DOCUMENTARY: Why and How IBM ended up creating the PC (and ended up choosing the 8088 CPU)
Вставка
- Опубліковано 12 тра 2024
- #IBM released the IBM PC 5150 in 1981. Internally, IBM went through massive hurdles to get a personal computer to the market to compete with the other microcomputers of the time. This documentary shows all of the happenings organisationally as well as the trade-offs and decisions that led up to IBM choosing the sluggish 8088 CPU.
Don't forget to LIKE this video and please SUBSCRIBE to my channel! Press that NOTIFICATION bell so you get to know when new great videos come out that you'll like! I really do appreciate your subscriptions, it makes the countless hours I spend on this channel feel so worthwhile!
If you really like my stuff and fancy joining this channel to get access to perks, you'll be helping me develop the channel for the future with new equipment and spurn me on for more content:
/ @alsgeeklab
Donate on Patreon: / alsgeeklab
Or buy me a quick coffee on Ko-Fi: www.ko-fi.com/alsgeeklab
On the web & social media:
www.alsgeeklab.com
Twitter: / alsgeeklab
Facebook: / alsgeeklab
Instagram: / alsgeeklab
Some of the main reference material:
BYTE Magazine, September 1990: archive.org/stream/byte-magaz...
"Bill Gates, Microsoft and the IBM Personal Computer" - Infoworld Newsweekly, August 23rd 1982: books.google.co.nz/books?id=V...
"Microsoft Announces 8086 Basic" - The Intelligent Machines Journal, Summer 1979: books.google.co.nz/books?id=F...
IBM Exhibits: www.ibm.com/ibm/history/exhib...
TOC:
0:00 Opening Credits
0:23 Introduction
1:08 The Rise of IBM
3:01 The Microcomputer revolution
5:50 IBM play catchup
7:23 Off the shelf components
10:00 Choosing the CPU for the IBM PC
11:52 MOS 6502 CPU
12:54 Zilog Z80 CPU
13:53 Other contenders
14:50 16-bit & the need for memory capacity
15:07 Texas Instruments TMS-9900 CPU
16:09 Motorola 68000 CPU
17:37 Intel 8086 CPU
19:21 Intel 8088 CPU - The PC's CPU
20:31 CP/M operating system
21:21 Why the weaker 8088?
21:35 Reason 1: Cost
23:23 Reason 2: Availability
24:04 Reason 3: R&D time
25:56 Reuse of the Intel 8085 compatability
27:58 Reason 4: Outside influence
30:11 Reason 5: Software availability
31:33 The first PC (1981)
33:50 Conclusion
#retrocomputing #vintagecomputing #ibm - Наука та технологія
My grandfather was an engineer at IBM since the Apollo missions, and was, I believe, assigned to the PC project after working on the System/360 and System/370 mainframes. I'm not certain what elements of which he had a hand in the design, but I do know that he was likely bounced between engineering lead and project manager from the PC era through the AT (for which he helped design the power supply, among other things), and may have been brought back from retirement to help manage an engineering team for the PS/2.
I've always meant to talk to him more about his tenure at IBM, and I suppose I ought to make the time since he's nearing 90. Still sharp as a tack and still a computer lover.
I'd love to have him on the show as a guest of honour! Find out what he did on the PC and we can tee something up!
You should ask him about computer aesthetics and the next step of [gaming] computers having screens inside the case for aesthetic purposes.
Hell, I'm planning to mod my case to have an LCD screen I'd salvage from some monitor to then show a translucid video feed through the glass side panel.
(and yes, I do believe that RGB is silly... Unless a light theme is used. It's like the old TVs in furnitures trend. The RGB lights can be set to specific styles and colours for ambiance and decorative purposes.)
My Dad is approaching 90, please have a sit down and a good heart to heart with your grandad to ask more of him on this and other topics.
The knowledge the older generations have is priceless and should never be underestimated!
Regards
Please act on this, his insights and knowledge about the behind-the-scenes story would be very important for the historical record of IBM, once considered the top name in personal computers. Interesting to many potential viewers, I'm sure.
IBM was a big part of my professional life for a long time, at least till 2008 came along🙄, as I worked in commercial entities programming enterprise software for firstly the System/38, a baby of the 360, and then later the AS/400. Midrange! 💜
And yes, I actually used CP/M on a Telxon product running on a clone early on back in 1984, pulling in store data overnight thru a dialup ... at the exact same time as feeding punched SKU tags, sent in from smaller stores in bags, into a Kimball reader attached to a 3741 key to 8" disk machine ! "Ka-chunk, ka-chunk."
A n c i e n t.
Good doc!
I've been designing software/hardware since 1972. I came across the 5150 motherboard when it was a 4X layout drawing being digitized by the firm that did the digitizing of my circuit boards. I visited the digitizing station of said firm to check on the work being done on one of my memory board designs. The man doing the digitizing said "Come into the secure digitizing room and have a look at the work IBM handed us this week." I looked at the 4X layout, noticed the ISA bus section & the CPU & coprocessor sockets and said "It looks like IBM is getting into the PC business." IBM had told the board digitizing firm it was just a new terminal. There was a massive new IBM facility just two streets away from which he had his technical contact persons. This was in Markham ON Canada in late 1980 or early 1981, so it seems IBM wanted the board design project to be far away from Boca Raton to keep it secret at the time.
Wow! Amazing that you got to see that, given how secretive IBM were about everything!
@@AlsGeekLab The weak link in the 5150 security was the people IBM farmed out some aspects of the design process to.
Your American Accent when quoting Bill Gates is pure comedy gold. I love it. Also, congratulations on a well researched and nicely produced documentary. Great work!
I believe the 8088 would have equated to 3000 nm process not 20nm. Fabrication processes didn't get that small until the 2010s. Great video!
I believe it was 20 microns (um,), not nanometers(nm).
I worked at Intel at the time, and I have some personal knowledge about why they went with Intel vs. Moto chips that most people aren't even aware of. Also, the choice of 8088 vs. 8086 was very simple: COST. Intel had an entire suite of support chips for 8-bit data while chips for 16-bit data were still in development and quite expensive. That also meant that if IBM went with the 8086 they'd have to buy the support chips from someone else, and that would have complicated things tremendously, and delaying production beyond their deadline. The 8088 had the same internal architecture as the 8086, which is where they went with the AT as 16-bit support chips became available. So they went that route instead of the 8085 since all of the software and peripherals would continue to work with the 8086.
BTW, the 8085 was originally introduced with a bunch of instructions that mirrored what was added to the Z-80 over the 8080, but at the last minute Intel decided not to document them. Internally, I kept hearing people say, "Intel makes sand". I was a software engineer, and for most of the time I was there, software existed ONLY to help them "sell more sand". They found that while hardware engineers who wrote lots of Assembly code loved the new instructions added to the Z-80, modifying their P/LM compiler to incorporate them was a whole 'nuther can of worms that added zero ROI in terms of improved sales. So eventually the additional instructions were removed from the 8085 except for two, and adoption of newer 16-bit CPUs like the 8086 and 68k made the Z-80 obsolete.
BASIC was the first widely used language on PCs, thanks to Microsoft, although it was an interpreter, not a complier. First real compilers for PCs were FORTRAN and P/LM, which I believe were both developed at the same place Gary Kildall taught (maybe by Gary himself), the Naval Postgraduate School in Pacific Grove, CA, where Gary's Intergalactic Digital Research Corp was also based. Internally, Intel licensed a version of CP/M from Gary that they called ISIS, as well as his P/LM compiler. They were used internally for years. Getting back to the earlier point about the extended instructions on the 8085, the hardware guys loved them, but when Executive Staff heard the budget that the compiler team proposed to add compiler support to P/LM for them, they were stopped dead in their tracks.
The Intel Development Systems, a.k.a. "blue boxes", sold for $30k and the software was free. Intel created these boxes to help customers "design-in" Intel's chips faster than chips from other vendors, reducing their time-to-market significantly, and support for those extra instructions weren't forecast to produce even a single additional sale, especially as the newer x86 architecture was rolled out that replaced the 8080 architecture (including 8085). So the 8088 was designed to provide an 8-bit replacement leading to the newer x86 architecture that used all of the existing 8-bit peripheral chips. If you wrote code for your 8080/85 boards in P/LM, then all you had to do was recompile it to run on the 8088 using the same peripheral chips. And replacing an 8080/85 with an 8088 was designed to be as simple as possible for Intel's customers. Our blue boxes did not support Z-80's, and as expensive as they were, they cut time-to-market in half, which was highly motivating to most customers. Also, the 8088 was much faster than the 8085 and was deliberately priced about the same, and at the end of the day, nobody ever cared about those extra instructions.
The strangest part about all of this was that IBM never expected to sell more than 10,000 PCs. They seriously believed they were just "toys" and would be replaced by much more capable low-end "mini-computers" IBM was designing within a couple of years. (They were seen as a stop-gap to get IBM into the game, but not a long-term play.)
The basic design of the PC came directly from the sample design in an Intel Databook. All IBM did was modify it to a bus based motherboard. It also could use some 8 bit cards IBM had for a 8085 based word processor. Namely the monochrome video card and the parallel printer card.
@@michaelterrell That wouldn't surprise me, as Time-to-Market was their #1 priority from what I'd heard. I suspect that's how Intel ended up becoming the world's largest motherboard maker. When they were coming out with a new CPU, an Intel FAE would make a design and publish it as an Application Note. But they'd be working on the design for months before the chips were released, and in close concert with the chip design team. A few customers would get "inside info" but when the chips were released, that App Note would also be released and companies usually started with that. Customers wanted more and more of the Mobo details, and Intel started giving out the gerber files and other things they'd created until it became clear that they could start building mobos that could also be shipped at the same time the chips started shipping. This also allowed them to build-in a lot of their other support chips, ensuring their chips were more likely to be used rather than competitors' chips. I guess that after a while companies just started saying, "Why don't you just build us complete boards? But with these changes..."
@@TheSchwartzIsWithYouToday Adapting the existing circuit probably save six months or more. Ive seen way too many designs over the decades that were nothing more that blocks of circuitry from databook examples, with little or no effort at optimization. One really ticked me off. Collins/Rockwell deigned a C-band receiver in the late '70s that used a large electrolytic in the video circuitry to couple two incompatible sub-circuits. They would run hot, and dry out over a few years, and damage the circuit board. A simple level converter would have eliminated that problem, and improved the video response.
It looked like they had just taken the design of a military product, and changed a few parts to put on the needed range of frequencies.
@@nicksterj yeah, I don't know what I was thinking there. I deleted it. IBM wanted to cap their costs. Unfortunately for them, they sold millions at a fixed price that MS delivered by the pallet, ready to be dropped into the boxes before shipping them off. MS was also able to sell the same thing as MS-DOS as a retail package that all of the cloned PCs would then use.
Heads up: the singular of Hertz is Hertz. You can’t singularity it as “Hert” 👍
Correct as it was the name of a person, Heinrich Hertz
I’m old enough to remember when the idea of a personal computer was met with, “What do I need a computer in my home for?”
I bet you have a lot of great insight then!
@@AlsGeekLab I am proud to be at the beginning of the home computer revolution. It was British companies chiefly Sinclair who pushed for the sub £100 home PC.
Tell me Al, have you seen Micromen the docudrama?
@@Zooumberg yep, seen it. Was a great series. I owned a zx spectrum and I still own a BBC Micro
I also met (typically older) people like that in the late 90s and early 2000s.
@@GlorytoTheMany my first TRS80 (in ‘77 or ‘78) had 4k rom, used cassette tapes for programs, and had graphics on par with an etch-a-sketch.
In 1968 at technical college here in the UK we were shown the first mobile computer - it only took 3 technicians to push it in to the lecture theatre - that computer was made by IBM, the floppy was a yard in diameter and had 1 Kb, the screen was very small with a huge box under it and the keyboard was above the hard drive and I remember it was less than 1 Mb.
As an IBMer I remember the launch of the PC. However, I wasn't involved in it. I was on the mainframe side and not aware of the PC until it was announced. This video, for me was GREAT! Thanks, BTW, at times the music was much too much, too loud.
Thank you - an awsome historical account of the genesis of the IBM PC. A lot of research went into the production of this video. In 1985 I took a college course on 8088 assembly language, and as you mentioned, working with the 8088's segmented registers was quite awkward, along with the 'crippled' 8-bit data bus. Having a 'flat' memory addressing architecture would have simplified programming and improved hardware performance. But as the video explains, decisions made at that time dictated the selection of the 8088 cpu. These limitations were present until the arrival of the 80386, also mentioned in the video.
A final struggle with the college course was that the PC clones in our computer lab used an NEC-8088 version which had a built-in hardware bug, which couldn't execute certain combinations of machine instructions. My instructor didn't fathom this possibility. One day I sneaked into an adjacent lab, which had newer/different PCs, and my text editor code ran flawlessly. In the PC industry, surprises always appear, providing endless awe and wonder.
A great presentation, kudos!
Thank you very much. Yep, the video took quite a significant amount of time to research and put together from what I remember (I made it back in 2019), but yeah thanks for noticing the hard work that went into it! Check out my new video on XENIX. A similar amount of effort went into that one!
To me the smartest mistake IBM did was open architecture and not buying Dos, Commodore, Atari, and all the others were closed architecture, Apple was open but Steve Jobs took Apple to closed. Innovation went to the open architecture and hardware thrived there. Originally with other brands of PCs you had to buy a different version of dos and software for that machine, Microsoft put it's foot down and said one version for all, that opened the door fo a software explosion. People forget that you had to buy Zenith OS and Zenith brand word-processing software. Before Microsoft said no more.
Someone built a card for the Heath/Zenith computers sold to the US military to all you to run regular MSDOS software. I think that I still have a couple if them, in storage.
The Commodore VIC20 and C64 had an expansion slot to use either a game cartridge, or third party module.
Well, I watched it on an Apple Mac and owned an Apple II back when. The Apple Corps of Dallas had its peak membership the month the PC came out and dropped as everyone was pushed to the PC at work."No one was fired for buying IBM"
I did work as a free lance on the earlier IBM 5120/5120 desktops which cost $32000 then with 32K of memorry 8" 1 meg floppies and no expansion. Thank you Dallas Museum of Art for building my career!
I worked for a Fortune 500 company's Austin location in the '80's and 90's. PC's, Ltd was my customer, later known as Dell Computer. Michael Dell had it figured out. His "showroom" was nothing more than a pegboard wall with components hanging like beauty products at Walgreen's. Everything was compatible. One could look down a short hallway and there was Michael sitting at his desk, usually on the phone. I continued calling on Dell until the crash of 2002, then called it quits. Wonderful time.
I remember buying one in the eighties. I learned to write some BASIC code by reading magazines and books, but knew little about computers. I had a young lad working for me part time, and who was doing IT in high school. He advised me to "buy one of these Pete, it's got a twenty megabyte hard drive, and it'll store all the data you'll ever need". I often pull his leg about that statement.
Well if everything was at a constant, 20mb was a lot back in 1984. But things change after all, who knew!
@@AlsGeekLab My first Hard Drives; Seagate 20Mb, Miniscribe 41Mb, Fujitsu 90Mb, Seagate 131Mb, WD1Gb, Fujitsu 2.6Gb, Seagate 4.3Gb, Quantum 20 Gb, WD80Gb, Seagate 250Gb & Seagate500Gb all IDE & so on from there. & There we have the evolution of the Hard Drive. I still have them all.
In essence, he's right. Today's data is more of graphics and videos, which take million times more space than simple text.
"I said 'need, not 'want.'"
The Miniscribe is a 50 pin Drive I'm going to have a look at what's on it today. I just got the Cable for it.
Interesting to see all those pictures of the IBM 5100. I used a 5110 a slight upgrade in 1978. It was an odd beast but very reliable and a safe business choice when the alternative was a PET with cassette tapes. I would love to see a video covering this family of pre PCs.
Layperson here. For me, a thorough history well written with many of the key players and their respective technologies made known to me for the first time. Well done and thank you.
No, Thank you for taking the time to pass on your kind words! Makes my day!
While I was born around the time that the first pc's came out, I still remember them fondly because our first computer eventually moved towards my childhood bedroom and I spent many hours exploring programs, games and programming on it. This documentary was a very nice watch, and I hope you'll keep up making them while I subscribe to your channel!
And you became a GEEK!
Great video. However, he was called Don Estridge, not Phil. Don was his middle name. He unfortunately died in the Delta crash in Dallas. It was due to wind shear, which was little known at the time. Another IBM employee of mine also died in the crash. They were all on the way from Boca Raton to LA.
Also, when we launched The 5550 personal computer in Japan we used the Intel 8086 as it supported 16-bit bus access which was needed to use the double-byte characters used by the Kanji language.
Thanks for your insights. Sounds like you have some great insider knowledge. I'd be happy to have a call and talk more with you on the subject, perhaps we could make a video?!
@Al's Geek Lab I would be glad to have a call. How can we set that up? I spent 25 years at IBM much of that in Japan.
@@danielsullivan5130 I am in New Zealand so we are only 4 hrs different. I could possibly do a call today if you are available. My email is in my about section here on yt, or send me a DM on twitter.
This many decades later a unified industr-wide design was inevitable, but nothing beats the excitement we lived during the 80s computing scene. From small companies to large ones, from hobbyist developers to huge teams, from weird hardware to absurd software -- anything really was possible and tried.
It sure was an exciting time!
@@AlsGeekLab To be savoured every month in "BYTE" magazine ;-)
The 1980s was the 'Cambrian Explosion' of the computing world
Wargames (Mathew Broderick) in the theater was my introduction to computers, and I begged for a computer every birthday and Christmas until mom saw a Word Perfect demonstration in 1985. Our first family computer was a Tandy 1200, configured as an IBM PC XT with an 8088, 640mb ram, 2× 5.25" floppy drives on the left, and a 20mb hard drive taking up the 2× 5.25" drive bays on the right, green monochrome screen, and daisy wheel printer. This was mom's word processor, and graphics cards showed the interlaced lines on each character, and gave her a headache. So graphics got returned to Radio Shack. Thank God for lawn mowing money, and the Commodore 64.
The success of the IBM PC was largely based on the open architecture and the unforeseen fact that there were millions of geeks lurking out there who just wanted to fiddle with the thing.
The open architecture ( which was likely because the PC was done in a skunk works fashion, the corporate lawyers and marketing ppl didn't know about it; IBM was anything but open. The fact they let MS have rights to windows prob led to major teeth grinding at Armonk,prob bc they never envisioned anyone but blue suits to have one).
The success wasn't the geeks. The IBM PC was too expensive, I had one circa 1984 w dual floppies, and a 13" color monitor, it was 5k ( be about 15k in today's money).
Its initial success was corporate,by having the IBM name on it it legitimized the PC. Before that with the CPM based machines and things like the Apple II it was the domain of Greeks and home users mostly ( some businesses,bc of VisiCalc in the apple II series, started using PCs). When the first clones came out, they weren't cheap either,but it was the clones that moved them into the geek market and the like.
With the PC companies didn't need things like dedicated word processors, and the person who once was called a secretary could do a lot of things on one machine. Likewise when they worked out networking ,( which in those days was expensive as heck, was like 1500/workstation) , allowed shared file servers and the like.
@@njlauren I also had an original 5150 around 1983 or 84. It was a hand-me-down from my brother and yes he paid about $4500 for it in 1981. Prior to that I had an Apple II which at the time was still somewhat "open architecture" as I was running CPM on it with an aftermarket card. I think that both Apple and IBM were trying to get the cat back in the bag and Apple had more success in that regard but for most of us the PC clone market was the ticket.
@@pi.actual
You are pretty much right there. IBM did attempt to put the cat in the bag again, on the PS2 series they came up with a proprietary bus structure where boards if I recall correctly had to get a license from IBM to operate in there. I think the pin structure was the.same but their bus mastering firmware made it closed unless licensed. It failed miserably,the most wacked out version was supposed to be able to run system 370 code. On the other hand IBM used to declock theyr cpu's and had firmware if you tried to hack the clock speed,would not allow PC to run. This was done to protect their mini computer line.
Apple with the Macintosh did in fact close it up. Everything in it was closed, you had to buy their stuff, and it was always lot more expensive. Ppl built Mac clones,esp when they shifted to Intel cous, but the problem was the OS, you had to hack that to get it to run with peripherals running on the box. Apple was so bad that way when they implemented networking it was a proprietary networking called Apple Talk that was incompatible with Ethernet networking . At the time someone I know went nuts trying to use macs and network them to other systems/devices.
Yes - the open architecture was a huge advantage and reason for the success of the IBM PC. In fact, IBM was counting on IBM PC customers to eventually purchase it's own add-on cards to boost its sales and profits. However, most customers simply bought the bare-bones system, then filled in the expansion slots with cheaper, better ISA-cards from third party vendors such as AST Research and a plethora of related companies. By the time IBM created its proprietary MCA bus to lock-in its customers, the PC-clone vendors simply created their own industry standard bus, I believe called the VESA bus, but memory evades me. In summary, the IBM PC created the common platform that helped propel the PC ecosystem forward.
@@michaelmeichtry316
The industry standard was the EISA bus, I think IBM was the VESA bus. Theirs was proprietary, and few board or device makers were willing to pay the licensing fees. Thing is the VESA bus was not exactly earthshaking, it had advanced bus mastering, it was fast, but the EISA standard did basically the same thing. The other thing IBM did ( which they tried to hide), they had code in the BIOS and in their bus that would keep speed well below what either cpu or bus could handle. The reason was simple, they were afraid of taking business away from their mini computer line. Interestingly IBM offered a compressor that would let you run system 36/38 apps on a IBM PC w the VESA bus.
This crap extended into their unix servers, they had the segmented crap there too,where even though they were unix, if you moved from a lower model to a higher tiered one you often had to buy new software, there were incompatibilities or you had to rewrite your own code to move up.
We purchased a Commodore 64 for our 8 year old fairly smart son when they first came out. Over halve of the little 64K of memory was used to run it. Had to insert a floppy disk then type in something then wait awhile then had to remove UT and turn it over and load that side. Think it took 12 to 15 minutes to load a very limited program to print out a 4 line hone made greeting card. Printer had a 4 color tape that printed less then 20 pages before it had to be replaced. Can remember going into offices back in the 1970 & 80's and seeing huge pile of IBM that we called punch cards in boxes all around.
I watched your documentary because I am a fan of microcomputer technology, as well as an end-user. I enjoy stories with people using the microcomputer. My own microcomputer, which is a Dell brand, is still my powerful typewriter. My first microcomputer was a Cybernet which model was an all-in-one PC, that I purchased in 2002. Thank you very much for showing!💙
I learned to program a computer before I ever touched one. Was stuck on a farm one brutal winter and was so bored I read books and built a VM in my brain that I would program 🤠💪👍
I am old enough to have been a young programmer in the late 70s. I did most of my work in assembly language due to the slow speed of computer systems. I looked at 6502 based systems and Z80 and found the 6502 sorely lacking and chose to support Z80 based systems. Later in the late 80s I was programming in 8086 asm and Pascal but would be daily frustrated with that stupid segmented memory architecture and this did not let up until sometime in the 90s.
I have an interview with a man who worked in Lotus on 123 coming up and he talks about that very problem. Quite interesting to hear how people had to work around the issues back in the day
I used to write code in 8086 assembly language and I have to agree that the segment:offset addressing was sometimes annoying.
Thanks for the "reminder".
I started as an operator on the IBM 360 mainframe, switched to the 370, got into mainframe programming using the IBM 3270 PC GX which emulated up to four terminal sessions connected to the mainframe. This PC had 640 k of RAM and two floppy drives -- when they gave me a 10 megabyte hard drive on an expansion card, I thought I would never have any disk space problems again.
They switched me to the IBM PS/2 running IBM's OS/2 -- which was the last efficient and reliable PC operating system I have ever used. When OS/2 was a joint effort between Microsoft and IBM, it had a few bugs, but when Microsoft dropped out and IBM took over, the next upgrade gave me a machine that ran much faster and NEVER needed a shutdown/restart, regardless of how badly I abused it. The longest time I ran it without a reboot was about 9 months. The only time a restart was needed was when a new expansion card had to be plugged in or we had a power failure.
OS/2 came with built-in Windows support which we used to run the e-mail client, word-processor, and spreadsheet. When we had one of the frequent Windows task crashes, OS/2 would clean up the mess without needing a restart. On each of our (about 100) PCs, we set up a full mainframe simulator for program development. This included several language compilers and both hierarchal and relational databases.
Then along came Microsoft Windows NT (or should I call it "Windoze Not Today") -- the company gave me a new machine with double the RAM and double the processor speed from my PS/2 -- and everything ran slower than half the speed of the OS/2 machine -- and I was lucky if this NT machine survived an entire week without a crash or hangup. And, since then, every new release of Windows has needed more RAM and a faster processor ("bloat-ware") -- and, in keeping with Microsoft's tradition, we still need to shutdown/restart a few times a week.
Oh for the "good old days" when computers were reliable.
Whilst MS windows (especially 10 and 11) are huge orders of magnitude more reliable than any other windows before it, I still totally agree with your sentiment. OS/2 would have been shipped on a few floppy disks compared with the Multiple gigabytes you need just for Windows to operate.
I have a 6MHz 286 AT which I'm often curious about installing an earlier version of OS/2 on it but never picked up the courage. I'm going to do a video on OS/2 through the ages at some point. However, my next video is on XENIX, Microsoft's UNIX variant!
@@AlsGeekLab do you have enough free disk space to set up another partition? If so, you could set up a dual-boot.
@@rbpalmer1 I guess I'd go down the route of emulation first then try a dual boot on real tin. But yes, my old PC is kitted out with compact flash so it has plenty of space
@Ron - we were also married to IBM at work - even down to Token Ring and DB2/2 .
I had to go to Taipei on a work trip and discovered that they'd diverged from policy by using Ethernet - a dead end standard, obviously.
I share your love for OS/2, but we had to ditch it because the non-IT part of the organisation found it too "hard" to use.
But I must disagree about efficient PC OSes - *nix variants now are utterly amazing. You can even run REXX :-)
As a manager, when you want the freedom to innovate and transform, you NEVER follow established R&D channels. The only option is to find a few engineers, that believe in the idea as much as you do, go hide in a Skunkworks barn and beg for forgiveness later.
What cracks me up, EVERY time, is when you show off your new thing and everyone loves it. The next comment from upper management is “why didn’t you tell us about this months ago”? I did, you said it was a crap idea, at which point you solidified me knowing it was brilliant.
Soooo true!!!!
I had to laugh when it was said it "only" addressed 1 mb RAM. That was an insane amount at the time. So was the DOS 2 limit of 32 mb hard drive partitions.
I built an IBM XT from boards that were tossed out at a local computer store. Kt took me a while to accomplish this task, everything was scrounged and salvaged. Then the clone world came into being and building XT type Pc’s was more readily available.
Like that Johnny Cash song - "I built it one piece at a time. "
The father of one of my friends was the service manager of the Leeds branch of IBM and I showed him a copy of 'Electronics Weekly'( I think it was) that reported that IBM was developing a personal computer. His reaction was to thrust the newspaper back to me saying disparagingly , "Kitchen table brigade". Ho hum........
I'm typing on a keyboard that came with a PS2, a model M, but running Linux.
I remember PCs being introduced at work and the only things that made them acceptable to our IT department were that they ran mainframe terminal emulation, and they had "IBM" on the front.
We have effectively returned to mainframes with cloud storage, and Microsoft, Apple, Google etc in control.
My wife and I were working at State Farm Mutual when IBM gave State Farm five IBM PCs (16 k RAM installed on the motherboard). Of the five PCs, the CEO, Ed Rust, my wife, and I each received one. As IBM produced new IBM PC products, we each received upgrades, then State Farm started buying IBM PCs and the 3270 - PC. State Farm employees were given discounts on PCs for home use. I bought my first IBM PC at a cost of over three thousand dollars.
In fact, as wozniak recognizes in his biography, what caused the explosion of microcomputers was the launch of visicalc, the first spreadsheet. This transformed the "videogame" in a empresarial tool, explodind the sales, and waked up IBM. I worked for IBM at that time. There are videos of dan bricklin, the visicalc inventor, telling the history.
Yes! Being a bit of a Luddite, I resisted personal computers (Apple at the time) tooth and tong. Then I found myself in a job where I had no choice but to learn to use the horrible things. I went to the Apple store on Friday and ordered a system for delivery on Monday. The salesman let me take home the manuals for VisiCalc and Apple BASIC. By the time I got done reading the VisiCalc manual I was hooked and couldn't wait for Monday morning. It's been an Apple / Mac / PC (clone) love affair ever since.
A breath of 80s... ahhh, those were the times... thank you so much for telling this story, I can't express enough how much I've enjoyed!
The first time I saw an Apple 2 in a store, it was $3,600. I came very close to financing one for purchase. I have always wondered why a PC featuring an 8086 was almost unheard of. Thanks for answering that question. With the new Apple ARM chips coming out, it is possible that the long rein of the '86 based architecture may be coming to an end. Only time will tell.
Yes, I think the x86 architecture is in its last days
The 8088 choice by IBM led to the absolutely worst year of my life. In 1985 I was asked to manage the development of Release 3 of Lotus 1-2-3 and the goal was to fit the code for a multi-sheet spreadsheet with other new features, backwards compatible with Release 2.2, and written in C (not 8088 Assembly Language) into a standard 640K original PC. No way would the code fit, so we attempted to partition the software and use Plink86 as an overlay manager. Despite the efforts of 50+ programmers (and 75 QA people) working 80-90 hour weeks for a year, we were unable to pull it off and Release 3 eventually would only ship on 286-based machines.
Wow that sounds extremely stressful but also super. interesting! Would you be willing to do an interview on it?
@@AlsGeekLab Sure. Let me know how.
BTW I mis-typed the date above - it was 1987, not 1985.
My email address is in the about section on this channel, drop me an email and we can talk about timing, or you can get me on discord discord.gg/mGtHmTaU (although I'm not on discord that often, so could be a little slow replying!)
But weren't the 286 memory addressing issues just as bad? IIRC, the memory bank switching had to be done by a kludge using the KEYBOARD electronics! It still did not have properly protected memory.
You documentary was a surprise. Great videos, research, etc. Congratulations.
Many thanks for your kind words!
If IBM had used, say a, 68008 (unfortunately not developed at the time of the PC design), and then gone to Digital Research, with CPM/68k, the future version of the PC would have been the equivalent to the Atari ST, with GEM (Digital Research's Graphical Environment Manager). They would have had a 'Windows' type of operating system, and definitely would have scooped the Apple world, and the Macintosh might not have been invented. Too bad they didn't go with the MC68000 cpu, we would have been spared a lot of problems with programming around 64k and 640k limits!
I think the world would have likely been a better place if it had!
great documentary, enjoyed it a lot. to think we are all using machines with the x86 architecture because 40+ years ago one company decided to use one particular chip over many others feels quite bizarre
Glad you enjoyed the video! It does indeed! The days of the x86 architecture are slowly coming to an end, but not for a while yet
I bought the TRS-80 when it came out in '77, and eventually upgraded to the PC world with the Tandy 1000, and then home-built PCs. I suspect a lot of TRS-80 owners ended up going that same route to a PC, rather than Apple or something else. My friends with Apple ][s all got Macintosh, but even most of them are now with a wintel product. Nice documentary!
There was nothing wrong with the TRS-80. It was as good, or better, than the Apple. The reason the Apple took off as a 'serious' computer, as a opposed to a hobbyist machine, was due to the first 'killer app' being written for the Apple.
@@MrSloika back in the day I remember almost all the old machines I think I've used at least a few times most of the early Apple computers even up to Apple 2gs which was actually quite a nice little machine if I remember correctly it almost looks very similar to some of the mcintosh's.
One of the reasons why I remember using the Apple 2 GS when I was in school it was determined that I needed to have a computer at home because of my issues basically having to have dictation taken for almost all school work we were looking into getting computer but at that time it was too expensive so we wound up using a computer to the school system for my ears basically they got a blank check to get the machine.
I remember originally they were looking into getting the Macintosh of that generation however problem was compatibility issue with the computers at school since at least at that point there was no compatibility between the older Apple computers so just Apple 2C and the Macintosh although I think there's a product later on dad brought that capability I can't remember if which machine had the processor the right slot that would take a basically Apple 2 on a card.
Later on I quoted book graduated to using Dragon dictate on a modern machine at least modern for the day probably at that time was with the fastest computer that I ever used basically optimize running as fast as he could bus accelerator and more to get the most performance possible for Dragon.
I do remember it required double the RAM for even the normal Windows install of the day!
One of the reasons for this probably the main reason if I remember correctly is essentially it had essentially having the parts of the program that don't was voice recognition pulling data from Ram since much faster than hard drive back in the day essentially using dad to accelerate the process since Ram was so fast compared to other storage media!
And this was the original version of dragon.
Back in the day it required only quarter second pause between words for dictation!
The only thing is you have to switch between things in order to use different commands and other things in theory you could even throw to mouse and keyboard out the window and run Windows entirely from voice I do remember there's a few games that could actually be played using only voice commands even got pretty good at playing Tetris just by voice of course when people heard how I was doing stuff on the computer everyone always wanted to see it working introductory believe that I was actually typing by voice it was beyond comprehension to some people some people like wow maybe the future is now then again I needed this technology back then I had even trained other people to use the software as well that first letterhead used both windows version and dos version I did a demonstration once add an accessibility conference I was the only one that could actually do this effectively since I was using it readily everyday and knew both windows and doors version well enough to be able to do this
Also there are a few things I remember back in the day if anyone remembers the old PC Junior.
I remember sometimes running software on there and it was meant for the IBM PC and after a certain point you'd actually get Drive overflow not sure exactly what would happen there because it's just at that time I really was not big into technology nowadays and just really never thought about it again strange home that comes up so much that you don't even think about what happened back in the day with things.
Although the PC Junior probably would be closer to that of the Atari in terms of things like gaming but yet I have a feeling some of the Atari systems were actually more capable in their own way which is kind of cool but even something that was originally thought of as just games such as the Atari 2600 became something more of a PC even in its own way
By the way I was the kid in school that would always be servicing equipment I was the one that would always be loading the paper in the printers and so on.
Also in school I did have a lot of downtime I was kind of ahead of the game you could say!
I had all kinds of access to things that most students didn't even know they are there!
It's kind of neat when you can almost like make your own extracurricular activities including learning opportunities.
This would be happening often times throughout years of my schooling even in the early years.
And yes I got credit for doing so this as well someone is classified technical education I was even doing a bit of recreational engine stuff before even high school.
A lot of stuff was trade specific that wasn't even offered through our school!
Sometimes I was actually server this evening very equipment that I was using to do all this as well.
Some of these pieces of equipment I still don't even know what the media even was that the equipment uesd.
There was one other one that used two different cartridges can't remember what type of film it was but it was on what look like 8 mm film or perhaps 16 some of the cartridges almost looks like an 8-track but oddball much larger pretty sure the audio portion control portion on the same cartridge and tape of course was an endless with cartridge probably the same with the film portion but not sure on this portion.
I do remember often using what they called the teaching projector but this is a different system entirely use a single cartridge for both audio and film.
This is what I would call Antiquated early multimedia!
Also very early in school they have perhaps kindergarten maybe later.
There was a system it was pretty large and heavy the audio was on an LP photograph record the slides if you'd calm that although War closely akin to a film strip but in a plastic slide in frame sort of thing think about your slides next to each other versus a film strip.
Essentially a beefed-up and more utilitarian version of that system from one of the companies that had the stories on film and the audio want to record but intended for Education applications.
One thing I remembered is you always held onto the film portion and somewhat with the record if you open the door to the turntable that Portugal to fill would be quite forcefully ejected oftentimes leaving the top of the machine don't think that was intended didn't work that way only knew about one of these thing you was so not sure if it was a flaw or a feature but if it was part of the system it shouldn't have been that could hurt someone especially somewhere to eat that youngin school that may not expect it
Also by the way I always remembered replacing lamps in the projectors as well a lot of times before the instructors had trouble with something they'd send somebody to the winter class on me in that the time and say something hey need your help don't take a minute.
And I remember quite often in school being called to the office at first people thought I was getting you in trouble not in the least.
It was usually getting somebody working for someone or other issues and more often than not it was getting the PA system working again we had all sorts of trouble.
Hard to believe I actually had a tool kit in my locker and later on the school actually help me carry a multi-tool!
Sadly one of the very reasons to have this was to have a blade if necessary!
They said it would be a good deterrent if ever necessary never had to go for it once in this respect but then again but that time had been issued pepper spray!
I was underage but I had to get a permit it was that bad that times it seems like a lot of times the most intelligent people in school or the ones that are picked on and demolished it off and but sometimes well that does take a turnaround.
At one point I hope will the school bullies and everyone says you were crazy to do that I said nope he was in a bad situation stranded and his vehicle.
Ever since that point didn't have nearly as many problems Tina most of the others as well as his cronies him and his cronies took care of the problem I remember that time that's why I had locked me in a locker well the other person let's just say he was not-so- lucky he got blocking the Janitor's closet overnight don't know how that happened but the principal said do you know what happened to that guy I said no you don't mean the whole story only one problem with being the smart guy at school is well other people try to pledge rise from assignments although problem is if you do it from a disk or otherwise that in itself yeah big problem.
By the way I did have to back up to the server for my machine used at school.
Back in the day I was using it on-screen keyboard at using the mouse the type that first before the other came along I needed.
I was the only one in school even running Windows everyone else was using dos!
One of the problems is we could not get any of the printers to play nice with Windows so I actually wind up having a friend in the back room which is just awkward!
Later on I wish the iomega Zip drive had existed had to use external hard drive since essentially flash drives did not exist back then But there again Not only was that use for data for removable hard drive but backing everything up keeping everything in sync had to use that system because of this basically all my machines had to have the same date on them since because of the voice recognition system later on was a real hassle originally I'd have to have an entire binder of floppy disks prior to this.
I don't even remember how largest hard drive was for the removable hard drive as I mentioned with plagiarizing before there was at least two people that got caught in the active actually copying my data disc no way you would not have somebody copying your disk in class it was too obvious what they were doing seldom would anyone copy anyting unless backing up something that was very important few and far between and even at home having two floppy drives with an asset for backing things up I did have usually two copies of everything still remember using the old Daisy wheel printer that I had access to it one place primarily for anyting of Greater importance or discounts better and could get a better grade we're just having something printed on Daisy wheel versus dot matrix and yes that was a thing back then it was considered typed work as it on a typewriter instead of a computer printout which was goes to show back when this was being done.
I even remembered if you remove the tractor feed edges of the paper you could sometimes even get a better grade just because of doing this weird I'd say nowadays just using plain paper seems kind of weird that that was a thing later on
YEP, You're 100% correct. "Hard core" TRS-80 (and other Z-80 business class computers) users that used CP/M (Like me) were NEVER going to Apple. The IBM PC (and clones) were natural choice. MS-DOS acted close enough to CP/M that it was not traveling to a different planet like an Apple ][. P.S. I worked for Radio Shack in the 1970s, So my choice of the TRS-80 was no accident (I had a Model I and later a Model III). Without bias I will say that the Tandy 1000 was the best bang for the buck PC of the mid 1980s. I'm not the only one that thinks so: ua-cam.com/video/mYHtojsaRkY/v-deo.html&ab_channel=The8-BitGuy
I’ve actually used one of those things before. 8088 PC by God! They had them at a state government facility I went to where disabled people went to learn how to navigate college before they actually started. It was in the mid-1990s. So, it was already an old machine then. But, it was really interesting to use. In those days, the top of the line was a 486 with an actual harddrive in it which that 8088 did not have. They had a couple 286 machines there that boasted 10 and 20 MB hard drives. Hard to believe nowadays. We measure those things in terabytes now.
IBM actually released its first personal or portable computer in 1975: the IBM 5100. It was followed by 5110 in 1978, 5120 in 1980, Datamaster in 1981 and just a month later model 5150 which we all know. The problem with those first ones was the price. Also they were not micro computers in the sense that their CPU was not a single microprocessor.
You can actually see IBM 5100 at 4:30.
I used a 5110 in 1978, it revolutionised the way we worked. We could do a crude (by excel standards) income tax type data model for 500 churches with 30 parameters in an hour per re calculation . A task that took 3 days by hand. i think the cast was about £10k with two 8" 1.2 Mb floppies and a dot matrix printer.
Still got it done though! That's the great thing, some times I wonder if we have advanced too much. A lot of the things we have today were supposed to make our lives easier, but have they truly made it better or have they shifted the problems, or worse, made life less pleasant?
Excellent video with a tremendous amount of detail. Thanks for making this
thank you so much!
Never get tired of hearing this story.
31:02 The original list price of the IBM PC was $1565 in the "bare bones" configuration (16K RAM, BASIC in ROM, monochrome composite video card, no monitor, no storage drive). I'm sure this was done just to create a "low ball" competitive advertised price point, but in practice never sold that way. The ad copy says "with the addition of one small device" it can be hooked to your TV for a display and a cassette recorder for storage. That would imply a cassette interface cable and an RF modulator was available to make it work. However, various sources have failed to find the IBM branded OEM parts for this. Likely that IBM virtually intended to never sell a PC for a TV and cassette (just the thought of using it is laughable). A more typical starting config including floppy drive and MDA monitor would be sold at nearly twice the mythical "less than $1600" price point.
Money, there were 100s of companies dabbling with making computers in the 70s and early 80s. the lists compiled by various magazines and books (US and UK) are astonishing.
The S-100 esp Cromemco's mil contracts would have been a massive green flag for IBM.
People were making 8 bit applications at the time. Having to code for a 16 bit machine would make you the odd one out of the pack for software development. Third parties being able to code for your machines was important.
From early on before microprocessors IBM had been working the industry for central computing, even from home with terminals. The home computer market put that plan they worked on for decades in trouble. But Internet, and most notably the Web Browser had put them back on track. And while we all have personal computers in some form, we also now heavily use a central computing model when we use the web. The blending of home computing and central computing gave us more than IBM ever thought of back in the 60s.
I've thought about this very thing many a time. To think that Bill Gates couldn't see the importance of the internet for a long time too is wild IMO. IBM should have doubled down on the internet and making the PC the forefront of that and reduce the costs of their PC's (potentially make a loss to make gains in the internet) rather than sell the PC business altogether.
I call it the pendulum of computing. We first had mainframes with punch cards, then next we had attached terminals. We then went to distributed processing on PCs and now we are back to central computing with terminals (home computers) where a lot of the work is web based and processing being done on servers. This centralized way of doing things is much more efficient as there is no software distribution or updates (other than the o/s)
@@rty1955 IBM probably new the home computer was on the way and tried to push it in their favor until failure from smaller upcoming companies that had no corporate control at the time. That's when they decided they needed to enter that market themselves. The are still around in a big way, but cloud and AI. It was the Internet that put then on track, but they had to act quickly. While it may seem like 2019 was late to purchase Red Hat, they been using it on their servers since the 90s.
@@delscoville they have a supurb o/s for transactional processing and have for decades. They also have a supurb o/s for batch processing. too. While gates and jobs pushed their way Ino schools, IBM did not. I taught mainframe operating systems in college, not done anymore. IBM is in the HARDWARE business, and they do make the best general purpose machines on the market. Its just a shame that most people have no exposure to one. To work on one is truely a joy.
I find it important to try and learn about the origins of anything instead of just taking them for granted. I would like you to make a video about Xerox and its contribution to turning the PC more user friendly and why it gave it away to Steve Jobs instead of launching what was to be a revolution. Great job. Thanks!
Thanks! I am actually doing a video on that topic in the future, so make sure you are subscribed and have notifications on. You'll see it pop up some time!
It's a misstatement to say that Xerox "gave it away" to anybody. They built the Star computer for many of the same reasons GM built the EV-1 electric car -- they were testing the market and trying to learn what people liked and didn't like. What GM learned was that dealers would revolt if they started selling these electric cars because they required so little after-market maintenance work, which is where most dealerships derive their profits. Xerox was never in the business of building general purpose computers. They just wanted to see how people liked different features they built into the Stars, some of which went beyond the machines themselves.
I worked briefly at a place once, and one day I took a wrong turn and went down a hallway that happened to be lined with a whole bunch of these Xerox Star computers. I asked someone what they were doing there, and was told their leases were up and Xerox was taking them back. I'd heard about them briefly, and tried to learn more, but nobody would say much about them other than Xerox had asked the company to be something like beta-testers for them. Folks liked them, but the software was limited. Not surprisingly, they talked directly to Xerox copy machines that could be used as printers, which was unheard of at the time. They were all replaced with Wang workstations.
They came from Xerox PARC, which was mostly a research place, and they didn't get into any kind of production -- that was never Xerox's intent. That's where Jobs took his team to learn about the Stars, and Xerox was more than happy to show them off. You can actually see the results of some of what Xerox learned from the Stars on large copy machines introduced in the mid-80's -- the touch screens and how they controlled the overall machines. Remember, Xerox wasn't into computers, they made high-end copy machines and office equipment. They happened to be embedding computers into their newer copy machines, which allowed them to transform how they were used (from a UX/UI standpoint) and make them run MUCH faster. To them, Stars were just a way to test a bunch of new technology and get feedback from users. They were never intended to be mass-produced.
One could also argue that Xerox "gave away" the laser-printer market because HP believed there was a market for small, hi-res laser printers set up on a departmental level in businesses, whereas Xerox's vision was that people would use networking to send print jobs directly to their behemoth copy machines to get print jobs done. People forget it was Apple who released the first low-end laser printer that had amazing capabilities at a very affordable price. This was another thing they learned from Xerox. I believe the "engine" inside of their printers was built by HP (or maybe Canon), but they used a motherboard with software they designed in-house based on PostScript to do the rendering.
Something most people don't know is that Xerox was one of the key developers of Ethernet -- there were three companies involved in its first iteration: DEC, Intel, and Xerox. Stars played a role in that, to the extent that they could be interconnected allowing data sharing. Large Xerox copiers introduced in 1980 had early versions of Ethernet inside of them that was used to move data through the machines. (These were things that were 15-20 feet long that were used for high-volume copying.) My team at Intel interviewed a guy who had been working at Xerox for a few years on that very technology, and we got quite a lot of insight into how they had been using Ethernet inside of their big copiers. They had the equivalent of a half-dozen IBM PCs (without the monitors) inside of these big copiers, using early versions of Ethernet that required a 12"x12" circuit board to implement, in addition to the main CPU board. It was VERY Expensive tech at the time!
Super informative. Keep up the great work
Thanks for the kind comments Curtis, if you've not already subscribed and shared this video i'd really really appreciate it, since I'm only a small channel. I have lots of new stuff like this I want to do this year
I dabbled a bit writing in 6800 assembler in the late 1970s and later some 6502 assembler. When the company I worked for got the 8088 based IBM PC I couldn't believe what had to be done for memory access on that chip. Still, the PC was the safe bet in the corporate world. The first application I wrote was a GUI front end for a dial-up TTY money transfer system. Since there was no hard drive at the time, just two single sided floppy drives, the whole application, written in BASIC, was stored on two 5 1/4 inch floppies with six tutorial/help discs.
I wrote a keyboard handler in Assembler for MS-DOS in the eighties. Yup, some memory access ;-) The funny thing is that bit of code, compiled circa 1990, still sort of works on dosemu under Ubuntu MATE 20.04.
Thank you for including the TI-99/4A and the TMS9900 👍🏻
How come this video doesn’t have loads more views? Brilliant. Enjoyed every minute of it. The Bill Gates accent was a massive bonus, completely unexpected and surprisingly accurate. I love the 5150. Arguably the most iconic computer of all time. I love it so much I had to buy one and restore it. It now proudly sits on a round white tulip-shaped table, next to a rose in a white vase. (Most people don’t get the reference.) Some very old designs, like a Porsche 911, a Rolex GMT Master or an IBM 5150 are utterly timeless. The products of genius minds well ahead of their time.
Thanks so much for your kind comment! I wish my channel had a bit more visibility, I guess I'm beholden to whether YT's algorithm loves me or not, or whether people share my videos.
I know what you mean about the aesthetic/designs, they really are timeless!
Had both a GMT Master and a 5150 but not at the same time, would have loved to have them now still. that was like... hmmm... about 30yrs ago now. But yes, very iconic, both of them.
But in terms of market influence, the Compaq 386 Deskpro also was important, taking PC leadership away from IBM forever. Of course, Compaq's credibility was based on its original Portable, a better PC than the 5150.
The first PC on my desk at work was a "Columbia" 8086-based clone from 1982. 2x 5.25" floppies, a full 640k of memory, 4MHz processor. It lacked the IBM BASIC interpreter, but anyone with access to an EPROM burner, an original IBM ROM no scruples about piracy could easily fix that. Later on, compilers became available and programs which took 4 or 5 minutes to run through an interpreter ran in seconds as a compiled executable. My first upgrade was to a -286 machine with a massive 20MB hard disk, which we were never going to fill 😂. As someone who has built many PCs over the years, I like the ability with the IBM PC and its descendants to customise the build with the best components for my, or my colleagues' needs. For that reason, I have never been a fan of Apple's closed architecture model and proprietary software.
the 8-bit data bus of the 8088 opened the door to compatibility with lots of 8-bit devices that already existed and were mature. For a project looking to get to market within a certain window of time (or else IBM risked that market moving on past them and leaving them behind), the 8088 was an immeninently more sensible choice over the 8086. Is a shame Motorola didn't come out with a 6809 that used segment registers too to address beyond 64K as then they might have been a strong contender. The MC68000 was a superb CPU but there was just no two sourcing of its manufacturing and IBM refused to let themselves be dependent on a solitary supply source for the CPU
The Z8000 would have been a better choice. second source AMD, SGS-Ates, Toshiba, and Sharp.
The 68000 was second sourced by Hitachi, Mostek, Rockwell, Signetics, Thomson/SGS-Thomson, and Toshiba
Motorola 68K was way ahead of its time. It had 32-bit ISA and as computers which use it showed even older programs took advantage of it if they were designed for 32-bit words where such operation on 8086 was dog slow, let alone on 8088... that said programs didn't often need even 16-bit words and often used 8-bit. This is especially true because on 8088 there was big performance penalty from using 16-bit memory/io accesses.
@@e8root But Intel was ahead of the pack in providing an optional floating-point processor, a good one with 80-bit internal precision. For a scientist, that was a big plus.
@@awuma best argument so far. I too would rather choose processor with optional 80-bit FPU, especially if it was the cheapest option
Definitely - the Boca Raton group needed to meet its deadline, and 8-bit off-the-shelf support chips like the 8255 were readily available and priced very favorably.
Our college had a IBM 360 back in the mid 70's. A IBM representative came into our class and told us IBM would never make a PC. They were not in the business of making PC's. Not long after that they made a PC.
Other than the loud music at the end of it that drowned out your quiet speech, making it hard to tell what you were saying, it was a good video.
All those CoCOs shown during the 6502 segment ran Motorola 6809 CPUs.
thanks for the low-down!
What a trip down memory lane!
Thanks.
When the pc was first introduced it had a choice of two operating systems:. UCSD Pascal, (an integrated os, language and debugger) [supported by soft tech microsystems] and MS Dos. Because the market was all about VisiCalc, and MS Dos had Lotus 123 (which wowed with integrated graphs you could display and print) the pc,msdos,lotus123 combo ate apples lunch. Oh, and UCSD Pascal faded to oblivion. Btw, the story about IBM, CPM and Bill Gates is way more interesting then you portrayed!
@ Paula, One thing that IBM did on the PC (altho limited) was the NUMERIC
KEY-PAD. We had one hooked up to Sys 34 then Sys 36. I loved using that thing when I could, because the regular terminals could not keep up with my numeric data entry. The regular terminals would lock up until things cleared which caused me extreme frustration. The Apple systems of the day didn't have that keypad and relies on somebody being a touch typist for numbers.
The story of IBM/CPM and Microsoft is a wonderful story, that is due its own video here of its own. It was just a footnote on this video. Indeed, IBM did offer CP/M as an optional purchase with the PC not long after the launch of the IBM PC, following a conversation with Gary Kildall.
I went with an Apple ][+ in '81. I was a software developer on IBM Midrange systems at the time, the S/34 and the S/36. They used 8 inch diskette drives, and you could not get IBM compatible 8 inch drives for the IBM PC. But you COULD for the Apple ][ line! I ended up doing a lot with my Apple ][, even adding a Z-80 coprocesser and running CP/M. It was a great machine! I still have it, though it has been years since I have powered it up.
I too am old enough to recall the "what do you need a PC for?". I did get a lot of old mainframe parts and managed to interface a Wyse terminal to my Sinclair Spectrum, a line printer (and I wrote a program to print out PI!). I also managed to reverse engineer a data entry keyboard to use on the spectrum which was a big deal back in the day as the keys where all using hall effect sensors...
Now, was the serial driver polling or interrupt driven? The original IBM BIOS polled the serial port... which was no good if you were using a multi-tasking add-on to MS-DOS.
Awesome vid man, kudos for getting Billy G to do some voice over too! 🤭
Bwahahaha! Yeah originally he was against doing it, but I said he could call me Al and and all of of a sudden he agreed. 72 hours later some men in suits arrived at my door and knocked me out. I awoke hours later, dazed, on a leer jet with a shorn scrotum and dressed up looking like Paul Allen in 1970s drag. Wrong Al. Still. Damn fine interview, worth every lap....
This video needed a lot of simplification, but some things are misrepresented. The 8088 was basically a 8086 with fewer external data bit lines. This mattered a lot when hand-designing a motherboard and bus with standard IC chips. The bus was probably not meant to be "open", but all the slots on the first PC were needed to add functions that could not be placed on one board without extreme cost due to production quality limitations.
I had an Apple II. God I can still remember how it felt picking it up, opening the lid.....I wish I had never sold it and probably for a song. Brings a tear to my eye.
I remember rescuing a 5150 from my IT teacher in upper school school back in the mid '80s. I think he had been given it by a local chemicals company who used it as a lab data logger and he didn't know what to do with it - the school network primarily being Research Machines based, there was no real use for it, so it was destined for the bin. In hindsight, I think I got a good deal, what with it still being only 5-6 years old and still reasonably current. Served me well for a few years before being upgraded to a 286. Still have the motherboard for it in my pc parts bin :)
Great story! Yes you got a good deal!
2 notes about the 6502..
- while it is 'risc like', it wasn't designed as a risc cpu.
- after the 1980s, 6502 derivates still saw a fair bit of use in calculators, and there are still microcontrollers around today which use a 6502 derived core.
I will have to learn a 6502 based PLC for my technician cert too. Our teacher said we'd have to program PLCs with binary because that's uniform with every PLC which is just dumb. All the PLCs used today can be programmed in a visual language (ladder diagrams) or C or assembly or BASIC (the documentation we got didn't even mention machine code)
@@loganmacgyver2625 You aren't wrong with regards to being able to use something more friendly than directly writing machine code, and no-one in their right mind would be programming a 6502 based PLC, or any PLC, or anything 6502 based, by manually assembling a binary. Yes, we used to do that on home computers decades ago, but even 'us old school 6502 coders' usually use modern tools for that now.
And yet, you are currently learning, and for that the 'dirty' hands-on experience of doing that at such a low level is far from dumb, it gets you skills which are extremely usefull when things don't behave as you expect or the documentation suggests.
And 'machine code' merely means doing the assembly manually, which really isn't that big of a deal, but very helpfull to learn about why assemblers work the way they do, and being able to verify their output.
Learning how to do this without modern tools means you'll be learning the absolute basics of this, which is, as your teacher says, universal, even when not every PLC is 6502 based.
At the end of the day, unless you're a programmer a computer is only as good as its software library. There's so many examples of fantastic hardware(not just limited to computers, but game consoles, phones and other things) that flopped hard due to lack of software.
Hitting the ground running with a mostly compatible, but more powerful Z80-ish CPU led to easy to produce software that was ready at launch, with a steady stream of others coming shortly thereafter. Add in easy expansion and it was almost an easy victory.
I found it very interesting for me. Thank you for this video!
Glad you enjoyed it!
I remember that IBM ran tons of commercials on television despite the fact that you could not find a PC in stock anywhere.
IBM's tag line for the PS/2 was "How you gonna do it? You're gonna PS/2 it!". Less-than-impressed software geeks quickly turned that into the somewhat less complementary, "Hey, IBM! PS/ON IT!". :-)
Several years ago, I was doing volunteer work for a local thrift store going through their donated computers cleaning financial and other personal information off of hard crimes and putting complete systems together where possible. I was given an early Thinkpad that needed a power cord in appreciation.
I found a power cord on eBay and I was good to go. I knew the original owner and she gave me the original bill of sale and that PC: $5,237. It was running Wndows ME and I don't remember how much memory or the hard drive size, but it was laughable.
Interesting story. Thank you!
Excellent documentary! Thanks for this :)
thank you!!
Great work! The only criticism I have is when you pronounced “Hertz” as “hert”. (Hertz is named after Heinrich Hertz, it’s not a plural. :P )
ok, I wasn't the only one who thought that was why it was being pronounced that way!
As some one who has lived through what you describe, I appreciate your insight and summary of the early IBM PC. I bought the ill-fated PC-Jr as my first computer, a big mistake! Still, I use an IBM descendant desktop to this day, a Dell computer.
Thanks David!
Since 1995, I have assembled my own PC's from selected subsystems, for Linux. As for laptops, IBM and then Lenovo mainly T-series have been my go-to machines since 2001, reliable performers usually bought in the second-hand market (which generally features ex-corporate units) and with excellent Linux compatibility. Compared with pre-PC computers, the commodity market for PC subsystems provides amazingly reliable gear if you choose with a little care.
A huge player in the dominance of the IBM PC format was the government of Taiwan.
Just as soon as someone reverse-engineered the BIOS and produced one which was close enough without being a direct copy, they decided the little island of Taiwan was going to become the world centre of IBM clone production, which it absolutely did.
They were churning out millions of motherboards, the early ones being direct copies of the IBM boards. Most of those pioneer companies still exist, production now being in mainland China.
Absolutely true. What a revolution that was!
The very early "clones" were not as good as those by Compaq, who perfected the clean room design process to reverse engineer a legal non-copyright-infringing but fully compatible BIOS. Othrer companies eventually produced such BIOSes, and these made the fully compatible clone market possible, such as the Taiwanese clones. The earliest efforts were not fully compatible (e.g. Hyperion, Columbia, Corona etc.).
The first motherboard that was close to acceptable quality was Diamond Flower International aka. DFI. They are still in business but no longer make motherboards.
Compaq did not reverse engineer the BIOS. First the source code to the BIOS was published in full in the Technical Reference that anyone could purchase for $150.00. Second, IBM attempted to file suit for copyright infringement but after a Federal Copyright judge reviewed the IBM and Compaq source listings, he concluded that Compaq had not infringed on the BIOS copyright.
@@MrOliverHBailey Yes true. I used to have a copy. The term Reverse Engineer is not really correct I suppose. More a case of fully understanding what every part of the BIOS did, in great detail, in order to write a fully compatible one.
At 19:48 there is a diagram showing the 8086 & 8088 address & data pins. The words seem to be in the right place, but the braces showing which pins are short. On the 8086 the brace showing the data bus is 13 pins long & the address bus is 17 pins long. On the 8088 the brace showing the data pins is 6 bits long & the Address brace is 16 pins long.
I personally think IBM purposely did not choose a fast CPU because they didn't want it to gut their mainframe and mini computers. Also the choice of DOS which was bad was on purpose as well as they want to sell mainframes if you wanted a stable platform.
Could be plenty of truth in that..perhaps we will never know...
So intel amd and cyrix all worked together to make slow cpus. And Microsoft had no competition for dos.
@@danc2014 No. At 3000 nm back in 1981 space was very very limited unlike the 5nm size circuits today. Multitasking? That required much more space and less chips per wafer? More l2cache? More (actually I think it was closer to 1990 before cache were on chips). Wanted more registers so less ram calls? Required more space and less chips per wafer. etc.
Remember the bigger the chip the less can be made from a wafer as costs to etch were made by wafer. when processing costs There were many duds too in those days. In essense a bad spot would be more contained if the chips were smaller for greater yields.
DEC and IBM also made their own chips so they were more advanced as Intel was a tiny company in comparison
Big Blue was NOT worried about thier mainframe market from the PCs at all...they knew it was a toy. That's why Big Blue never got involved with it. I was working on IBM mainframes and all tje people I spoke to about the PC just laughed it off. The PC could NEVER do what the mainframe does. It still cant.
That makes so little sense.
I should add: This my first computer., bought new for $3000 for the setup. (Robbery!) My 7th grade son learned computing and DOS on this machine. I modified it for 640K and 2 disk drives. When IBM quit selling the PCjr, I bought close out mother boards and other pieces from which I put together computers for my brother and sister. It gave them a start in computing as well. So the PCjr served a good purpose, even if it was expensive for my first machine.
These guys were geniouses.. Thanks a lot to them..
Used to be a warranty repair shop for Commodore. They actually produced a PC clone too which had some interesting bugs.
Looking at the original PC designs you could see some errors that became standard. The printer port was missing one wire to be a fully bidirectional eight bit parallel port , all the chips would have supported it, but no one would change anything after it was in production.
I used to repair C64, C128 and VIC20 boards. Everyone else just swapped them out and sent them back to Commodore. The biggest problem was ailing RAM, followed by the PLA chips. I only scrapped two boards. One that the owner decided to install JiffyDos, and pried the ROM chip out of the broad, ripping out all the plated through holes. The other board was full of intermittent internal layers.
I saw one of the Commodore PC clones. Our club in Orlando used it to run our BBS.
@@michaelterrell
Did chip level too 😁 when commodore started leaving the unused pins out of the power supply connectors, the number of blown RAM went way up. The plugs would fit in partly rotated and the 8 RAMs protected everything else, by becoming crowbars.
Friend had a C64 that died from emp from a nearby lightning strike. 18 chips, fixed as a challenge. There was not a single electronic device in his house that survived, the C64 was packed in its box, so he hadn't even looked at it until after the insurance claims.
Yeah, having someone else try, and fail, to fix a board generally means the original problem and progressively less likely ones.
Funniest, a chip replaced by jumper wires. 🤦♂️
@@garygough6905 I lost a nearly new SVGA monitor to lightning. It was on my computer desk, with the video cable, and power cord wrapped around its base, and not connected to anything. A battery powered digital thermometer exploded, and my phone line was fried all the way to the Central Office. There was a mash behind my home, and there were constant strikes. My shop had no windows, but it would light up enough to see, with every strike. There was a report of over 1100 strikes in under a half hour that day.
My 5 MHz Leader scope had been stolen, so I was troubleshooting with a logic probe and a 500KHz scope. Most bad RAM had a bad undershoot, so I used around a 10 Hz sweep and looked for that. A 'calibrated finger' found some, even faster. Cold was dead. warm was likely working, and hot was defective.
I used a lot of solder wick, but I used the NASA approved 'Wet Wick' method of dipping the copper braid into RMA flux. I learned that method at a defense plant where I did QA on the PRC77 radios.
The first computers that I repaired used the Motorola MC6800 chip.
I have a SWTPC computer in storage. I'll probably list it on Ebay, since I'm getting too old to spend time on some projects.
My first PC was built in a used case, and with boards that I had repaired.
Did you ever see the fliptop XT cases? I have a couple of them.
Don't forget that the IBM PC was the 5150 because the tape based system you showed was a BASIC/APL computer called the 5100. There was also the 5110 and 5120.
Is that a TV commercial at 4:21 for a 5100?
I was an avid PC user as a hobby, starting with my first S-100 8080 based, I remember when the rumors of IBM releasing a PC started getting hot and talking to my boss and saying, if IBM comes out with a PC we better be all over it, it will legitimize buying PC's, long regarded in fortune 500 companies as toys. Those 3 little letters, IBM, changed all of that overnight and the rest as they say is history.
Very true!
What a great video and well researched
It would have been good to have asked one of the dirty dozen why they chose the 8088, assuming any one of them are still alive
Thank you 💗💗. Yep,. I would love to know too!
Great vid! Keep e'm coming!
Thanks! Will do!
My father worked for them in the late sixtys. I started with the Vic-20. I didn't know my father, but I followed the same path.
"Watching on a descendent of the 8088..." Not me. Mac guy since 1985. Though I've had to be a DOS/Windows user at work. Enjoyed it. Glad I watched!
Thanks for making this documentary. It was very interesting and entertaining. I'm subscribing =)
By the way, what's your accent? where are you from?
my accent is Scottish. Thanks for subscribing!
@@AlsGeekLab Nice accent! Thank you for making the videos and taking the time to answer. Have a great day!
Well done! New sub here, this I think is the best distilled summary of the history behind what most of us use in everyday life today, in a good level of detail and told in a short enough way that you don't need to be a nerd to understand. I'm a consummate nerd who got his feet wet programming a VIC-20 in the '80s, grew up through C-64, Amiga, Apple, PC, to spend a long time on DEC and other larger systems ending up as a systems architect and software developer. Nowadays it's rare for me to ever get into detail with people, but I sure wish there were a way to refer people to something like this video when I was a teenager and more outwardly excited about what I do.
In 1981 8088 was not underpowered, actually it was fast minimally like other PC of his age. Also limitation of 1MB od addressing RAM (640k for DOS) was not problem, because in 1981 was 64k pretty standard. 640k was also enough in late 80s and then came 80286 as standard, so for 8088 was never 640k real limitation..
I agree. All too often the earliest and least capable PC setups are compared unfavourably to desktop computers released years later, by which time the PC had already moved on.
@@vapourmile Yes, they forget, that other 16 bit machines like the Amiga only had 512 KB. And don't forget, the 8086 and even the 80286 still used bank switching to really address 20bit of memory space: the processor splitted the memory into smaller 16 bit chunks (64 KB big), that you could address with a quite complicated memory adressing (which made it possible to address every 1 KB of you 1 MB memory with the upper 16 bit hex codes and the specific byte of this chunks were addressed by the lower 16 bit hex code - thus giving you three options for jumps: small (8 bit range, for jumping up to 256 bytes up and down), normal (addressing your specific 64 kb block) and "far" (for jumping farer away than 64 kb).
Also in this architecture you were able to define specific 64 KB blocks as "Code segment" (CS), "Data segment" (DS), and "extra segment" (ES) and even "Stack Segment". When your computer program or data was under 64 KB big, you never needed to fully address the Data segment or Stack segment (where temporarely data was stored), which made your code shorter (the DS/ES/CS/SS was coded into the opcode of the machine code).
This was also intentional as they didn't want it to gut their mini computers and mainframes. Limitations and a terrible OS was intentional even if 640kb was unrealistic in 1980 IBM wanted to make sure it stayed that way in the later part of the 1980s
Actually in 1981 64 KB was much, far from standard. When IBM PC was released it had two memory options: 16 KB and 64 KB. In 1981 64 KB of RAM was $400.
The 640 KB was a huge limitation. Sure 286 came but the success of DOS which operated in real mode meant it could not be utilized fully. Various gimmicks were used to circumvent it like LIM EMS and XMS memories. In the early 90s there were DOS extenders that could allow breaking the barrier but only switch to Windows made it history.
Also note that not everyone had the newest CPU. I used 8088 until 1991. In fact I got it only in 1987.
I worked for a company called ViewStar in 1990 that ran a LISP-based document imaging system for PC's that used a DOS extender for running the LISP interpreter. The base configuration was an 80286/80386 with 5MB (!) of memory. I encountered them ca. 1988 as an IT analyst at a bank that was implementing an Accts Payable app at the time. The PC hardware was a Compaq 386/20e (640K RAM) with a 4MB expansion card. The 4MB module cost as much as the base machine.
The software stack was MS-DOS + TSR's for LAN drivers + DOS extender shim in the lower 1MB of RAM (originally 640K until Quarterdeck's QEMM and Microsoft's XMS came along) and the LISP runtime in the 1M-5M space. It even had a mouse-driven GUI that was visually inspired by the Macintosh and ran on 19" monochrome monitors at 1600x1200 pixels.
It's unfortunate that Xenix never really took off on the 80286/80386. By the time all of the lawsuits with SCO were over, it didn't matter anyway. But of course, now we have Linux, and UNIX did go on - we have the BSD-derived distributions and the remaining commercial variants.
I use to service and repair several types of banking terminals (as we called them) and i saw the whole pc uprising which ended up terminating several thousand jobs which actually repaired these communication devices in the customer offices. I repaired sever type terminals and even saved a complete line once by replacing glitch producing transistors. I should have gone to the Company president to show why his largest computers suddenly started passing... instead a supervisor refused to pay me my over time which helped me produce this result.
Just being here able to watch a video on a computer about the evolution of computers in just the past 60-70 years is completely crazy. We have no choice but to thicken our skin against wonder itself.
This is awesome yet so sad at the same time. I love watching old vintage videos such as this but what’s sad is knowing just about all of the folks in the videos have passed ❤
Yes, there are a few still alive, but I think all/most are in their 90s now. I would love to hear from any that are still alive
Great Documentary Thanks Buddy 🙂
Glad you enjoyed it
According to IBM, I got the first one sold, under a large GM corporate buy that I set up. It worked fine but the manuals were not ready so they gave us the original printing proofs to use.
Whoa! So you had like serial number 1 or something super low?
That’s right. IBM provided their sales records to me so I could verify the in-service date and serial number for warranty purposes. I think I had the first 40 before a different owner in Chicago showed up on the printout.
@@alanmayfield5476 very cool! Imagine how much those would be worth nowadays!
It would have been interesting for you to talk about why IBM chose Charlie Chaplin as its mascot, and what other mascots they were considering at the time (one was the Muppets.)
It was easier. Charlie Chaplin had died recently (in 1977) and Kermit the Frog sounds a lot like Bill Gates.
My first PC was a Schneider that had the processor board and floppy drive inside the keyboard. I learned turbo pascal on it and still program in FreePascal/Lazarus to this day.
This is a brilliant video, thanks a lot
Thanks! Glad you like it!
I had an 8088 emulator card for the Amiga. It was interesting, but Amiga DOS was so much easier and the GUI was lightyears ahead of Apple or Intel pc's.
My first computer was an Ohio Scientific Superboard (single board computer). It had a built-in keyboard, 4k of RAM, a 6502 CPU (can't remember its speed but I re-clocked it to 1MHz). Can't remember what capacity the BASIC-in-ROM was (8k I think?). You had to make your own 5V power supply and modify a TV to see any output. It had a cassette interface for program load/storage - it worked very well (mid to late 1970's I think). A lot of water has passed under the bridge since the days I used to fiddle with that.