My apologies if this is too OT...

I've been reading up on the IBM 5150. I'm looking to add one to my collection but the eBay scalpers are really keeping the prices high.

The obvious advantage of the 5150 back in the day was IBM. But when you remove IBM out of the picture and look at the hardware itself, it's really nothing to brag about!

I mean, don't get me wrong, I love all vintage computers. But how was the 5150 better than the Apple II? In my opinion, it wasn't. And I don't say this as a fan-boy of any one computer.

I don't have any real scientific evidence to back this up, but when I read how the 8088 takes an obscene amount of cycles (IMHO) to do anything and the massive amount of support chips it needs, it's a small wonder it became so popular and important. Again, the secret sauce were the three magic letters, "I", "B" and "M".

Commodore certainly could have been a great contender for the business world. In fact, they were headed in that direction. Commodore wanted to drop the "toy" image of the C64. While the Amiga had some great success in the broadcasting world, it never caught on in the business market. People didn't think of Commodore when they thought of spreadsheets and databases. Despite Commodore releasing some PC compatible machines later on, their image (and marketing strategy) were too tarnished for the "serious world" of computing. Not to mention the computer industry associated Jack Tramiel with Commodore AND Atari later on. I don't think Commodore or Atari were ever able to forgive and forget how Tramiel treated them. But that's a story for another day.

What about Apple? Apple came really close with VisiCalc (not their product). People were buying VisiCalc and then asking what machine they needed to run it on. Imagine if Apple had really pushed the Apple II into businesses instead of developing the Apple ///. Hindsight being what it is, it's hard to believe Apple didn't see how they literally had a free pass into the business world before they slammed the door on it.

I guess Apple actually still wins in the end. While IBM is hardly bankrupt, they called it quits in the low-end / home computer business years ago. Apple is now worth more than IBM. I don't want to say Apple was lucky, but I do feel the iPhone (and to lesser extent, the iPod) is what pulled them from the ashes of forgotten computer companies like Gateway 2000 and Cromemco.

From what I read, there were at least a few people at IBM that wanted to use the 6502 or Z80 for the 5150. Now that would be a fantastic world indeed. Instead of the x86 architecture of today we would be using 4 GHz '816 variants.

It's fun to imagine these "what if" scenarios. I mean, the 6502 was really a prime candidate for the 5150 in my opinion. It had more than one source, the performance exceeded the 8088 (and maybe the 8086 in some scenarios) and the cost seemed right at the time. But it was IBM's previous dealings with Intel that kept them two together.

So even though the 6502 is my all-time favorite processor, and the 5150 is such a "complex, simple machine", I still find myself fascinated by the history of the 5150, the appeal of the machine to business users, and the utilitarian nature of IBM.

_________________
Cat; the other white meat.

Well, having recently retired from Big Blue a few years back... and having been transferred to Boca Raton back in 1984 to specifically work with the Personal Computer line of products... I'm in a pretty good position to address pretty much anything around the earliest of machines through the PS/2 lineup, as I provided the support for key customers (meaning the fortune 500 ones) and had sign-off authority on most of the hardware announcements.

So... the parts themselves... in short, the basic idea was to use off the shelf components 100%, meaning no IBM proprietary circuitry or components. Certain choices were made for very specific reasons... and some IBM history in changing computer architectures which caused some grief for their customers. If you looked at Motorola, they has zero compatibility from their 8-bit 6800 MPUs to their newer 68000 MPUs, so anyone investing in their 8-bit MPU program would have to start from scratch when moving to their 16-bit/32-bit MPU program. Intel had a history of carrying old code along.... plus the IBM Displaywriter (hardware box) used a 8086, so there was already a relationship with Intel.

Also, there was the fact that the IBM PC was more along the lines of "test the waters" and see... and the thinking/planning was the original 5150 (all IBM machines at that time had a 4-digit machine type) would have a production life of approximately 250K units. The bit that was IBM content was the chassis and the keyboard (which by itself was worlds better than any of the existing personal computer type machines). It was also designed and built to have a long life span, as that was typical IBM product design and engineering.

I don't recall anybody thinking about the 6502 or Z80 back then... so I can only consider it a rumor started elsewhere. There was a bigger internal discussion about the 80286 however... as it was a truly botched chip... rumor had it that Bill Gates had some input around it... the problem being, once you bring the chip into protect mode, there's no set of instructions to get back to real mode... hence the BIOS and keyboard workaround to tickle the chip into real mode and intercept the vector. This was done with the 16-bit releases of OS/2 (1.x) so you could run a protect mode environment and still have a DOS box. The real discussion was around dropping the 80286 entirely (for protect mode OSes, aka OS/2) and going directly to the 80286, which had the ability to run DOS virtual machines. Microsoft wanted to go this route entirely (after pushing IBM to bring out a 80286 machine) after realizing the inherent shortcoming in the 80286. Unfortunately... this was back in the mid-80's.... and IBM actually intended to fulfill it's promise to it's customers who bought the range of 80286 products (the original PC-AT, XT-286 and entry level PS/2 models). This was mostly the start of the IBM/Microsoft divorce, followed by many other internal discussions around windowing technologies, device drivers, etc.

All in all, I lived through all of this internally at IBM... as my daily job. To be perfectly blunt... the only reason that Intel and Microsoft are big and successful today is IBM's inclusion of them for the 5150 and follow-on products. The IBM brand and historical customer service is what legitimized using smaller processors in the business environment. Earlier adapters and software for the PCs provided connectivity to both mid-range and mainframe machines (via 5250 and 3270 emulation) and were quite a bit less expensive than an actual 5250 or 3270 terminal.... and allowed the end user to do much more. That was the start of terminate and stay resident programming, as you can hot key from a DOS prompt over to the terminal emulator, work there and hot key back to your DOS app.

How Microsoft came into the picture is a whole different story... for a different time however

_________________
Regards, KM
https://github.com/floobydust

First of all...wow! I had no idea you had such a personal (work) connection with IBM. Especially during the early PC days. I would love to pick your brain on the things you remembered and what IBM was like back in the day.



Yep, I've read that many times. I don't want to keep drawing parallel to the Apple II, but you could almost say the same thing about it. Woz especially wanted an open design using off-the-shelf components. But for him, it was probably more of a necessity considering their small budget. It's actually amazing IBM went that route.

But, as you know, IBM couldn't produce the 5150 if they did it through the normal "IBM way". Not in a year, anyway.




That's interesting to know. The fact I can run many programs for the 5150 on my modern Dell laptop is incredible. The same can't be said about Apple (running Apple II software on a Mac without emulation or extra hardware). Or Commodore (running C64 software on the Amiga).





HA! Little did they know...lol. I guess that's a "good type of problem" to have. IBM *really* understood the most profitable customer out there, the business customer. But it seems they didn't know it well enough in the beginning! I read in the "Blue Magic" book that it was said that IBM understood why the Apple II was such a success BETTER than Apple did.



I completely agree. Those keyboards are built very well (well, other than the PCjr chicklets...lol).



I've read somewhere (Wikipedia IIRC) that IBM's quality control would even inspect the rosin flux inside the solder.




I'm pretty sure I read that in the "Blue Magic" book on IBM. However, the book clearly states that some of the content may not be super accurate because most of it was taken from peoples' memory instead of documents from IBM. So perhaps someone falsely remembered that in the day and it became part of the book.



That sounds like a logical opinion. Where were they before IBM married them with DOS and the 5150? I'm sure they would have pulled through. Maybe they wouldn't be as big as they are today. But I think they would have survived. But I completely agree that IBM had a monumental role in their rising. Probably more so with Microsoft that Intel.



If I ever wrote a book, it would be on the rise of personal computers starting in the 70's and include companies like IBM, Microsoft, Apple, Atari and Commodore (among others). I think there are so many interesting and important stories that haven't been told.

Thanks for the walk down memory lane!

_________________
Cat; the other white meat.

IBM also threw piles of 5150s at high schools to get people entrenched into their ecosystem. Those schools then used them to teach word processing and general computer use, which fed well into their business user climate.

_________________
WFDis Interactive 6502 Disassembler
AcheronVM: A Reconfigurable 16-bit Virtual CPU for the 6502 Microprocessor

Computers in schools.... I don't recall us (IBM) tossing tons of them into the schools so much, I do recall Apple getting their machines in more schools early on however. One thing is certain... when I was in the field in NJ back in the latter half of the 70's and early 80's, every high school and secretarial school (yes, they had specific schools for training secretaries) were littered with IBM Selectric Typewriters... I know only too well, as I was in Office Products for about 18 months and trained on all typewriters, including Typebars, Selectrics and the newer electronic Selectric which had memory and record/playback capability. As time went by, the PC starting to replace the typewriters everywhere.... so that was a more logical transition.

The 250K units... yea, everyone chuckled about that The PC created a new industry and customer base for IBM... one which was moving pretty fast, both for hardware and software. IBM partnered with Microsoft for a ROM Basic interpreter, but... according to some sources internally, approached Dr. Kildall for CP/M-86... in short, that deal didn't fly for a variety of reasons, which gave Bill Gates an additional "foot in" by acquiring 86-DOS from Seattle Computer Products. The earliest version (DOS 1.0) pretty much had the same identical function calls for the first 32 (as CP/M). IBM did hose up legally however... giving Microsoft all the rights for MS-DOS (which IBM paid for all development costs for newer versions) for clones (thinking we're IBM... who on earth would buy a cheap clone instead of a real IBM machine). If there was ever a scenario where "hindsight is 20/20" this one takes first place for an eternity. Much later... and too many DOS licenses to count, we did get an IBM branded CP/M-86 from Digital Research. unfortunately it was simply too late... the DOS bandwagon was the biggest game in town and CP/M-86 became moot. I still recall sending a dozen or more cases of it (CP/M-86) to the dumpster when clearing out one of more storage closets at the Boca facility.

IBM was a lot faster to develop this type of hardware than you might think... as we didn't have to design and develop the chips first!!! The PC-AT came out in 1984, following the PC-XT in 1982... I did much of the actual documentation reviews (for the PC-AT) before announcement in early 1984... which also included the original EGA adapter.... and later daughter card to expand the video memory. Those were fun times...

As for Intel and Microsoft, yes... both would still be around, but not the giants they are today. In short, Intel couldn't supply IBM's volumes for chips, so even with the PC and PC-XT, we were second-sourcing from AMD... I still have a static box with AMD 8088 chips pulled from old motherboards. By time the PC-AT and 80286 were in production, it was so bad that IBM started making the 80286 CPUs in house... I still have one here at the house which was made at our fab factory. We did the same for memory as well... the industry couldn't supply our demand, so we started making memory expansion cards with our familiar square substrate chips with alloy caps on them (IBM designed/manufactured DRAM). Needless to say, AMD got a large boost from the IBM PC as well... so did other chip manufactures like Nec, as Intel couldn't supply enough 8272 (FDC chips) and we second sourced the Nec u765. The same story holds true for Western Digital.... as we used their disk controller interface chips for the hard drive controllers on the PC-AT and some follow on machines.

The PC, XT, AT, and PS/2 lines were all manufactured down here in Boca Raton.... and all of the DOS and OS/2 development was mostly done in Boca as well... sans the database and networking components. When the decision to scale down the site was announced, software development moved to Austin (mostly) and hardware development moved to Raleigh. The end of an era in some ways. After the site was sold, the one building out in the rear (building 51... which we internally referred to area 51) was given to Palm Beach County. Our industrial line of PCs were made there as well as the early robotic machines.

Palm Beach Country turned Bldg 51 into the "Don Estridge High Tech Middle School" I had an office there for a few years before going home office. I also got to visit the school multiple times... many memories indeed. Shortly after it opened, a handful of IBM employees, myself included, were approached to see if we had any of the older PC hardware pieces kicking around. The intent was to create a technology display at the school to showcase what happened in Boca previously. With timing being everything... I had literally just disposed of over 30 pounds of boards, processors, diskette and disk drives, power supplies, cables, adapter cards, etc.

The 80's and 90's were quite a time.... I'll always remember them fondly and smile about them in my old age. I worked there for 38 years... and I can say that pretty much all were a great experience, some better than others of course. Attached is a pic of some older CPUs I had kicking around... a couple 80286 chips, one made by Intel, one by IBM and a pair of early 486 chips, one being the first Overdrive processor, where the internal CPU clock had a 2X multiplier over the bus clock... the sign of things to come!

_________________
Regards, KM
https://github.com/floobydust

That was such a great read. Thanks for sharing all of that.

During your time in Boca, did you happen to run into anyone in the software/hardware business that would qualify as famous to nerds like me? Kildall, Gates, etc.?

_________________
Cat; the other white meat.

Yes, thanks KM for your story!

Several points here which I'd like to pick up on. Might IBM have chosen the 6502? Was the 8088 a bad choice? Might things have been totally different if IBM had not made the PC?

So, the possibility of a 6502-based PC. Yes, it was in the air, but not many people knew about it. Personally, I don't think it had a realistic chance, because although memory banking techniques were well known and widely used, there was a strong feeling that 64k address space was not enough. Possibly, from a marketing perspective or for other reasons, there was also a feeling that 8-bit micros (whatever that means) were not appropriate, and a 16-bit micro (whatever that means) was a necessary part of the spec.

We wrote about the 6502-based PC here - it was about IBM buying Atari, or at least co-producing some derivative of Atari's 800 series. Possibly it was not a serious possibility but a political manoeuvre. There was a physical prototype made too:



As it happens, we got some first-person commentary on that post, from a Tom Hardy, which I'd like to preserve here for posterity:


Was the 8088 a fine choice? I think so. Second-sourcing was crucial, as was confidence in having reliable volume production in good time. That rules out Motorola. The 8-bit bus of the 8088 is a big help in keeping cost down, even at the price of some performance, which gives headroom for improvements later at the top of a range. The fact that Intel's and Zilog's chips were clocked faster than RAM is a non-issue: the important thing is being able to access RAM at a given rate, to achieve a given performance. I suspect at this time there was no great focus on CPU clock rates as a measure of performance, which has always been a dubious thing unless you are comparing very similar CPUs.

What if IBM had not made the PC? I think something big would have happened anyway. Moore's law was going to continue, micros were going to get faster and cheaper, and businesses really wanted more compute power at department and desk level. Someone, be it Apple or Commodore, Olivetti or Siemens, Thomson or Amstrad, would have driven volume into business purchases, with reliable well-built machines. Possibly we would have had more diversity. Possibly that would have driven software houses to use different techniques and made more portable software.

I very much recommend Jimmy Maher's history series at https://www.filfre.net/ - there's a huge amount of material there, and it's well-researched and well-written. He starts the PC story here.

It would be very interesting to see where the world would be if Compaq was not able to successfully clean room the IBM BIOS and start the floodgates of "clones" everywhere to the point of ubiquity.

IBM may have helped legitimize small computers in business, but it was the clone market that exploded. Obviously IBM alone could make a sustainable market for that product, as demonstrated by the PS/2 series, which only businesses bought. Very few private people would purchase those vs any of the popular compatibles. And, of course, the PS/2 was able to leverage the bulk of the software available.

And, respectfully, that was MS-DOS that empowered that. IBM provided the hardware and original BIOS, but MS-DOS provided the abstraction layer that empowered the software. There were many, early, MS-DOS compatible machines that were not necessarily IBM Compatible.

Had IBM controlled both the hardware as well as the OS, who knows where we, or anyone (Intel, Microsoft, etc.) would be today.

While Compaq did do this, I'm reasonably sure they weren't the first. They released the first portable PC clone, but I think another company beat them to the first PC clone.

The name escapes me at the moment.

**EDIT**

https://en.wikipedia.org/wiki/IBM_PC_compatible

https://en.wikipedia.org/wiki/Columbia_Data_Products

However, the Wikipedia article says that a citation is needed to confirm this. So I'm not 100% sure on this.

_________________
Cat; the other white meat.

Well, to try and answer a couple inquiries and some comments on later posts, here goes:

Did I meet any important folks along the way (like Gates, Kildall)... not really, Gates generally met with IBM execs, but he was within about 5 feet of me at a customer conference IBM held in Dallas back in the late 80's or early 90's... again, he was talking with some execs. He presented at the conference as the special guest... and also mentioned that OS/2 would be the most important OS going forward.... well, he sorta changed his mind a bit on that one. Others.... hmmmm, hard to say. I was at Borland a couple times, as we were looking to license much of their stuff, Philip Kahn was usually around, came into the meeting and said hi, etc. Other companies I had visited over the years included Lotus (before IBM acquired them), Novell, Microsoft of course, Intel a few times, but can't think of any others off hand.

The clone wars... well, as the clones sprang up.... Compaq was among the first, I even met with them a few times during the OS/2 days, as they did some preloads on some specific machines as well. As for cloning the PC BIOS, IBM made it pretty easy, as the entire assembly listing was published in the appendix of the Technical Reference manual... original PC, XT, AT, etc. This was stopped when we moved to the PS/2 set of machines. IIRC, Jack Blackledge was the IBMer that wrote the original PC BIOS, but I'd have to dig out one of my old reference manuals... and yes, I still have a large collection of original PC hardware and software manuals. I was level 2 support for hardware and operating systems for many years in Boca and one entire wall of my office was stacked with a collection of the manuals.

Initially, there was only the name, DOS, which was marked on the manual binder edge, but was printed as "IBM Personal Computer Disk Operating System" in the manual. Hence, it was coined PC-DOS early on. When Microsoft started marketing and selling their own version of DOS, it was coined MS-DOS instead and was slightly different, as it had to work against other machines' BIOS, which legally couldn't be a verbatim clone of IBM's (BIOS). Side note, I also have the original BYTE magazines where Steve Ciarcia designed and built a PC clone motherboard... named the Circuit Cellar MPX-16, November 1982 issue.

The PS/2 machines moved the ball beyond the original ISA bus and for good reason. Chet Heath was the lead IBM architect that created the Microchannel Architecture... I had met and worked with Chet on numerous occassions... internal meetings, customer meetings and even at Comdex one year. Brilliant guy actually.... another brilliant mind on the MCA was Les McDermott, who I also ran into in numerous meetings. The clone vendors came up EISA.... as they didn't want to pay a license fee for MCA (couldn't blame them... their prices would be un-competitive mostly). After a few years of PS/2, I moved into PSLOB (Personal Systems Line of Business) which eventually morphed into IBM Software Group (makes me a pre-charter member). I focused on software going forward but kept up to date with the hardware as well. Still, the PS/2 line were excellent machines, robust, reliable and very solid... however, the price premium made it a difficult sale, eventually in the corporate world too... as the technology was advancing so quickly, that nobody cared about using any PC type machine for more than a few years, then replacing it with a newer, faster, cheaper box with more memory, more storage and faster everything and better video.

One nice feature that came with the PS/2 machines was better DASD (IBM term) meaning hard disk drive storage. The old ST-506 interface was showing it's age... so we started with ESDI drives. These were the first disk drives that had logical block addressing and better yet, on disk defect mapping, via spare sectors on each track. The result, no more long and exhausting low-level formatting of new drives.... 100% of the addressable data was good. This also prompted me to write, what was the very first hard drive cloning software (Packcopy). A side benefit of the LBA scheme allowed a smaller drive to be copied to a larger drive... which could then be extended on the parition. Packcopy was used for a long time internally... preloads, site machine rollouts, billable services, classroom training, even debugging of new file systems, like HPFS.

Trying to predict what might have happened had IBM not ventured into this space.... or waited longer, which could have changed the CPU selection, software, etc. is well, anyone's guess is just as good as anyone else's.... given some background knowledge of how things worked out.

Big Ed... you never cease to amaze me in how you find obscure pieces of information.... after seeing the pic of the Atari 800 in IBM clothing, I do recall Tom Hardy's name... but like you said, that was such a small confidential project, it was never disclosed to anyone internally.

On the trivia side... I still have a lot of the early commercials which got digitized years ago... which are the original ones with the Charlie Chaplin character and the later ones featuring the MASH TV crew with the PS/2 line. Fun stuff.... even one of the Bond movies (Golden Eye) featured an OS/2 Warp splash screen in the movie as it was shown initially in the theaters... Boris' machine near the end when everything was going up in flames.

Final trivia... Don Estridge, he was the IBM exec that led the original development team that created the IBM PC. I never knew how many were part of that team as the count varied depending on who you spoke to. On a sad note, Don and his wife were on Delta flight 191 that crashed at DFW in Dallas on 2nd August, 1985. They were among the casualties... their two daughters were in Dallas attending school. In any case, that was a sad time in Boca for the entire site. Don's original team had small lapel pins with a rose on them, a special link from those original TV commercials with the Chaplin character, as "Charlie" always finished each commercial with a red rose. At Don's funeral, his entire team placed their lapel pins on the coffin as a final farewell to their friend and leader.

_________________
Regards, KM
https://github.com/floobydust

The 8088 was the obvious choice, despite being only marginally faster than a 1mhz 6502 on many tasks.

- Supported 1mb RAM
- 16-bit integer math
- Math Coprocessor with high precision available (AMD 9511 and similar were lower precision, hot and slow)
- Can use cheap 8080 peripheral chips and 8-bit bus

By picking the 8088 IBM stumbled on an almost secret sauce - more memory space than almost any other computer in 1980-81, and a math coprocessor. Suddenly the spreadsheet is even more powerful and useful. Boom you end up in most of the offices in America.

The 68000, Z8000, and NS16032 were not ready and available in the quantities needed by IBM. No commercial operating systems or compilers readily available - remember everything had to be off the shelf!

Really it just came down to timing. If IBM were willing to wait a year they would have had more choices with more elegant architectures, but the 8088 was the best at the time.

You might be able to make an argument for a 6809 with an MMU. A 2mhz 6809 with MMU would have easily outperformed a 8088 PC, but Motorola had been clear the 6809 was the end of the line and the 68000 was the future. Why invest in a dead-end platform? No math co-processor for business spreadsheets.

The other 16-bit chips out there in 1980 were problematic, the NI1600 (so slow...) and Fairchild Nova clones were not practical for IBM to use either.

If I recall correct, software could be ported from the 8080 to 8088 easily (compared to 8080 to 6502 or 6800). So part of what helped make IBM's PC an early success if that existing 8080/cpm business software could be quickly and easily ported to the 8088/ms-dos. cbmeeks said Commodore was a contender in the business world but I think they had already lost due to that port of cpm to dos, by then IBM had already one.

floobydust, I don't think Microsoft would have been around. The history I recall is Bill gates daddy licensed IBM mainframe software. Then IBM needed an OS for their new PC and anyone in the mainframe world thought pc's were toys so daddy tossed the sale to his son who wanted to get into the tech business. Son Billy then found a company and bought their software and licensed it to IBM.

If IBM had not done the PC or had written their own in house or had bought or subcontracted someone else then Microsoft would not have been. A lot of younger people think Bill Gates was smart or a good businessman but in reality he just worked out a license deal exactly the same as was standard in the industry at the time, he did nothing new or different or smarter than any of the other software salesmen, he just happened to get lucky.

Hmm, that hyper-condensed history doesn't work for me at all. I do recommend Jimmy Maher's uncondensed histories - he writes well, and has done his research.
https://www.filfre.net/2012/05/16/

As Ed said, your history isn't quite up to snuff.

IBM first approached Digital Research about getting CP/M running on the nascent PC, due to CP/M being a de facto standard at the time. Theories vary about why it didn't happen, but the one that has been generally accepted is IBM had a deadline to meet and Gary Kildall of Digital Research failed to respond in time to IBM's query. Gates approached IBM at the right time and there you have it.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

There's a lot of water under this bridge now... one such story I heard inside the blue walls was, Dr. Kildall purposely went out riding one of his horses knowing that the IBM team were visiting to strike a deal for CP/M (-86). Apparently he thought they would have no other option and he was in the position to do things his way. As the story goes.... the IBM team didn't appreciate the attitude, waited for a while and eventually left without having the meeting. Further.... it was noted that Mrs. Gates (Bill's mom) knew the IBM CEO (John Opel) at the time and that allowed an additional door to open (beyond ROM Basic). However.... just stories that filtered around the halls... not sure if anyone can actually validate them.

In any case, I'm still adamant that the only reason Intel and Microsoft are the giants they are today is solely because of IBM's original PC and the industry that grew from it. Things would have been quite different if Motorola's 68000 was chosen as the CPU and a proper OS developed up front, thinking Unix from Bell Labs. But again... the original IBM PC was really a "testing of the waters" for the acceptance of a personal PC from IBM. Who knew....

_________________
Regards, KM
https://github.com/floobydust