PDA

View Full Version : Your favorite IBM 5150 reflections, historical facts, technical trivia and scuttlebut



dmemphis
January 30th, 2017, 09:19 AM
I was unsure that this might belong in the history forum, but
since it is machine specific, I chose here, plus I really
want an audience with 5150 enthusiasts. Pardon the faux pas if
this isn't quite right etiquette.
Thanks in advance!

Wild stuff will have to be vetted or reference provided! :)

Chuck(G)
January 30th, 2017, 09:45 AM
I remember my 5150; miserably underpowered thing, almost painful to use with 64KB and one SS floppy drive. I had various CP/M machines that worked much better. After DOS 2.0 came out, I expanded the thing with a Quadram card and my own SA-1000 hard disk setup and things were a bit better. Eventually I jettisoned the thing for a cheap XT clone and then a PC AT clone.

lyonadmiral
January 30th, 2017, 10:00 AM
I swear I was born 30 years ahead of my time, of course I think part of that has to do with being raised by much older grandparents; early 70's and early 80's.

When I was in high school 1995-1999 there was an interesting schism. The computer science department was all Apple. Apple IIc's when I entered "middle school" but LC 580's without CD-ROM connected to a Power PC based server. The business department however was PC and way behind. Mind you most of my business classes happened in my junior and senior year, so 98 & 99' however that year the machines were mostly IBM PS'2 Model 25's, a 486/33 clone, and a Pentium 75 clone. The reflection of course comes from the fact that I chose to use way in the back of the room with the 5150 with single 5.25 and aftermarket hard disk to do my work. Unfortunately since the senior-itis was so bad I almost failed. Still, best classroom experience since no one was going to steal my work. The PS/2's and the clones had 3.5" only.

The software we used mostly was Electric Pencil (FUN!) and PFS: First Choice 2.0 (also FUN!)

dmemphis
January 31st, 2017, 07:33 PM
How about things like this:

According to a widely quoted account of by Bill Gates, one such found here:
https://books.google.com/books?id=lB4PAwAAQBAJ&pg=PA97&lpg=PA97&dq=8088+video&source=bl&ots=3UQU-V5-kR&sig=OFyc4hl_21DFq-eGSbeAjZaiIMU&hl=en&sa=X&ved=0ahUKEwiL4bDQuerRAhUmrVQKHVt8CH4Q6AEIVTAJ#v=on epage&q=8088%20video&f=false

The 5150 motherboard allegedly took only 40 days to design once the 8088 was chosen as the processor.
The designer, Lou Eggebrecht, authored the book "Interfacing to the IBM PC" and today is a board member of Broadcom/Avago.

krebizfan
January 31st, 2017, 08:21 PM
There are many stories about the IBM PC development.
Mark Dean's explanation of the choice of Intel over others can be heard https://www.youtube.com/watch?v=cYPwVOUcCnc

Some years ago, there was a retrospective involving members of the team responsible for actually building the prototype IBM PC but I can't find a link to it now.

br44
February 3rd, 2017, 09:51 AM
I was very young when I first experienced the IBM 5150. At the time I didn't even know what it was called besides computer because I was so young. It was not the first computer I ever used, that would be the commodore 64. However the IBM was the first computer that made an impression on me enough to want to get into computers.

The computer its self was a gift from one of my mothers co workers. The lady was a doctor and she could afford expensive things like that. She gave it to us because we had nothing and by then the 5150 was way outdated. It came with the monitor, keyboard, printer, tons of paper, and a printer paper box full of 5.25 floppies. Some of my earliest video gaming memories were on that computer with things such as sokoban, jeopardy, chess, and wheel of fortune. We always treated that computer like a gem. We went to great trouble to keep it clean and dust free. It even came with these rubberized plastic dust covers for everything.

Unfortunately my parents finally got rid of it when they got a more modern computer :( . They replaced it with a Micron with a pentium II in it and windows 95. That computer is also nostalgic for me but that is beyond the scope of this post.

SpidersWeb
February 6th, 2017, 09:40 PM
I grew up in a country that rarely spotted an IBM outside of a large corporation, and was a few years late, but through the 90's I became quite the PC enthusiast and started digging in to the history more, my first collectible was an IBM 5160 (not a 5150) this would've only been around 2001.

I wrote to IBM asking for more information about the original IBM PC and the PC XT, and I actually got a package in the mail. They sent me several original product photo prints and a letter thanking me for my interest.

I've since lost the letter, but that's a memory that will stick with me for a long time. The photographs are framed around my computer room in the house.

d5aa96
February 7th, 2017, 08:12 AM
I worked at IBM in Boca Raton in the mid 80's and knew many of folks who worked on the development of the 5150 PC. I learned that many of the 5150 manuals contained "easter eggs" put in by the people who authored or edited the manuals in the form of code samples that contained the names of actual people and places important to those who worked on the project.
For example, the BASIC manual contains many code samples that reference Boca Raton and Delray Beach and other addresses and names of real places and people. Many of the numbers used in the examples are the ages, birthdays, or street numbers important to the authors.
Rick

lyonadmiral
February 7th, 2017, 08:26 AM
I worked at IBM in Boca Raton in the mid 80's and knew many of folks who worked on the development of the 5150 PC. I learned that many of the 5150 manuals contained "easter eggs" put in by the people who authored or edited the manuals in the form of code samples that contained the names of actual people and places important to those who worked on the project.
For example, the BASIC manual contains many code samples that reference Boca Raton and Delray Beach and other addresses and names of real places and people. Many of the numbers used in the examples are the ages, birthdays, or street numbers important to the authors.
Rick

That is real neat. The 80's will always be my favorite decade and some-days I would like to live it again...

bobba84
February 7th, 2017, 01:46 PM
I worked at IBM in Boca Raton in the mid 80's

Have you got any pictures of the operations there? Especially assembly line / etc? They always fascinate me!

For the record, I got my first 5150 two years ago, shipped from the USA to Australia. Lots of fun!

dmemphis
February 7th, 2017, 03:38 PM
I worked at IBM in Boca Raton in the mid 80's and knew many of folks who worked on the development of the 5150 PC. I learned that many of the 5150 manuals contained "easter eggs" put in by the people who authored or edited the manuals in the form of code samples that contained the names of actual people and places important to those who worked on the project.
For example, the BASIC manual contains many code samples that reference Boca Raton and Delray Beach and other addresses and names of real places and people. Many of the numbers used in the examples are the ages, birthdays, or street numbers important to the authors.
Rick

That's a real treat. Thanks exactly what I was looking for. I'll try to look some of those up.
If you come by any direct references please pass them along.

Stone
February 7th, 2017, 03:41 PM
Have you got any pictures of the operations there? Especially assembly line / etc?Was that a serious question or did you omit the smiley?

nc_mike
February 7th, 2017, 04:04 PM
I worked in a Big Blue mainframe programming lab as a new hire and received one of the first PCs that came off the line. After playing with it for a week I was asked by my lab management what I thought of it, and I replied, half joking "Our 3081 had a baby - and its going to eat its mother" A few short years later I realized that my assessment was no joke as mainframe sales plummeted while companies shifted their budgets to buy PCs by the truckload.

bobba84
February 7th, 2017, 05:18 PM
Was that a serious question or did you omit the smiley?

It was a serious question...

Casey
February 7th, 2017, 08:11 PM
Yes, the original machine was unimpressive at best, which is why I bought an Epson QX-10*, with that lovely 640x400 monographics monitor, 256Kb of ram when the PC still only came with 64, and two 380K floppy drives! The keyboard was very nice as well. When I finally bought a PC compatible (Compaq Portable) the keyboard layout drove me nuts. The Portable was nice, but I discovered that Compaq wasn't hardware compatible. Shoulda bought an XT. Eventually ended up with an XT "turbo" clone. Ran a fidonet BBS on that for a fair amount of time.

Interesting tidbit about the QX-10 keyboard: the layout bears a strong resemblance to the later Model M 104-key layout... :)




*At the time I was unfamiliar with the saying "no one ever got fired for buying IBM."

krebizfan
February 7th, 2017, 08:49 PM
On the other hand, in 1983, the QX-10 with 256k was a $3,000 system while the 64K IBM PC was a $1,300* machine. Adding drives and memory resulted in a system that fairly closely matched the fully loaded QX-10 in both price and specs but the lack of Valdocs was probably the biggest advantage of the IBM PC.

* Price cuts thanks to the introduction of the XT.

Scali
February 8th, 2017, 12:59 AM
I was very young when I first experienced the IBM 5150. At the time I didn't even know what it was called besides computer because I was so young. It was not the first computer I ever used, that would be the commodore 64. However the IBM was the first computer that made an impression on me enough to want to get into computers.

For me the PC made the opposite impact.
We've had a C64 since 1984 or so, and at around 1988 my dad bought a PC clone. It was an 8088 at 4.77 MHz with 640k, 2x 5.25" DD drives, no HDD and Hercules graphics. I later found out that it had a 'turbo' as well, allowing it to run at 9.54 MHz. But my initial impression would have been very similar to what a 5150 or 5160 would be like.
And I found it rather underwhelming compared to the C64. It felt sluggish, and everything was very 'dusty' and 'industrial'. We later got a color monitor, because we found that the video adapter was also compatible with CGA and Plantronics. But that didn't really improve my impression. The ugly 4 colours were no match for what I was used to on my C64. And it was just terrible at most games, with no sprite hardware, and no ability to scroll. And don't get me started about the 'sound' :)
I never quite understood why it was so successful.

Real IBMs were not that common here. But I eventually got one about 2 years ago (a 5160), when I went back to my roots, and needed a 100% compatible machine for 8088 MPH.
Up to then I used two clones, a Philips P3105 (which we had in school when I was young), and a Commodore PC20-III (basically the same as my first PC, which was a PC10-III, aside from having a HDD onboard).

vwestlife
February 8th, 2017, 05:03 AM
I worked at IBM in Boca Raton in the mid 80's and knew many of folks who worked on the development of the 5150 PC. I learned that many of the 5150 manuals contained "easter eggs" put in by the people who authored or edited the manuals in the form of code samples that contained the names of actual people and places important to those who worked on the project.
For example, the BASIC manual contains many code samples that reference Boca Raton and Delray Beach and other addresses and names of real places and people. Many of the numbers used in the examples are the ages, birthdays, or street numbers important to the authors.
Rick
And in the DOS manual they frequently used "FOO.BAR" as an example file name. As a young kid when I first read that, the reference totally flew over my head...

geneb
February 8th, 2017, 05:28 AM
And in the DOS manual they frequently used "FOO.BAR" as an example file name. As a young kid when I first read that, the reference totally flew over my head...

Keep in mind that "FOO.BAR" doesn't really link to FUBAR - at least in this kind of computer reference. (google "foo bar baz") :)

g.

Trixter
February 8th, 2017, 10:00 AM
My favorite "easter egg" related to the 5150 is the same thing we exploited in 8088 MPH: CGA's 80-column text mode does not force disabling the color burst signal on the composite video output. By all rights it should, since 80-column text is nearly unreadable with it turned on, but for some reason it isn't forced off. VileR used that quirk to paint some pretty amazing 1000-color images by exploiting the color artifacting present in that video mode combination.

My second-favorite is how the PC speaker can be tied to PIT channel 2. By doing this, the PC could not only output a square wave independently from CPU control (ie. it keeps playing in the background), but the PIT has a 1-shot mode that allows you to pulse the speaker at very short intervals, which you can abuse to produce some crude pulse-width modulation. This combination allows you to reliably and consistently produce sampled sound output from the speaker, albeit at a low volume.

I've always wondered how intentional the speaker/PIT stuff was; I've ordered Eggebrecht's book and hope I'll find an answer in there. If not, I might track him down to ask him personally.

offensive_Jerk
February 8th, 2017, 10:23 AM
My second-favorite is how the PC speaker can be tied to PIT channel 2. By doing this, the PC could not only output a square wave independently from CPU control (ie. it keeps playing in the background), but the PIT has a 1-shot mode that allows you to pulse the speaker at very short intervals, which you can abuse to produce some crude pulse-width modulation. This combination allows you to reliably and consistently produce sampled sound output from the speaker, albeit at a low volume.

I've always wondered how intentional the speaker/PIT stuff was; I've ordered Eggebrecht's book and hope I'll find an answer in there. If not, I might track him down to ask him personally.

Is that what they did for that "Magic Mushroom" song from the air freshener commercial that played on the PC speaker?

Cimonvg
February 8th, 2017, 01:14 PM
hello
a new observation (but utterly useless) information . When you run ibm 5150+inboard 386 and install Netroom3 . Upon start of the RM386.exe the relay controling the cassette tape port "clicks"
.. kind of funny, i think :)
/cimonvg

bobba84
February 8th, 2017, 01:50 PM
hello
a new observation (but utterly useless) information . When you run ibm 5150+inboard 386 and install Netroom3 . Upon start of the RM386.exe the relay controling the cassette tape port "clicks"
.. kind of funny, i think :)
/cimonvg

Strange! Maybe it probes that memory address or whatever to see if it's in a real 5150?

Cimonvg
February 8th, 2017, 02:41 PM
yah, kind of..
When RM386.exe starts, one click is heard. Then when i press Ctrl+Alt+Del and the pc begins to reboot another click is heard (perhaps a click when entering protected mode and at reboot a click on the way back to real mode ??)..
/cimonvg

vwestlife
February 8th, 2017, 02:48 PM
Some demos and animations toggle the 5150's cassette relay to simulate the sound of a ticking clock or to make buzzing sounds (as I'm sure we've all done by putting the MOTOR command in a FOR-NEXT loop in BASIC!).

Trixter
February 8th, 2017, 04:48 PM
Is that what they did for that "Magic Mushroom" song from the air freshener commercial that played on the PC speaker?

Yes, although their method was a little odd; instead of just letting the PIT handle the timing, they continuously polled the timer to see if it was time to do something. Or a variant of that, I can't remember. But it wasn't the typical straightforward way of using PIT channel 0 to time the samples and PIT channel 2 to time the pulses.


Some demos and animations toggle the 5150's cassette relay to simulate the sound of a ticking clock or to make buzzing sounds (as I'm sure we've all done by putting the MOTOR command in a FOR-NEXT loop in BASIC!).

Indeed!


https://www.youtube.com/watch?v=qifVDyeug0k

Casey
February 8th, 2017, 04:53 PM
Hate to think what the freight was on that sucker...

Casey
February 8th, 2017, 05:02 PM
Yes, Valdocs sucked rocks. The QX-10 was a fairly decent performer under regular CP/M. Once had a demo, doubtlessly written in assembler, which looked like the "arty" demo that used to get shipped with Turbo Pascal. Bloody thing flew thru the graphics.

My memory tells me there was a closer match in price. I'm thinking ~$2,400 or $2,700 for the QX-10... It's been a long time. I do recall a friend telling me all the PC had available was a "buggy word processor," and that the DOS was a cheap copy of CP/M. Was that price with 2 floppies and a CGA display? I might have gotten it before the XT came out. As I said, it's been a long time.

Either way the PC was the wave of the future. :-/

dmemphis
February 8th, 2017, 05:31 PM
Valdocs was ambitious for its time. If it had not been buggy it might have changed the course of a lot of events.
We are straying off topic here, but several years ago contemplating keyboard designs, I found their ideas regarding the
HASCI keyboard (Human Application Standard Computer Interface) fascinating
http://electrickery.xs4all.nl/comp/qx10/doc/pg2v_ch1.txt
And wow, according to this, Valdocs was planned for the ST:
http://mcurrent.name/atarihistory/tramel_technology.html
Also saw that portions of Valdocs were written in Forth. Forth keeps popping up odd places I look.
http://electronicdesign.com/blog/rising-star

Ok, back to those early IBM machines...

dmemphis
February 8th, 2017, 06:11 PM
Yes, although their method was a little odd; instead of just letting the PIT handle the timing, they continuously polled the timer to see if it was time to do something. Or a variant of that, I can't remember. But it wasn't the typical straightforward way of using PIT channel 0 to time the samples and PIT channel 2 to time the pulses.



Indeed!


https://www.youtube.com/watch?v=qifVDyeug0k

I ran across that video a ways back and forgot about it. Thanks for bringing it up.
I may try to port that to qbasic sometime to make it handier.

dmemphis
February 8th, 2017, 06:41 PM
Yes, Valdocs sucked rocks. The QX-10 was a fairly decent performer under regular CP/M. Once had a demo, doubtlessly written in assembler, which looked like the "arty" demo that used to get shipped with Turbo Pascal. Bloody thing flew thru the graphics.

My memory tells me there was a closer match in price. I'm thinking ~$2,400 or $2,700 for the QX-10... It's been a long time. I do recall a friend telling me all the PC had available was a "buggy word processor," and that the DOS was a cheap copy of CP/M. Was that price with 2 floppies and a CGA display? I might have gotten it before the XT came out. As I said, it's been a long time.

Either way the PC was the wave of the future. :-/

>Either way the PC was the wave of the future

It was, despite the criticisms of what it was and what it wasn't,
given the quick pace that it was put together they got a lot of things right.
Here's my list of some of the right stuff:
- provided promise of nearly 10x increase in memory over the common computers of the time.
(even if it was &(*&%()$*% segmented)
- provided a solid slot architecture for expansion.
- provided awesome documentation in the tech reference line.
Notable examples I have from other companies are good were the Model I tech ref
and the Osborne 1 tech ref. But interestingly these didn't lead to a large clone market.
- A decent DOS command/batch language set for the time- that was reliable and reasonably fast.

Trixter
February 8th, 2017, 08:07 PM
Valdocs was ambitious for its time. If it had not been buggy it might have changed the course of a lot of events.

It needed speed as well as less bugs. I started laughing when I read "Entering text becomes a disconcerting pastime when the screen display lags as many as 60 characters behind your typing, and you lose characters."

I'd love to read a write-up or see a video demonstration of what gave Valdocs so much potential. Topic for another thread.

Trixter
February 8th, 2017, 08:10 PM
I may try to port that to qbasic sometime to make it handier.

It won't run at the correct speed if you do. All of the timing was written on an IBM 5150 with ROM BASIC. qbasic on the 5150 will make it run slower, and qbasic on any other system will make it run too fast.

dmemphis
February 8th, 2017, 08:32 PM
It needed speed as well as less bugs. I started laughing when I read "Entering text becomes a disconcerting pastime when the screen display lags as many as 60 characters behind your typing, and you lose characters."

I'd love to read a write-up or see a video demonstration of what gave Valdocs so much potential. Topic for another thread.

>It needed speed as well
Agreed. It wasn't the first failure due to performance. Context MBA in particular comes to mind.
I have yet to run either of those.

There's quite a bit written about Valdocs. I guess it really was the first integrated office wysiwyg package offered on a microcomputer. I imagine that did create quite a buzz.

Scali
February 8th, 2017, 11:29 PM
Is that what they did for that "Magic Mushroom" song from the air freshener commercial that played on the PC speaker?

If you want, I should have a disassembly of that routine somewhere.
As Trixter says, it uses a slightly peculiar approach, but in essence it is PWM.
Why do we see various approaches to PWM, I wonder. Was it because these were early attempts, and they had not perfected the technique yet? Or was it because RealSound(tm) patented the 'perfect' technique already, and they had to do something different to avoid the patent?

One thing I'd like to say about PWM is that not everyone seemed to quite get it right. Early demoscene productions weren't always that great. The MOD player in Crystal Dream can use PWM on PC speaker at various rates, but they did not seem to map the 8-bit samples to the correct values for PWM at these different rates, so it sounds very distorted.
The MOD routine that reenigne did for 8088 MPH (the final version) has carefully selected values for PWM, giving you maximum dynamic range with no distortion. The result is pretty 'crystal clear'.
The RealSound stuff also sounds very good. These people knew what they were doing to push the technique for maximum quality.
There are two ingredients to the technique I'd say:
1) Pre-process your data so you generate the proper PWM values for each sample
2) Perform the per-sample timing with minimal jitter

In various routines I see/hear at least one of these two going horribly wrong.

dmemphis
February 9th, 2017, 10:33 AM
>It needed speed as well
Agreed. It wasn't the first failure due to performance. Context MBA in particular comes to mind.
I have yet to run either of those.

There's quite a bit written about Valdocs. I guess it really was the first integrated office wysiwyg package offered on a microcomputer. I imagine that did create quite a buzz.

Just one more thing on Valdocs for those interested with related history,
you should be aware of Roger Amidon and all he did on the east coast USA.
http://obits.nj.com/obituaries/trenton/obituary.aspx?page=lifestory&pid=180015471

Trixter
February 9th, 2017, 02:46 PM
Why do we see various approaches to PWM, I wonder. Was it because these were early attempts, and they had not perfected the technique yet? Or was it because RealSound(tm) patented the 'perfect' technique already, and they had to do something different to avoid the patent?


Access started using RealSound in 1988 with Echelon (the box and materials state "Patent Pending" next to the RealSound logo). My research shows that most games with PWM were winging it based on techniques the programmers learned on other systems, with hard-coded timing loops or watching-the-timer loops. Right around 1987, proper PWM with both PIT channels arrived in a few games; a year later, RealSound emerged. After that, I know of at least one game (Manhole for DOS) that used a deliberately different technique, deliberately patented as well, as a "screw you" to Access and their patent. So I think it was a mixture of everything, including a small amount of prior art that should have made the RealSound patent invalid but for some reason wasn't. (I can send you my research document offline if you're curious; email me)



The RealSound stuff also sounds very good. These people knew what they were doing to push the technique for maximum quality.


I interviewed Steve Witzel two decades ago and he mentioned they took the additional step of having an audio engineer master the audio (likely with dynamic compression and limiting) for the 3" speaker before digitization.

Sorry to hijack the thread! I will try to get it back on track with a nostalgic memory of my very first IBM 5150 experience: Downloading software from a BBS using PC-Talk (https://en.wikipedia.org/wiki/PC-Talk) in 1983. I was a young teen, and a friend's wealthy father had a PC with a 300 baud modem. We dialed a Chicago BBS (we didn't know there was more than one BBS, we called it "dialing the BBS"), browsed around to find Bushido (http://www.mobygames.com/game/dos/bushido), and downloaded it directly to floppy using xmodem (https://en.wikipedia.org/wiki/XMODEM). This should haven taken about 10 minutes but took nearly 30 because we had to keep restarting the transfer due to noisy phone lines, and xmodem-crc wasn't implemented in PC-Talk yet (probably wasn't invented yet?) We then needed to download whatever the unarchiving program was (I want to say it was a .zoo archive but it may have been an early LBR (LU, LU86, LAR, LUPC) or an old SQ format) and then unarchive it from one floppy drive to the other. We were then greeted with this:


https://www.youtube.com/watch?v=nYuO_i2SOgM

It certainly shaped my future. I was quite the digital miscreant after that (and still am, I suppose). Prior to that afternoon, all of my trading had been via disk swapping Apple II games.

dmemphis
February 9th, 2017, 05:37 PM
Its all good!
The PC speaker details are right in there with the topic title "5150 reflections, historical facts, technical trivia and scuttlebut".

Scali
February 9th, 2017, 11:30 PM
Right around 1987, proper PWM with both PIT channels arrived in a few games; a year later, RealSound emerged.

I would say that pretty much proves that this wasn't a feature that was deliberately designed into the 5150 by IBM.
Otherwise I would expect the technique to be showcased (and documented) way earlier (possibly even by IBM themselves at introduction).


I interviewed Steve Witzel two decades ago and he mentioned they took the additional step of having an audio engineer master the audio (likely with dynamic compression and limiting) for the 3" speaker before digitization.

Yes, and it shows.
I firmly believe that audio programming should always be done by/with someone who has musical experience. I think Rob Hubbard is an excellent example of this. He was originally a studio musician, and taught himself to program. His compositional qualities and good ear were the driving force behind his music routines.
People with a trained ear can often pick out details that a normal programmer can't. And if you're a composer, you will probably have more creative ideas about what you want to do with a computer musically. So you may try out tricks that a normal programmer would never think of.

br44
February 11th, 2017, 09:47 PM
you have to keep in mind I was born in 1993 so I experienced these computers long after they were outdated.

When I experienced the commodore 64 I might have been 3 or 4 years old. All I remember was a computer hooked up to a tv with boring text. The IBM on the other hand I actually got to play games on. My only other gaming experiences at the time consisted of the gameboy and the virtual boy. As you can imagine... compared to those two systems the 4 ugly colors of the IBM was hot stuff. The only other thing I can compare it to was a VERY early and vague memory of a monster truck game on a console of unknown make that didn't work 90% of the time and would freeze up if you so much as breathed on it.

dmemphis
February 12th, 2017, 08:49 PM
you have to keep in mind I was born in 1993 so I experienced these computers long after they were outdated.

When I experienced the commodore 64 I might have been 3 or 4 years old. All I remember was a computer hooked up to a tv with boring text. The IBM on the other hand I actually got to play games on. My only other gaming experiences at the time consisted of the gameboy and the virtual boy. As you can imagine... compared to those two systems the 4 ugly colors of the IBM was hot stuff. The only other thing I can compare it to was a VERY early and vague memory of a monster truck game on a console of unknown make that didn't work 90% of the time and would freeze up if you so much as breathed on it.

>... I experienced the commodore 64 ... All I remember was a computer hooked up to a tv with boring text

Oh boy you got gypped! That should have been the memoriable one... I hope you have one to put
through its paces now, if not get one!

Casey
February 12th, 2017, 10:32 PM
The S-100 bus for CP/M provided a popular & common bus during the CP/M era.

The QX-10 came with 256K memory, but I didn't know at the time that the Z80 could only address 64K. It sure was handy having a 192K ram drive, though! :)

Yes, having 64K as a base was very good, but the CP/M field had a much more mature software base, and a much larger selection. Enough stuff was ported over to the PC to make it workable.

The docs for CP/M were arcane at best.

Casey
February 12th, 2017, 10:35 PM
I never experienced that level of lag while typing, but then I naturally stop after a few lines to think. :)

The promise -I think- lay in the combination of "integrated software" (remember when that was all the rage?), an easy to use keyboard that included things like a STOP key, and a WYSIWYG word processor. It was something else to hit the BOLD key and see boldface on the screen, as well as italics & underline. It's a pity they never leveraged the graphics this machine was capable of.

Casey
February 12th, 2017, 10:38 PM
It included what was called a "non document" mode; that is text only. I used it to write source for my Fortran 80 compiler, but that's another story.

...At one point I actually owned the IBM Assembler & IBM Basic Compiler, but I ended up throwing them away because I stopped using them... Sigh... If I only knew what they'd be worth today...

Casey
February 12th, 2017, 10:51 PM
The C64 had sound & graphics capabilities that allowed their owners to laugh at IBM owners. :) The also had access to more powerful peripherals than did IBM owners, at least during the XT/AT era. I recall at one point my friend Tom (C64 owner) was nearly the only 2400bps caller to my BBS. The PC didn't get anything like the Commodore sound until the Sound Blaster came out, and that cost as much as a full-up C64. That's not to mention the graphics. It wasn't until the EGA/VGA era that PCs gained something equivalent to the original Commodore.

I was also flabbergasted when I read an article on the comparative efficiency of a Motorola 6502 and an Intel 8088. Note: was checking my memory via Wiki, and glad I did. My recollection was that the Commodore used a 6502, but apparently it used a 6510, a modified 6502. I expect the comparisons still hold.

In either case apparently the 6502 used more single-cycle instructions than did the 8088, so it was more efficient despite it's lower clock speed.

Scali
February 12th, 2017, 11:28 PM
That's not to mention the graphics. It wasn't until the EGA/VGA era that PCs gained something equivalent to the original Commodore.

In terms of colour and resolution, yes. But other than that, they were still quite primitive standards, and did not offer any hardware sprites, and only very limited hardware scrolling capabilities.
You needed a fast CPU to compensate for this with brute force. Eg, a game like Commander Keen needs at least a 286 to get perfect scrolling and animation at the maximum framerate. That's a whole lot more CPU power than the 1 MHz 6510 CPU that can do the same on a C64, with the help of the VIC-II chip and its capabilities.


My recollection was that the Commodore used a 6502, but apparently it used a 6510, a modified 6502. I expect the comparisons still hold.

Yes, like the Atari VCS uses a 6507. They are both slight variations of the 6502. The 6507 just has less address lines connected, allowing it to fit in a smaller package with less pins, a cost-saving measure. The 6510 has a special 'tri-state' selector on it, allowing other chips to share the address bus. It also had an 8-bit IO port rather than the 6-bit on a regular 6502.
In both cases, the changes to the original 6502 have no effect on the actual performance of the chip. The core is exactly the same as a 6502.


In either case apparently the 6502 used more single-cycle instructions than did the 8088, so it was more efficient despite it's lower clock speed.

Yes, you could see the 8088 as an early case of the 'MHz myth': it ran at nearly 5 times the clockspeed of an 6502. However, it also needed 4-5 times as many cycles for a lot of operations.
The main equalizer was the memory: 8088 systems and 6502 systems generally used the same type of memory, with the same access times. On a 6502, a byte access would take 1 cpu cycle. On an 8088 it takes 4 cpu cycles.
It also doesn't help that the 8088 is really an 8086: a 16-bit CPU, on an 8-bit bus. Because it is quite an advanced CPU, it has a rather complex instructionset, where many instructions are encoded with 2 or more bytes. If you hook it up to a 16-bit bus (as was originally intended with the 8086), this is not that big of an issue. In the case of the 8088, a lot of time is wasted just waiting for the instructions to come in from memory. The 6502 has a much smaller and simpler instructionset, resulting in more compact code.
When optimizing for 8088, the first thing you do is to find the shortest possible instructions for a piece of code. 99% of the time, the shortest routine is the fastest routine, because the execution speed is almost entirely dictated by the time it takes to get the instructions from memory.

alecv
February 13th, 2017, 01:28 AM
I saw my first 5150/5160 in USSR 1987 during USIA Exhibition: Information USA

Announce:
http://www.upi.com/Archives/1987/06/05/US-computer-exhibit-opens-in-Moscow/8040549864000/
http://www.csmonitor.com/1987/0623/awick.html

Photos
https://habrahabr.ru/post/249651/
(text in Russian, use google translator).

Video:
https://www.youtube.com/watch?v=JVwH47niko4
(in English)

I played IBM's 'CAT' game on CGA.

dmemphis
February 13th, 2017, 07:35 AM
The S-100 bus for CP/M provided a popular & common bus during the CP/M era.

The QX-10 came with 256K memory, but I didn't know at the time that the Z80 could only address 64K. It sure was handy having a 192K ram drive, though! :)

Yes, having 64K as a base was very good, but the CP/M field had a much more mature software base, and a much larger selection. Enough stuff was ported over to the PC to make it workable.

The docs for CP/M were arcane at best.

>The QX-10 came with 256K memory, but I didn't know at the time that the Z80 could only address 64K. It sure was handy having a 192K ram drive, though!

RAM drives were a godsend in the times before affordable hard drives.
Perhaps every platform of the day managed to make use of extra memory as RAM drives.
It was an excellent solution for more mass storage space and speed.
Software developers lived by them.
Even the PC/XT platform had RAM drive solutions, even in the real mode DOS days.
In fact I'm probably going to set one up on a AST memory board that goes past the
640K boundary.
Today we put in a solid state drive for much of the same reason... access without
the physical drive movements and enjoy similar benefits.

dmemphis
February 13th, 2017, 01:00 PM
I saw my first 5150/5160 in USSR 1987 during USIA Exhibition: Information USA

Announce:
http://www.upi.com/Archives/1987/06/05/US-computer-exhibit-opens-in-Moscow/8040549864000/
http://www.csmonitor.com/1987/0623/awick.html

Photos
https://habrahabr.ru/post/249651/
(text in Russian, use google translator).

Video:
https://www.youtube.com/watch?v=JVwH47niko4
(in English)

I played IBM's 'CAT' game on CGA.


Very cool historical video. I loved the folks remarks of their experience at the end.
Thanks!

Casey
February 16th, 2017, 11:52 AM
In terms of colour and resolution, yes. But other than that, they were still quite primitive standards, and did not offer any hardware sprites, and only very limited hardware scrolling capabilities.

Well, we were discussing the original PC/XT, no? :) No 286 there!

As for the 8086, I used to have an original Deskpro with that cpu. Would run at 4.77Mz for compatibility and around 9Mz in "turbo" mode. Quite the speedster. If memory serves the AT&T machine used that cpu as well.

Tandy actually used the 80186 in one of their systems. Go figure.

Yes, it was cheaper to use the 8088 to save money on the bus & peripherals. I expect today Intel would have called it an 8086sx. :cool:

To be honest I find the XT form preferable. I have both Compaq & Zenith desktops with proprietary power supplies and form factors. For example, an older Deskpro (8086 or 286) will happily take any standard 5-pin DIN keyboard, the connector is on the front of the motherboard, not the back.

I've sometimes thought about hacking the frame on either a Zenith or a Compaq desktop to take a "standard" ISA motherboard & power supply. The Zenith might be a better choice in that it also has a rear keyboard connector and the hard drive bays are easier to remove/install. The Deskpro system is just asinine. :mad: Can't afford to buy a new case these days, especially considering shipping...

vwestlife
February 16th, 2017, 02:32 PM
Tandy actually used the 80186 in one of their systems. Go figure.

To be fair, the Tandy 2000 and most other computers that used the 80186 were designed before the IBM PC became such a dominant force in the personal computer market that both hardware and software compatibility with it was deemed mandatory. They figured that just having software compatibility would be good enough, as it was with 8-bit CP/M systems, and that offering hardware improvements over the IBM PC would attract buyers even if it wasn't compatible with the PC. (Such as the Tandy 2000's faster CPU, 16-bit data bus, expandability to 768K of RAM, twice the floppy drive capacity, higher-resolution video, built-in I/O ports, a better keyboard layout, and expansion cards that could be installed and removed without needing to open the case.)

http://www.vintagecomputing.com/wp-content/images/retroscan/billgates_tandy2000_large.jpg

MrArgent
February 16th, 2017, 10:22 PM
I recently got my hands on a Amstrad PPC640 that, itself, is a contemporary of the XT/AT that uses a v30. Go figure.

dmemphis
February 17th, 2017, 01:00 PM
I recently got my hands on a Amstrad PPC640 that, itself, is a contemporary of the XT/AT that uses a v30. Go figure.

Nice those are really interesting industrial designs!

T-R-A
February 17th, 2017, 07:15 PM
FWIW...Oddly enough I noticed today that the place where I work stocks 80C188 chips for repair on some boards (Yeah, I know it's not a 80186, but close). 12MHz/SMT PLCC package made by AMD. Thought of this thread when I saw them....

Scali
February 18th, 2017, 02:19 AM
Tandy actually used the 80186 in one of their systems. Go figure.

Philips offered the very obscure ":YES" system in 1985: https://en.wikipedia.org/wiki/Philips_:YES
It has an 80186 CPU, and comes with an OEM version of MS-DOS.
I guess it was one of the last systems where the manufacturer bet on the wrong horse: they figured DOS-compatibility would be all they needed, and a faster CPU and better hardware around it would seal the deal.
Others had already figured out that a PC doesn't have to be good or fast, it just has to be cheap and compatible.

Something that we saw once again in the days of moving to 64-bit: the short-term poor solution of x86-64 extensions won out over moving to newer 64-bit architectures.
And with ARM we see history repeating itself: ARM keeps getting extended with new instructionsets and execution modes, to remain backward compatible, at all cost.