrain.conmicro.cx...[SNIP]
> Not all EISA boxen were PC-class. My SGI Indigo 2 has an EISA 10/100
> Ethernet card. No, it neither has nor needs a configuration floppy, as the
> configuration is done in the driver.
I hated those floppies. People had a habit of losing them. Stupid solution,
if
the config came with the driver it would have made sense... :(
I think that some Apollo workstations used EISA too. I think I'd have
preferred
MCA on balance all other things being equal... MCA had a ton of ground lines
on it, classic case of IBM hyper-engineering in action. :)
Cheers,
Rupert
MSCHAEF.COM
CBFalconer <cbfal...@worldnet.att.net> wrote:
>Steve O'Hara-Smith wrote:
>>
>... snip ...
>>
>> MCA was technically much better than ISA, it may even have prevented
>> PCI if it had been licensed differently.
>
>The thing that killed MCA was that, to get a license, the clone
>makers had to license their previous and future usage of the ISA
>bus. Not being overly inclined to hand money over to IBM, most
>clone manufacturers declined that munificent offer. I believe
>licensed MCA usage also involved buying special interface chips
>from IBM.
What's your take on the board space issue? I remember when MCA first came
out, people were complaining that the MCA form factor allowed half the
board space of a full ISA card.
-Mike
Peter Ibbotson
news:HVxz8.11088$Lj.8...@bin4.nnrp.aus1.giganews.com...
I think the reasons it failed were:
a) You couldn't plug existing "classic" ISA cards in
b) You had to the config file somewhere (Pain in the bum for roving
engineers)
c) You could use a simple PAL to do the logic for decodes etc. It required
significant amounts of glue logic to do a simple board (I don't remember
PGAs being available circa '86 when MCA was announced). This meant that many
engineers ignored it.
d) IIRC you needed different drivers to talk to the board from those for an
ISA system.
I don't remember board space being a significant problem but then I was
doing an ASIC design so it was basically one ASIC plus an SRAM on the board.
PCI got around this by offering backwards compatible slots on the
motherboard and by then all the glue logic fitted easily in a Xylinx type
PGA so most engineers could implement a single chip solution (I would
imagine by now there is a good drop in design for a PCI bus).
--
Work pet...@lakeview.co.uk.plugh.org | remove magic word .org to reply
Home pe...@ibbotson.co.uk.plugh.org | I own the domain but theres no MX
Anne & Lynn Wheeler
> I have a couple of IBM MCA RS/6000s. They make great platforms for the P/390
> personal mainframe system, much more so than the PCI versions that followed;
> you can stuff an MCA P/390 board in any MCA box, and it'll work, while the
> PCI version is much, much pickier.
the problem with MCA for RS/6000s wasn't the protocol itself or even
doubling the rate ... it was having a common bus hardware with the
ps/2 and suggestion that the rs/6000s should use ps/2 adapter cards
instead of designing their own. there was a joke if the rs/6000s had
to use the ps/2 scsi disk adapter card ... it would run as as ps/2.
there were benchmarks of the ps/2 16mbit MCA t/r adapter card against
the PC/RT designed 4mbit ISA t/r adapter card. While 16mbit t/r might
have had higher aggregate thruput ... the cards were designed for
office desk top thruput. Any single 16mbit MCA t/r card had lower
sustained thruput than the PC/RT designed 4mbit ISA t/r card. Not so
bad for random office desk top use ... but serious problem for
engineering client/server ... especially where the "server" needed to
sustain the aggregate of all the individual clients.
--
Anne & Lynn Wheeler | ly...@garlic.com - http://www.garlic.com/~lynn/
Steve O'Hara-Smith
Tim Shoppa <sho...@trailing-edge.com> wrote:
TS>
TS> > MCA was technically much better than ISA, it may even have prevented
TS> > PCI if it had been licensed differently.
TS>
TS> I have the gut feeling that all pre-PCI bus improvements failed
TS> because the buyer of your typical consumer PC didn't see any need
TS> for a better bus until the mid-90's.
They were right too! Especially after the video went to VLB (and
the discs too if you wanted a high performance machine). The heaviest thing
on the bus after that was probably the ethernet (quite rare too) - and even
an 8 bit ISA ethernet could saturate the link at 10 Mbps, 16 bit ISA had
plenty of capacity (but not for long).
Pete Fenelon
> They were right too! Especially after the video went to VLB (and
> the discs too if you wanted a high performance machine).
Ah, joy. The only time I've ever had a "high-end" PC (relative to
the prevailing standards of the day) was back in mid '93, when I
bought some DX2/66s for a new project. VLB graphics (Spea/V7 Mirage),
VLB cached disc controllers (Promise DC4030), single-speed Sony
SCSI CD-Roms on a Future Domain controller... the PCI-based P133s we bought
nearly 2 years later weren't all that much faster in interactive use.
All goes to show that it's not CPU power that makes a machine, it's the
stuff around it :)
pete
--
pe...@fenelon.com "Irk the purists, irk the purists, it's a right good laugh."
CBFalconer
I've been wrong before :-) (once). At any rate my opinion is
clouded by the fact that I was never interested in volume, just
the ability to create one-off special peripherals, and was using
my own quite satisfactory 10 year old buss design at the time.
Jim Stewart
Excuse me for jumping in here... Both licensing and board space were
non-trivial issues. At the time MCA came out, I had designed several
ISA cards and was tempted to do an MCA design. As mentioned earlier,
the bus required some unusual chips. I also seem to recall that you had
to get a unique ID from IBM and this costed money. The board size was
small, but even worst was the form factor. In those days, and even
today to a lesser extent, the ideal form factor is square and the
further you deviate from that, the harder the printed circuit layout
becomes. In those days, where most boards were laid out by hand with
tape on Mylar, the small squashed shape, coupled with extra bus chips,
made the MCA board outline look real unpleasant to work with.
jim
gray...@yahoo.com
> On Tue, 30 Apr 2002 04:05:25 GMT
> Andrew Gore <dic...@earthlink.net> wrote:
>
> AG> Long before I bought my first "IBM-compatible", as they were
> AG> known back then, I recall IBM began boasting that their new line of
> AG> business PC's had "microchannel" architecture. Thiis was, maybe 1988?
> AG> IBM was beginning to get competition from 'clones' like Compaq (and
> AG> Osbourne?), and needed to have some 'edge' over the clones. I recall
>
> I recall it being rather more aggressive than that - IBM had some
> trouble with some major manufacturers, but not much. They all had a
> nightmare on their hands by the name of Taiwan, from where PC clones
> poured out at really low prices. At the time it was noticed that while
> the market was awash with PC clones there were no Apple clones, the catch
> was to clone an Apple you had to license some bits. Microchannel was IBMs
> attempt to kill off the clones, if you wanted to biuld an MCA product you
> had to license MCA from IBM. It worked (sort of) almost no maker of cheap
> clones licensed MCA, unfortunately for IBM almost nobody bought it either.
> IBM was left with the least compatible PC clones on the market and a face
> full of egg. Meanwhile the rest of the world went VLB (a hack) and then PCI.
>
> MCA was technically much better than ISA, it may even have prevented
> PCI if it had been licensed differently.
>
I think Alan Sugar (Amstrad) downloaded some money to IBM , hoping to
grab a share of the MCA generation . Rumor is that Alan is not the
sort to make a fool of . Must have been fun when he realized he had
taken the wrong branch of the road .
Tim Shoppa
> Not all EISA boxen were PC-class. My SGI Indigo 2 has an EISA 10/100
> Ethernet card. No, it neither has nor needs a configuration floppy, as the
> configuration is done in the driver.
I used some DEC Alpha 2100's with EISA busses for a couple of years in
the early 90's, when they were really hot sh*t. But they had the
dreaded EISA Configuration Utility which you had to boot from floppy,
after going into the Alpha's firmware and telling it you wanted to
switch to *WINDOWS MODE*, where the floppy drive became A: and your
system disk became C:. Ughhh!
I think there was an emulator that actually ran 80x86 code underneath
the whole ugly mess, so that you could use PC-clone configuration
utility floppies (I think that I had to do this once to set up the
EISA RAID controller or something. I've supressed most of that
memory...)
Tim.
Rupert Pigott
news:5427AC62E41033FC.1640524C...@lp.airnews.net...
[SNIP]
> becomes. In those days, where most boards were laid out by hand with
> tape on Mylar, the small squashed shape, coupled with extra bus chips,
> made the MCA board outline look real unpleasant to work with.
You're kidding ! Mind you I'm guessing you're talking about a particular
market segment. AFAIK most boards I saw being designed were done
using a CAD package - or even a paint package (eeek). :P
I have to agree that MCA (or PCI for that matter) does not suit
people hacking up a one off or ultra-low volume board on the cheap,
and that ISA is a better fit... However when it comes down to people
designing graphics cards, SCSI controllers, high speed network
adaptors or server motherboards MCA makes a lot more sense.
MCA certainly *did* introduce an overhead of it's own due to
the physical packaging, but there were good reasons for that...
ie: Engineering the damn thing properly. I bet MCA did better in
terms of cards & motherboards sold than NuBus. :)
Cheers,
Rupert
CBFalconer
> > becomes. In those days, where most boards were laid out by hand with
> > tape on Mylar, the small squashed shape, coupled with extra bus chips,
> > made the MCA board outline look real unpleasant to work with.
>
> You're kidding ! Mind you I'm guessing you're talking about a particular
> market segment. AFAIK most boards I saw being designed were done
> using a CAD package - or even a paint package (eeek). :P
Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
knives, on light boxes equiped with a precision 0.1 x 0.1 inch
transparent grid. Two sided (i.e. complex) boards were done with
multiple layouts, carefully registered, or the high tech way was
with multiple colored tapes, allowing separation at photoreduction
time. Routing was an art. Reliable plated through holes
connected the two sides. Cheap production meant single sided
boards, with a minimum of jumpers.
After we got those carefully produced photo-reductions, someone
had to produce the silk-screens to actually fabricate the boards.
Rupert Pigott
[SNIP]
> Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
> knives, on light boxes equiped with a precision 0.1 x 0.1 inch
Hehe, I'm very well aware of how boards were designed in the
olden days. I would just be very surprised if people were still doing
that en masse when MCA was introduced... That was the 90s,
right ? :)
Cheers,
Rupert
jrla...@shell.golden.net
Rupert Pigott <dark.try-eati...@btinternet.com> wrote:
>
>Apparently the license was a one off payment, and a modest one at
>that (few tens of thousands). To be honest I think the real reason
>why MCA flopped was that it required extra R&D and marketing
>effort in a well established market.
I'll go along with this. At the time we were putting an AMD 29K RISC chip
on a card to plug into an MCA machine. It was a tight fit. Physically the
chip was almost as wide as the board was. Electrically it also took most
of the 5V power that each slot was alotted.
--
john R. Latala
jrla...@golden.net
Peter Ibbotson
news:aangqt$pl5$1...@knossos.btinternet.com...
Nope certainly by '87 PS/2 Model 80 was on the market, although I would
agree that most folks would have been using CAD for their layout.
MSCHAEF.COM
CBFalconer <cbfal...@worldnet.att.net> wrote:
>Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
>knives, on light boxes equiped with a precision 0.1 x 0.1 inch
>transparent grid. Two sided (i.e. complex) boards were done with
>multiple layouts, carefully registered, or the high tech way was
>with multiple colored tapes, allowing separation at photoreduction
>time.
Is this also how multi-layered boards were done? Or were multi-layered
boards even around then?
-Mike
MSCHAEF.COM
Peter Ibbotson <spa...@ibbotson.co.uk> wrote:
>
>Nope certainly by '87 PS/2 Model 80 was on the market, although I would
>agree that most folks would have been using CAD for their layout.
Does anybody here remember the issue with 32-bit slots on the Model 80? I
seem to remember that they did not implement one of the more important
features for bandwidth. Maybe matched memory cycles?
-Mike
Jim Stewart
>
> "Rupert Pigott" <dark.try-eati...@btinternet.com> wrote in message
> news:aangqt$pl5$1...@knossos.btinternet.com...
> > CBFalconer <cbfal...@yahoo.com> wrote in message
> > news:3CCF33F4...@yahoo.com...
> > [SNIP]
> > > Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
> > > knives, on light boxes equiped with a precision 0.1 x 0.1 inch
> > [SNIP]
> >
> > Hehe, I'm very well aware of how boards were designed in the
> > olden days. I would just be very surprised if people were still doing
> > that en masse when MCA was introduced... That was the 90s,
> > right ? :)
> >
>
> Nope certainly by '87 PS/2 Model 80 was on the market, although I would
> agree that most folks would have been using CAD for their layout.
The transistion to CAD was long and painful and depended upon how much
money a company wanted to spend up front. By that time, I'd had a
couple of boards spun by outside consultants using minicomputer based
systems. I also still had an extensive collection of tapeup tools and
supplies.
The transition for me was Tango CAD software that would run on an AT
clone. My wife got pretty good at working with it and shortly
thereafter I hauled all my tapeup stuff to Mike Quinn's (an East Bay
surplus dealer).
Douglas H. Quebbeman
>
> It's worth pointing out that MCA lived into the mid-90's on
> several IBM machines, as did EISA on Compaq and other clone
> maker's high-end machines, so they weren't complete failures:
> they both had markets where they were relevant and (presumably) useful.
>
> The real thing that will stop EISA living into the future in
> hand-me-down machines is that you need that EISA Configuration
> Utility floppy that came with your motherboard to add or remove
> a card. And that floppy rarely gets handed down with the machine...
Dell servers, if they include any non-PCI slots, include one
or more EISA slots. Fortunately, they also write a small hidden
partition to the boot drive that contains (among others) the
EISA config utility.
Of course, many people upgrade these servers without recreating
that tiny hidden partition, so your comment w/r/t hand-me-downs
still stands.
-dq
Bob Willard
Multi-layered boards were used by 1970, probably before. I recall
seeing one that was >12 layers in 1971, and I doubt if that was a
record.
--
Cheers, Bob
CBFalconer
> "MSCHAEF.COM" wrote:
> > CBFalconer <cbfal...@worldnet.att.net> wrote:
> >
> > >Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
> > >knives, on light boxes equiped with a precision 0.1 x 0.1 inch
> > >transparent grid. Two sided (i.e. complex) boards were done with
> > >multiple layouts, carefully registered, or the high tech way was
> > >with multiple colored tapes, allowing separation at photoreduction
> > >time.
> >
> > Is this also how multi-layered boards were done? Or were multi-
> > layered boards even around then?
>
> seeing one that was >12 layers in 1971, and I doubt if that was a
> record.
Not for the competitive commerical world. Those things went into
the space race and the MAD race. Manufacturing yields were not
impressive.
Bob Willard
>
> Bob Willard wrote:
> > "MSCHAEF.COM" wrote:
> > > CBFalconer <cbfal...@worldnet.att.net> wrote:
> > >
> > > >Oh the innocence. Artwork was laid at 2x or 4x with tape, xacto
> > > >knives, on light boxes equiped with a precision 0.1 x 0.1 inch
> > > >transparent grid. Two sided (i.e. complex) boards were done with
> > > >multiple layouts, carefully registered, or the high tech way was
> > > >with multiple colored tapes, allowing separation at photoreduction
> > > >time.
> > >
> > > Is this also how multi-layered boards were done? Or were multi-
> > > layered boards even around then?
> >
> > Multi-layered boards were used by 1970, probably before. I recall
> > seeing one that was >12 layers in 1971, and I doubt if that was a
> > record.
>
> Not for the competitive commerical world. Those things went into
> the space race and the MAD race. Manufacturing yields were not
> impressive.
Actually, the one I referred to was in a video terminal which was
intended for the commercial world. I don't recall the vendor's name;
I was working at a trade show in Germany, and it was a European outfit.
I was astounded by their ability with multi-layer cards; my (then)
employer did 2-layer cards, and sub-contracted for 4-layer cards with
only power & ground on inner layers. 12-layers or more, on fairly
large cards with standard (FR4?) material, just blew my mind.
--
Cheers, Bob
Russell P. Holsclaw
> seeing one that was >12 layers in 1971, and I doubt if that was a
> record.
All of the backplane boards used in the IBM System/360 CPU's and many of its
peripherals, based on the "SLT" technology (Solid Logic Technology), were
4-layer boards. The two outer layers carried logic signals, and the two
inner layers were for power and ground distribution. With all the logic
signals running parallel to ground/power planes, noise and crosstalk were
minimized, and the combination acted as a fixed-impedance transmission line.
This technology was announced by IBM in 1964. I don't know how long before
that it was actually developed.
Cards were plugged into these boards using gold-plated steel pins that
protruded through the boards on both sides. On the "card side" the pins
formed a connecter for the cards, and on the other side, the pins were
square shaped in cross-section to form wire-wrap connecting points.
Engineering changes in the field were done by using a special powered
"delete" tool to scrape off connections on the outer planes, with new
connections made via wire-wrap. I used to do a lot of those. :-)
I gather that in the System/370 systems (1970), there were more layers
added, and the delete tool could be set to do either a one-layer or
two-layer deletion. ("surface delete" or "deep delete"). I wasn't working on
hardware by that time, so I don't know the details, but that suggests a
6-layer board. ... and in large-scale production.
--
Russ Holsclaw
Jim Stewart
The ground support computer system for Pershing 1A missiles had a
central interface assembly (the adapter, if there's any old Pershing G&C
techs out there) whose motherboard was said to have 12 layers and cost
USD 20,000. The system was deployed by 1970 and was probably designed
in the mid 60's.
jim
Mike Tomlinson
writes
>I think Alan Sugar (Amstrad) downloaded some money to IBM , hoping to
>grab a share of the MCA generation .
No, you're thinking of Apricot. They made several MCA-based machines.
--
Security-wise, NT is a server with a 'Kick me'
sign taped to it. -- Peter Gutmann
Roger Johnstone
<ste...@eircom.net> wrote:
> I recall it being rather more aggressive than that - IBM had some
> trouble with some major manufacturers, but not much. They all had a
> nightmare on their hands by the name of Taiwan, from where PC clones
> poured out at really low prices. At the time it was noticed that while
> the market was awash with PC clones there were no Apple clones, the catch
> was to clone an Apple you had to license some bits.
There were plenty of Apple II clones. The early Apple II computers used no
custom chips, just about everything was done with common 7400 logic chips.
Just like with the IBM PC/XT, this made it very easy to clone. The only
proprietary part was the ROM, but a lot of the Taiwanese clones came with
empty ROM sockets amd left it up to the buyer to 'obtain' <nudge-nudge,
wink-wink> their own ROMs. Franklin was the largest cloner until about 1985
when they got caught by Apple copying parts of the Apple II ROM. Now they
make electronic pocket dictionaries.
The Mac on the other hand was almost impossible to clone because of the ROM.
The Apple II/II+ only had a 12KB ROM which was mostly Microsoft BASIC (and
of course any company could license that), but the first Mac had a 64KB ROM
full of highly optimised machine code, and the ROM only got larger with each
model after that (it eventually reached 4MB before sense prevailed, it's
down to a reasonable 1MB in current models). There were several unauthorised
clones of the Mac, but all the ones I know of that actually shipped relied
on installing a ROM chip, usually from a Mac Plus. Their advantage over
something like, say, the Mac Plus you just rendered inoperable was that they
were faster and/or portable.
--
Roger Johnstone, Invercargill, New Zealand
Apple II - FutureCop:LAPD - iMac Game Wizard
http://homepage.mac.com/rojaws
________________________________________________________________________
"But what is it good for?"
Engineer at the Advanced Computing Systems Division of IBM
commenting on the micro chip, 1968
Steve O'Hara-Smith
"Roger Johnstone" <roj...@mac.com> wrote:
RJ> In article <20020430101020....@eircom.net>, Steve O'Hara-Smith
RJ> <ste...@eircom.net> wrote:
RJ> > was to clone an Apple you had to license some bits.
RJ>
RJ> There were plenty of Apple II clones. The early Apple II computers used no
'twas the Mac I had in mind - as you said:
RJ> The Mac on the other hand was almost impossible to clone because of the ROM.
Eric Smith
> Hehe, I'm very well aware of how boards were designed in the
> olden days. I would just be very surprised if people were still doing
> that en masse when MCA was introduced... That was the 90s,
> right ? :)
April of 1987.
I was at the 1987 Apple Worldwide Developer Conference at the Santa Clara
Convention Center, where the Macintosh II and Macintosh SE were being
introduced. IBM was had an event in the SCCC on the same day to introduce
the PS/2 line. A few of us went over to the IBM area to see what we
could find out about it. But since we hadn't been invited, we didn't get
to see anything. (Not surprising, the IBM invitees wouldn't have been
welcomed into the Apple conference either.)
When I did learn about the PS/2 line a few days later, I was impressed
with how little improvement had been made as compared to the PC AT.
The MCA bus was technically superior to what later became known as ISA,
but they really should have kept a few ISA slots in the machine (as was
later done with VESA local bus and then PCI).
The VGA display wasn't that much better than EGA. An improvement, to be
sure, but nothing very exciting.
What was really sad was that the PS/2 line didn't even initially use
the 386. I presume that they didn't use it since they'd deliberately
planned for OS/2 to only require a 286. Even so, the 386 would have
provided a performance boost. Obviously a pair of terrible decisions
in hindsight.
Rupert Pigott
news:qhn0vdl...@ruckus.brouhaha.com...
> "Rupert Pigott" <dark.try-eati...@btinternet.com> writes:
> > Hehe, I'm very well aware of how boards were designed in the
> > olden days. I would just be very surprised if people were still doing
> > that en masse when MCA was introduced... That was the 90s,
> > right ? :)
>
> April of 1987.
Corr blimey... MCA is *that* old. :P
[SNIP]
> When I did learn about the PS/2 line a few days later, I was impressed
> with how little improvement had been made as compared to the PC AT.
Well they were shooting for a high level of compatability with the PS/2
line. The cynical among us thought that the main innovation was the
mouse port. :P
> The MCA bus was technically superior to what later became known as ISA,
> but they really should have kept a few ISA slots in the machine (as was
> later done with VESA local bus and then PCI).
The 25,30 and 40 models had ISA slots, no idea if they were introduced
on the same date though.
> The VGA display wasn't that much better than EGA. An improvement, to be
> sure, but nothing very exciting.
VGA was a big improvement - maybe not in terms of resolution, but in
terms of the palette and the frame buffer depth it was a little better. The
standard certainly allowed for growth after those initial offerings (thank
god :P). VGA was also much easier to code for (IMHO).
> What was really sad was that the PS/2 line didn't even initially use
> the 386. I presume that they didn't use it since they'd deliberately
The model 70 and 80 used 386s, according to IBM the launch date was
April 1987... Wow, still can't get over how long ago that was. :)
> planned for OS/2 to only require a 286. Even so, the 386 would have
> provided a performance boost. Obviously a pair of terrible decisions
> in hindsight.
Anyone remember how much a minimum spec Model/80 cost ? I already
had a look at the various Mac II prices from back then and I swear I
nearly died of shock. :)
Cheers,
Rupert
Steve O'Hara-Smith
"Rupert Pigott" <dark.try-eati...@btinternet.com> wrote:
RP> The cynical among us thought that the main innovation was the
RP> mouse port. :P
It would seem to be the most succesful feature of the PS/2, I'd
say the cynics were about right.
Helmut P. Einfalt
Mod. 30 -- 8086 desktop, ISA only
Mod. 50 -- 80286 desktop, MCA
Mod. 60 -- 80286 tower, MCA
Mod. 70 -- 80386 desktop, MCA
Mod. 80 -- 80386 desktop, MCA
Later, the range was extended to include 486 and 586 up to the P90 in pure
MCA flavor). The highest pure-MCA machine were the 9595A (-xYx, P90 upgrade
board) and the Server500 (same), the highest MCA/PCI machine was the 6 x
Pentium Server720.
Structurally and functionally, the MCA bus of the time was pretty close to
what PCI was almost a decade later. Busmastering for all cards was a
standard feature, and the speeds achieved over the MCA bus were astonishing.
One of the greatest features, however, was the modular design of the
machines. Constructed for fully automated minimum-manpower-assembly, they
were composed of snap-together parts including floppy drive and HD design
(no cables on the first series models!), and what few screws there were,
were thumbscrews. You could and can take any of these machines
(30,50,60,70,80, also the later 90 and 95) apart and put it together again
in less than five minutes, requiring nothing more than a coin (in case one
of the thumbscrews had been tightened too fast and got stuck...).
There were quite a few more interesting innovations:
-- a cleverly devised airflow design,
-- sturdy housing (not these flimsy sheet-metal boxes or -- horribile
dictu -- plastic cases of today)
-- low maintenance cost resulting from the design (if necessary, the
technician came, popped out the defective part, put in a new Field
Replaceable Unit (FRU) and was off in less than five minutes),
-- the hardware configuration design. Each and every adapter had a unique
Adapter Description File (ADF) and its ADF code hardwired. Upon boot, the
machine would recognize the changes in the setup and ask for the ADF
(supplied on diskette), letting you set up all hardware features from the
keyboard -- no jumpers, no dip-switches...
-- the information panel on later models (8595, 9595). POST error codes were
standardized 8-digit figures displayed on-screen from the very first Mod.30.
The Hardware Maintenance Manual would give you additional information.
However, when the video adapter was faulty (or the machine was running as a
server w/o KB and screen), these figures were not accessible. So IBM added
the front panel showing the whole of the POST checkpoints as well as any
error code. Not that you'd need it very often once the machine was up and
running, though... yet, compare that to the cryptical beep sequences other
MFs use to the present day... My Server500 would tell me not only that it
found a defective memory module, but also in which of the eight slots it is
located -- all that with the monitor off...
These and similar "little" enhancements made them to be fine machines, and
extremely reliable. Contrary to popular folklore, the MCA PS/2s were a huge
success with corporate buyers, not with the "common customer", however. You
paid a (high) price, but you got excellent HW support (if necessary) -- not
exactly what the emerging "home buyer" market wanted or needed. And it was
the customers' tendency to buy cheaper (=ISA)rather than better as well as
internal frictions at Big Blue that brought about the end of MCA just when
the 64-bit MCA was almost ready for marketing...
Peripherals, too, matched IBM's PS/2 line in quality, at least in the first
years. I had to say goodbye to my very first keyboard I bought with my
Mod.30 in late 1987 just a fortnight ago after over 14 years of daily
service... (and I still haven't checked whether it's only the -- unpluggable
and replaceable -- cable...)
Helmut
still running half a dozen of these beast for dedicated tasks
--
Brought to you by IBM PS/2 power
comp.sys.ibm.ps2.hardware
Rupert Pigott
news:ab8jhk$bf8$01$1...@news.t-online.com...
[SNIP]
> These and similar "little" enhancements made them to be fine machines, and
> extremely reliable. Contrary to popular folklore, the MCA PS/2s were a
huge
I can vouch for that. There was a lab of the suckers at Uni, they must have
been
3 years old already when I got there and were still there 3 years later...
This
was with 24/7 abuse from Sports Science types seeing how many choccy
biccies they can stuff into the floppy drives. Haha.
They were very very tough little machines. Within 2 weeks of those same
students being let loose on the brand spanking new (and rather pricey)
Apricot
486DX2/66's half of those fancy new machines were dead...
Respect due to IBM for PS/2s - they were damn fine machines. R.I.P.
Cheers,
Rupert
Tony Ingenoso
The 25mhz A21's fix all that evilness.
"MSCHAEF.COM" <msc...@eris.io.com> wrote in message news:BLSz8.10972$n7.7...@bin8.nnrp.aus1.giganews.com...
Tony Ingenoso
were/are electrically very quiet compared to clone boxen. They were designed to survive in environments that would have other
machines twitching in spasms of resets and hangs ;->
"Rupert Pigott" <dark.try-eati...@btinternet.com> wrote in message news:aam963$jhr$1...@paris.btinternet.com...
>
> ...MCA had a ton of ground lines on it, classic case of IBM hyper-engineering in action.
Tony Ingenoso
would have left large market segments open. In 1987, the 286 was in its heyday, every vendor had a machine in that space. Also a
factor was the extreme cost of the 387 NPX compared to the 80287. This cost difference was so large, Compaq shipped some of their
first 386's with a socket for the 287 (yes, the 386 chip can work with a 287 with some restrictions)
You also simply had AT customers who were "comfortable" with the 286 chip, understood its known bugs (which were not nearly as
severe as those in the B1 step 386's of the period) and weren't in a hurry to move.
Seriously - those first 386 chips barely worked at all. We (I was working at IBM in Boca at the time) put all sorts of grody code
into OS/2 and PC DOS to work around bugs in the 386's. There was a lot of code in the OS's to work around 286 bugs too of course,
but those workaround were straight forward compared to the B1 386 errata. The errata were so bad, there was serious debate about
not supporting the B1 chips under OS/2 2.x, and handing out some free D-step chips to any customers with the early 70/80's that
would be affected.
"Eric Smith" <eric-no-s...@brouhaha.com> wrote in message news:qhn0vdl...@ruckus.brouhaha.com...
>
MSCHAEF.COM
Helmut P. Einfalt <hp.ei...@t-online.de> wrote:
>The line introduced in April 1987 was:
>
>Mod. 30 -- 8086 desktop, ISA only
>
>Mod. 50 -- 80286 desktop, MCA
>Mod. 60 -- 80286 tower, MCA
>
>Mod. 70 -- 80386 desktop, MCA
>Mod. 80 -- 80386 desktop, MCA
Was the 70 part of the initial release? The Byte magazine I have describing
the release only talks about tht 30, 50, 60, and 80. I thought the 70 came
later.
Other noteworthy PS/2's:
Model 25 - Low end, all in one box
Model 50Z - the Model 50 with 0 wait states, rather the 1, as in the 50
Model 55SX - Model 50 with a 386SX chip
Model 30/286 - A Model 30 with an 80286 and 16-bit ISA slots
-Mike
--
http://www.mschaef.com
Helmut P. Einfalt
>Was the 70 part of the initial release? The Byte magazine I have
describing
>the release only talks about tht 30, 50, 60, and 80. I thought the 70 came
>later.
>Model 25 - Low end, all in one box
That was an attempt to imitate the Mac-all-in-one-design. Same as on the
PS/1 series. Horrible ISA boxen...
>Model 50Z - the Model 50 with 0 wait states, rather the 1, as in the 50
...*and* upgradeable to 2 MB on board... Swapped my Mod.30 after one year to
a Mod.50, and upgraded that one to 50z later (mind you, I paid the money as
a private user... eeek!)
>Model 55SX - Model 50 with a 386SX chip
nice reliable office machine. the larger footprint Mod.30 case, but more
elegant than the 50/70 boxes
>Model 30/286 - A Model 30 with an 80286 and 16-bit ISA slots
This basically was an upgraded AT...
But since we are at it:
P70 / 75 "Lunchbox" -- the portable with the plasma display
9533 "Pizzabox" -- the "green" PC (ISA-based), in a pizza-parlour-take-away
sized case, with the latest in laptop/notebook technology squeezed into the
tiny case, 4 x PCMCIA, total power rating: 24 (!) W... complemented with an
appropriate LCD screen (9507 -- anyone got one to spare?) and the QuietTouch
space saver keyboard with the little mouse stick....
Prop in a bigger 2,5" HD and replace the PCMCIA riser with a ISA NIC,
install Linux on it and there you go with a nice MP3 server... *g* (friend
of mine did exactly that)
Helmut
Peter Ibbotson
news:ab8r78$jkk$01$1...@news.t-online.com...
> IIRC yes, at least when they were introduced in Europe, that is...
>
> >Was the 70 part of the initial release? The Byte magazine I have
> describing
> >the release only talks about tht 30, 50, 60, and 80. I thought the 70
came
> >later.
>
> >Model 25 - Low end, all in one box
>
> That was an attempt to imitate the Mac-all-in-one-design. Same as on the
> PS/1 series. Horrible ISA boxen...
I seem to remember it wasn't an ISA slot but some rejigged connector that
was an ISA bus with different pinout and order. We had one in the office,
but I honestly can't remember any more (Apart from wondering what the hell
IBM were upto)
Rupert Pigott
news:5KRB8.9701$AC2.40...@newssvr17.news.prodigy.com...
> Every other signal on the connector was a ground to reduce crosstalk and
help noise supression. It worked - the MCA machines
> were/are electrically very quiet compared to clone boxen. They were
designed to survive in environments that would have other
> machines twitching in spasms of resets and hangs ;->
I know. I loved it when I saw a PS/2 machine... I knew the critter could
be ignored, it would get on with it's work without complaint or reboot. :)
I'm a big fan of generous grounding. My other rather perverted fetish
is ECL logic because of the signaling it uses - which also happens to be
pretty resiliant to noise... I'm seeing a pattern here...
It's weird, but there's some kit out there where you actually forget you
are using it, because it just bloody works. This is good, it allows you to
solve problems with computers rather than solve computer problems.
Cheers,
Rupert
Joshua Hesse
:9533 "Pizzabox" -- the "green" PC (ISA-based), in a pizza-parlour-take-away
:sized case, with the latest in laptop/notebook technology squeezed into the
:tiny case, 4 x PCMCIA, total power rating: 24 (!) W... complemented with an
:appropriate LCD screen (9507 -- anyone got one to spare?) and the QuietTouch
:space saver keyboard with the little mouse stick....
I have one of those! (found it in a basement storeroom a couple years back)
No LCD tho... :-(
:Prop in a bigger 2,5" HD and replace the PCMCIA riser with a ISA NIC,
:install Linux on it and there you go with a nice MP3 server... *g* (friend
:of mine did exactly that)
Does he have some tips for doing that? (and which flavor of Linux won't
choke on the floppy drive) I ended up putting OS/2 2.1 for Windows on it,
which nuked the Win3.1 stuff. (Didn't have the last two Windows floppys.
Am also missing a bunch of other config disks.)
-Josh
--
"I have also mastered pomposity, even if I do say so myself." -Kryten
"You scream at them at the top of your lungs and then hit them over the head
with an immense wooden mallet? You're weird, sir." -Dave Brown on girls.
Mike Tomlinson
>Mod. 30 -- 8086 desktop, ISA only
>
>Mod. 50 -- 80286 desktop, MCA
>Mod. 60 -- 80286 tower, MCA
>
>Mod. 70 -- 80386 desktop, MCA
>Mod. 80 -- 80386 desktop, MCA
You forgot two of the more popular models:
Model 30 286 80286 desktop, ISA
Model 55SX 80386SX desktop, MCA
>You could and can take any of these machines
>(30,50,60,70,80, also the later 90 and 95) apart and put it together
>again
>in less than five minutes, requiring nothing more than a coin (in case
>one
>of the thumbscrews had been tightened too fast and got stuck...)
Yeah right, try changing the planar in < 5 mins.
Helmut P. Einfalt
>
>Model 30 286 80286 desktop, ISA
>Model 55SX 80386SX desktop, MCA
These were quite some time later...
>Yeah right, try changing the planar in < 5 mins.
Changing the planar required unscrewing the ports in the old models, and a
couple of screws in the 9595, I admit. Would need a screwdriver for that,
too. Might take a total of seven to eight minutes to fix that...
But then -- how often did you have to change the planar?
Helmut
Bradley Parker
> I seem to remember it wasn't an ISA slot but some rejigged connector that
> was an ISA bus with different pinout and order. We had one in the office,
> but I honestly can't remember any more (Apart from wondering what the hell
> IBM were upto)
It's a regular 16 bit ISA slot in the 9533 (PS/2 E).
I have two of the "Mighty E"s at home. Nice little boxen, absolutely
quiet! Use a couple 10/100 cards in the PCMCIA slots and you have a nice
cable modem/dsl gateway that uses almost no power.
I tried using one as a gateway with a 56K modem hanging off the serial
port. The ISA NIC worked fine but, for reasons only known to IBM, the
serial UART is a 16450 and wouldn't support fast enough transfer rates to
work with my modem. Every other PS/2 I've owned has buffered 16550 UARTs.
Someone else in this thread wanted to know how to get the floppy to work
under Linux... floppy=thinkpad IIRC. This boot time parameter shows the
"E"'s lineage although the planar looks similar to the model 35/40 units.
Certainly not being built as a MCA bus machine was a big screw up by IBM
IMO. Still, considering it's low power usage, low profile, and LCD screen,
the PS/2 E was ten years ahead of it's time.
Brad
--
Bradley M. Parker Systems Programmer II
Center for Computer-Aided Design Computing/Simulation Services
The University of Iowa E-Mail: b_pa...@ccad.uiowa.edu
208 ERF http://www.ccad.uiowa.edu
Iowa City, IA 52242 Phone: (319) 335-5723
MSCHAEF.COM
Bradley Parker <par...@ccad.uiowa.edu> wrote:
>
>Someone else in this thread wanted to know how to get the floppy to work
>under Linux... floppy=thinkpad IIRC. This boot time parameter shows the
>"E"'s lineage although the planar looks similar to the model 35/40 units.
What were the models 35 and 40?
-Mike
Tony Ingenoso
M35 came in the small Model 56 style case, M40 came in the big M77 style case.
"MSCHAEF.COM" <msc...@eris.io.com> wrote in message news:6GXB8.120472$Lj.94...@bin4.nnrp.aus1.giganews.com...
Peter Ibbotson
news:Pine.LNX.4.21.02050...@lacuna.ccad.uiowa.edu...
> On Tue, 7 May 2002, Peter Ibbotson wrote:
>
>
> > I seem to remember it wasn't an ISA slot but some rejigged connector
that
> > was an ISA bus with different pinout and order. We had one in the
office,
> > but I honestly can't remember any more (Apart from wondering what the
hell
> > IBM were upto)
>
> It's a regular 16 bit ISA slot in the 9533 (PS/2 E).
>
> I have two of the "Mighty E"s at home. Nice little boxen, absolutely
> quiet! Use a couple 10/100 cards in the PCMCIA slots and you have a nice
> cable modem/dsl gateway that uses almost no power.
Doh! I was thinking of the PS/1.
Mike Tomlinson
<0009...@bigred.unl.edu> writes
>which flavor of Linux won't
>choke on the floppy drive)
'linux floppy=thinkpad' at the LILO prompt.
Mike Tomlinson
<hp.ei...@t-online.de> writes
>But then -- how often did you have to change the planar?
Too often, while I was an IBM engineer. The Model 50z planar wasn't the
most reliable design ever, and Seagate hard discs in the 55sx suffered
from stiction.
Model 80s could be a pain too - if they were well populated with cards,
those needed to be removed along with the disk cage and about 10 screws
before the planar could be removed. The case would without fail be
locked and the luser would have lost the keys.
Model 56s and 77s had dodgy power supplies.
The worst aspect of the PS/2s IMHO was the lack of a dust flap on the
floppy drives, meaning they inevitably filled up with dust and crap.
They appeared to be the main air inlet for the machine's PSU fan :)
Helmut P. Einfalt
>locked and the luser would have lost the keys.
Oh yeah... not to forget the password, and if applicable, even the PAP
(Privilege access password) -- and if *that* was unknown, *nothing helped
except changing the planar...
>Model 56s and 77s had dodgy power supplies.
Magnatec from Italy, to be precise. The other PSs (e.g. Schrack from
Austria) worked fine. But Magnatec definitely were a pain...
>The worst aspect of the PS/2s IMHO was the lack of a dust flap on the
>floppy drives, meaning they inevitably filled up with dust and crap.
>They appeared to be the main air inlet for the machine's PSU fan :)
Well -- they *did* come with a dummy cardboard floppy to be inserted in the
FDD for transportation... one could use that to keep the bunnies out...
(still have mine from the Mod.30, 1987 -- *g*).
Goran Larsson
Mike Tomlinson <mi...@jasper.org.uk> wrote:
> The worst aspect of the PS/2s IMHO was the lack of a dust flap on the
> floppy drives, meaning they inevitably filled up with dust and crap.
> They appeared to be the main air inlet for the machine's PSU fan :)
Instead of sucking air into the box through a filter it must be much
cheaper to blow air out of the box and use the floppy drive as a
filter. I always assumed that it was designed that way to save the cost
of a filter on the fan. The floppy drive will probably work ok until
the warranty expires.
--
Göran Larsson http://www.mitt-eget.com
Heinz W. Wiggeshoff
>
> Instead of sucking air into the box through a filter it must be much
> cheaper to blow air out of the box and use the floppy drive as a
> filter. I always assumed that it was designed that way to save the cost
> of a filter on the fan. The floppy drive will probably work ok until
> the warranty expires.
Then there's option C: no friggin fans at all. I've suffered much
irritation and expense with 486-DX4 and AMD K6 CPU and power supply
fans, to the point where I spend most of my FreeNet time on the
286 system. Was I losing my hearing? The machine was sooo quiet,
with only the sound of two hard drives purring. Never any of that
squealing into failure mode crap.
Checked the back of the machine a few days ago: the power supply fan
had packed it in long ago. No complaints from the CPU, FPU, 7 Mb RAM,
two HD, two floppies and the modem.
Now that's engineering.
Stefan Skoglund
> I recall it being rather more aggressive than that - IBM had some
> trouble with some major manufacturers, but not much. They all had a
The big bad point was IBM's requirement that MCA-licensees should buy
an ISA-license from ibm too.
MicroChannel was eons better than ISA but you still had to
have install disks with hw description files on.
Stefan Skoglund
> (30,50,60,70,80, also the later 90 and 95) apart and put it together again
> in less than five minutes, requiring nothing more than a coin (in case one
> of the thumbscrews had been tightened too fast and got stuck...).
You forgot the RS 6000 machines !!!
Stefan Skoglund
> >Model 55SX - Model 50 with a 386SX chip
>
> nice reliable office machine. the larger footprint Mod.30 case, but more
> elegant than the 50/70 boxes
You forget the onboard SCSI which my machine do have.
Rupert Pigott
news:ZAT1k7A9...@jasper.org.uk...
> In article <ab950l$j0g$02$1...@news.t-online.com>, Helmut P. Einfalt
> <hp.ei...@t-online.de> writes
>
> >But then -- how often did you have to change the planar?
>
> Too often, while I was an IBM engineer. The Model 50z planar wasn't the
> most reliable design ever, and Seagate hard discs in the 55sx suffered
> from stiction.
>
> Model 80s could be a pain too - if they were well populated with cards,
> those needed to be removed along with the disk cage and about 10 screws
> before the planar could be removed. The case would without fail be
> locked and the luser would have lost the keys.
We had a shipment of 6 PCs direct from HP some years ago... They
fell apart if you breathed on them, no screwdrivers necessary. That was
lucky as it happens because the motherboards were just waving around
vaguely so that they were grounded against the RF shield foil on the
plastic case.
Never touched a HP widget since, that was so piss poor that they just
lost my custom for the rest of my life. Doesn't help that they're helping
INTEL push that IA-64 BS down our necks too. We've spent 10
years getting a solid ANSI C compiler together, now they want to
throw in an architecture which requires *MAJOR* upheaval in the
compilers to get anything close to current desktop PC performance.
Screw them and the PC they rode in on.
Cheers,
Rupert
Rupert Pigott
news:abcbud$2fk$1...@freenet9.carleton.ca...
I'm looking at putting together my next machine based on an embedded
controller PC... Check out the EBC363 from www.nexcom.com... You
can run the sucker fanless... Looks like a great option for me. Plus it has
something approximating to a 3D accelerator on board. :)
Cheers,
Rupert
Helmut P. Einfalt
I didn't -- but I can't comment on them from experience...
I'm waiting to get one, though -- and then up into the big AIX adventure...
Tony Ingenoso
"Goran Larsson" <h...@invalid.invalid> wrote in message news:GvtCz...@approve.se...
Tony Ingenoso
who was current on royalties or (more likely) had some interesting intellectual property could do a patent swap with IBM and get
away without paying monetarily - IBM and most big corps love to do patent swap deals. A "carbon copy" cloner with no IP of their
own to speak of didn't have anything to negotiate with other than cash - which they were reluctant to part with ;->
"Stefan Skoglund" <ste...@ebox.tninet.se> wrote in message news:3CD9B54B...@ebox.tninet.se...
Walter Rottenkolber
Rupert Pigott wrote in message ...
>Heinz W. Wiggeshoff <ab...@FreeNet.Carleton.CA> wrote in message
>news:abcbud$2fk$1...@freenet9.carleton.ca...
>> Goran Larsson (h...@invalid.invalid) writes:
>> Then there's option C: no friggin fans at all. I've suffered much
>> irritation and expense with 486-DX4 and AMD K6 CPU and power supply
>> fans, to the point where I spend most of my FreeNet time on the
>> 286 system. Was I losing my hearing? The machine was sooo quiet,
>> with only the sound of two hard drives purring. Never any of that
>> squealing into failure mode crap.
>>
>> Checked the back of the machine a few days ago: the power supply fan
>> had packed it in long ago. No complaints from the CPU, FPU, 7 Mb RAM,
>> two HD, two floppies and the modem.
>> Now that's enginee
>Cheers,
>Rupert
>
That's one big thing I like about my Kaypro II, no fan, no noise. Ram disk
instead of hard drive -- some of the early hard drives could sound like a
jet on takeoff. Computers should be seen and not heard.
Jerry Pournelle keeps recommending a PC case that is ultra quiet. The local
library got some new Dell's with the flat screens that seem much quieter
than earlier PCs.
Originally IBM picked the noise level of the fan based on studies of average
office noise levels, about 65 db IIRC. Unfortunately these studies were made
when manual typewriters were in common use.
Walter Rottenkolber
Heinz W. Wiggeshoff
>
...
> office noise levels, about 65 db IIRC. Unfortunately these studies were made
> when manual typewriters were in common use.
MANUAL? Check the noise from a daisy wheel terminal, perhaps supplemented
by some dot matrix printers for the higher frequencies.
Back in 84, one weasel-dolt manager expected shit hot design and
programming, but surrounded me with _four_ Trendata 4000 doing non-
stop word processing, (ATS/360 offspring) ! Add in the hen party
chatter, and one can see why that outfit is mams up.
greymaus
And he complained that there was no humor!
--
Greymaus;
Follow up , don't e-mail , my killfile is savage ;
Next Year In Hak Nam ;
Mike Tomlinson
<ste...@ebox.tninet.se> writes
The 55SX doesn't have onboard SCSI.
Stefan Skoglund
> The 55SX doesn't have onboard SCSI.
MY has -- maybe it isn't really a PS/2 55SX but what is it then ?
Tony Ingenoso
to the riser card at the top.
There was a line of machines called PS/55 - these were Japanese models, quite different than anything else. They had different
planars, had the "Atlas" DBCS video cards, etc.
"Stefan Skoglund" <ste...@ebox.tninet.se> wrote in message news:3CDC37E6...@ebox.tninet.se...
Mike Tomlinson
<ste...@ebox.tninet.se> writes
>MY has -- maybe it isn't really a PS/2 55SX but what is it then ?
Some of the 486SLC CPU based models did have SCSI. 56SLC maybe, but I
can't remember for sure. Different case to the 55SX, with a tamper
switch and 2.88Mb floppy.
Helmut P. Einfalt
>can't remember for sure. Different case to the 55SX, with a tamper
>switch and 2.88Mb floppy.
The Mod.55 had MFM/RLL standard
The Mod.56/57 had SCSI on-board
A nice collection of links ot the various models' features (mostly IBM
original material) can be found at
http://ps2page.tripod.com/ps2specs.htm
The following links are (expanded and enhanced) mirrors of Louis Ohland's
"Ardent Tool of Capitalism" page, *the* repository of the common PS/2
knowledge gathered in comp.sys.ibm.ps2.hardware:
http://homepage.ntlworld.com/tim_n_clarke/Systems.htm
http://www.walshcomptech.com/ohlandl/
Additional information on PS/2 and MCA machines is provided by Peter Wendt's
collection at
http://members.aol.com/mcapage0/mcaindex.htm
(Good place for finding ADFs).
Each of these sites has a couple of links to related material (various tools
and tweakers, processor upgrades IBM never even dreamt of, Linux on PS/2s,
and the like).
Ir you're ever looking for information on PS/2s, check these sites (and the
above NG, obviosuly) -- and if you can't find it there, you probably won't
find it at all...
Helmut P. Einfalt
> :install Linux on it and there you go with a nice MP3 server... *g*
(friend
> :of mine did exactly that)
>
>Does he have some tips for doing that?
See Peter Wendt's
http://members.aol.com/mcapage0/ps2elinx.htm