https://every.to/the-crazy-ones/the-misfit-who-built-the-ibm-pc
In a burnished-oak corridor outside the committee room at IBM's headquarters in August 1980, two engineers pace nervously. Eventually, a door opens. Their boss, Bill Lowe, emerges from the board room next door. Before they can say anything, he smiles and nods. They laugh. They can't quite believe it. It's official. IBM is going to try and build a home computer.
Bill Lowe kicked off this ambitious project, but he wouldn't be the person who would finish it. That role would fall to his successor, a humble, cowboy boot-wearing mid-level executive, out of favor and kicking his heels in the IBM corporate backwater of Boca Raton, Florida. He would take Lowe's project forward, one nobody else in the company wanted. Just 12 months later, on August 15, 1981, a computer would launch that would change the world: the IBM PC.
This is the story of Don Estridge, the man who brought the IBM PC to market and changed business and home computing forever. In just five years he created an IBM division that almost nobody else in the company wanted to exist. By 1983, it had seized 70 percent of the microcomputer market and was valued at over $4 billion ($12 billion today). Under Estridge, IBM's PC division sold over 1 million machines a year, making it the third largest computer manufacturer in the world on its own. This story is based on contemporary accounts in publications such as InfoWorld, PC magazine, Time, and the New York Times, as well as books such as Blue Magic by James Chposky and Ted Leonsis; Big Blues by Paul Carroll; and Fire in the Valley by Michael Swaine and Paul Frieberger.
(Score: 5, Insightful) by Rosco P. Coltrane on Saturday June 08 2024, @08:16AM (20 children)
They let Microsoft ream them in the ass with the OS, and they failed to prevent Compaq from reverse-engineering their BIOS.
Well done IBM: you kick-started a gigantic industry and utterly lost control of it.
(Score: 5, Informative) by canopic jug on Saturday June 08 2024, @08:53AM (8 children)
The stab in the back that IBM got from M$ over OS/2 is well documented but mostly forgotten. IBM foolishly agreed to collaborate with M$ on OS/2 while allowing M$ to intermingle copyright throughout the code base. Further, IBM had an agreement for M$ to produce productivity software so that there would be something for regular business drones to use. M$ used its time to not only work on NT in secret, but build out applications for NT to the complete neglect of OS/2. So right before launch, when it was already too late, M$ backed out of the applications deal and began promoting NT. Because there were no word processors, spreadsheets, or simple databases for OS/2 it floundered. Because M$ owned copyright to a significant portion of the operating system's code base, even the OS could not be salvaged on its own, at least not in a time scale which mattered.
There were also the riders in the SDK contracts which required Windows95-certified software to have an NT version too. The rider in the corresponding NT SDK forbade the same programming team from developing the software for competing operating systems. That clobbered Apple too, and tripped up Linux, but finished off OS/2 completely.
That stab in the back was just the beginning of the treatment. The old crew at IBM new that and, better, knew how to make a high profit from Linux. Thus the highly profitable investments in Linux two decades ago. That generation is all gone. The current crew appears as a bunch of moles from M$ and work against IBM from the inside to further M$ goals at the cost of IBM's survival.
Money is not free speech. Elections should not be auctions.
(Score: 5, Interesting) by RamiK on Saturday June 08 2024, @11:21AM (5 children)
It wasn't a Machiavellian plot of intrigue and betrayal so much as a Machiavelli paper on governance. IBM had the same corporate disease that kept Microsoft from going into mobile: Their mainframe executives recognized home computers as a threat but fought any full-scale project proposals since that would divert funds and personal from their own divisions. Instead, they claimed the world would be run by a handful of mainframes and everyone will connect to them with cheap low-end terminals. So, the Entry Level Systems team ended up with just a dozen engineers that could only barely spec out an open architecture and out-source components and software despite knowing the risks just to meet deadlines on a tight budget and skeleton crew.
It's basically an unavoidable outcome of human individualism to have organizations shoot themselves in the foot when markets demand restructuring that would come at the expense of high-position members.
compiling...
(Score: 4, Interesting) by stormwyrm on Saturday June 08 2024, @11:51AM (3 children)
Numquam ponenda est pluralitas sine necessitate.
(Score: 4, Interesting) by RamiK on Saturday June 08 2024, @01:46PM (2 children)
Windows CE was using desktop window decorations that couldn't be manipulated without a digitizer so the platform was only suitable for full-screen embedded interfaces (vehicular media, gps, medical equipment, industrial machinery...). The politics for doing anything that didn't run Office was so bad that, putting aside how the xbox was done in secret internally, despite shipping Windows Vista with the Windows Media Center back in 2002 and the Kodi predecessor (Xbox Media Player) coming out later in that same year, Microsoft still failed to adapt Windows CE to that very obvious remote control use case and offer it to manufacturers which would only have required them to compile the damn thing and have it launch instead of the explorer.exe shell. In fact, not only they had nothing to offer by the time MK802 TV stick came out in 2009, they still had nothing to offer by 2014 when Android TV came out.
Them trying to outsource GUI toolkits and and buying Nokia is exactly the repeat of IBM outsourcing Microsoft for it's OS.
compiling...
(Score: 5, Informative) by Snotnose on Saturday June 08 2024, @02:51PM (1 child)
WinCE (very suitable name IMHO) was 100% unsuitable for any embedded application as defined by anyone outside Microsoft. In 2000 we were designing a new consumer product and the customer wanted WinCE. I called Microsoft. "What's the minimum RAM" "that's not defined". "What's the maximum Interrupt Response time?" "that's not defined". "How much RAM will each process need?" "that's not defined". "How much ROM does each process need?" "that's not defined". "How long does it take to switch tasks?" "that's not defined". "How big is the kernel with options A, B, and C?" "that's not defined"
In other words, the most basic of questions any embedded guy is going to ask all got answered "that's not defined". So unless your $100 consumer widget has a full fledged PC under the hood WinCE was useless.
Of course I'm against DEI. Donald, Eric, and Ivanka.
(Score: 2) by RamiK on Saturday June 08 2024, @04:18PM
It was mostly a misnomer for the point-of-sale type stuff they're currently shipping windows 10 IoT for and the odd NAS / gaming console. Basically, each CPU vendor ended up with slightly different specs per their hardware and partner board makers with Microsoft being 2-3 licenses removed from the edge developers. That is, you'd get marketing material from the board developer after seeing their stuff in a trade show with relevant specs and BoM including windows licenses and such before ordering a reference board sample and signing an NDA. You'd then muck about with it to see if you can get stuff to compile following their instructions and whether it suits your needs before ordering.
Basically, it was embedded to their customers but not their customers' customers.
compiling...
(Score: 5, Touché) by Nuke on Saturday June 08 2024, @12:28PM
Turns out they were correct, 40 years later. Except the "mainframes" are cloud server farms.
(Score: 3, Funny) by c0lo on Saturday June 08 2024, @11:53AM
Nobody was fired for buying Microsoft
(:very large grin: - thank God, I'm not old enough to have forgotten the original)
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 5, Interesting) by Snotnose on Saturday June 08 2024, @01:50PM
Not quite. I ran OS/2 back in the day and loved it. The only programs I had any trouble with was Falcon 3, where I didn't get voice, and stuff that competed with Microsoft (word processors, spreadsheets, etc).. Everything else ran great. In fact, when Win95 came out more programs worked under OS2 than Win95. The difference? MS Word and Excel ran fine on Win95 but not on OS/2.
I went to a Comdex in early '95, a couple weeks after Microsoft said they were releasing their new OS by the end of the year. I figured this gave IBM a 9-10 month head start so I made a beeline for the IBM booth. Asked about OS/2, got a blank look. Guy looked at his buds in the booth and asked "Anybody heard of OS/2?" Non of the IBM sales droids at Comdex had even heard of OS/2.
That was when I knew OS/2 was doomed.
Of course I'm against DEI. Donald, Eric, and Ivanka.
(Score: 3, Interesting) by stormwyrm on Saturday June 08 2024, @09:06AM
Numquam ponenda est pluralitas sine necessitate.
(Score: 3, Touché) by Thexalon on Saturday June 08 2024, @11:04AM (4 children)
It's worth pointing out that Apply had put out the Apple II years earlier, so that cat was already out of the bag, so to speak. The IBM PC was to a large degree IBM trying to protect itself from being overwhelmed by that competition.
But yeah, that was a bad move on the OS licensing deal. I wonder if they tried getting a really stripped down BSD working on that hardware, because it very much existed at the time and might have been a viable option.
"Think of how stupid the average person is. Then realize half of 'em are stupider than that." - George Carlin
(Score: 4, Informative) by canopic jug on Saturday June 08 2024, @01:46PM (3 children)
It's worth pointing out that Apply had put out the Apple II years earlier, so that cat was already out of the bag, so to speak. The IBM PC was to a large degree IBM trying to protect itself from being overwhelmed by that competition.
Apple hit it big at the tail end of the 1970s with VisiCalc, which put an affordable spread sheet in every small business.
Up until then, it was CP/M which was the mainstay of microcomputers. However IBM had just gotten worked over by the Department of Justice for its long running anti-competitive practices. Since that was before Little Bush, the DOJ still had teeth [www.ecis.eu] and the penalty for IBM was to choose either hardware or software but not both. They chose hardware, which meant they were on the lookout for an operating system. CP/M was well-established. What happened next has become muddied with time and m$ revisionism [wired.com]. Too many of the 'investigations' and retrospectives have ties directly to m$. Be that as it may, IBM cancelled the meeting with Gary Kildall about CP/M and Bill Gates' mother who knew people on the board at IBM got him a meeting instead. Here is where it gets really problematic. Early reports claimed that he sold what would be PC-DOS to IBM and only then (if I recall the sequence correctly) bought a CP/M clone called QDOS off of Tim Patterson for peanuts which then became PC-DOS.
From there, M$ inherited IBMs software half of the monopoly which they then leveraged into illegal, per-processor fees [justice.gov] which gained them royalties for every microcomputer CPU shipped regardless of whether it had an M$ system or any system at all.
Money is not free speech. Elections should not be auctions.
(Score: 2) by janrinok on Saturday June 08 2024, @02:51PM
I had forgotten much of this until I read your comment. I used CP/M for a while but then fell for the MS sales pitch for my desktops, with CP/M being relegated to my 8080/Z80 based boards. After Windows 2000 I switched to Linux beginning (I think) with Mandrake.
Thanks for the reminder...
[nostyle RIP 06 May 2025]
(Score: 3, Interesting) by stormwyrm on Saturday June 08 2024, @04:03PM (1 child)
TFA itself presents its own version of the events in question. Apparently as early as September 1980, Gates had already been asked to provide a BASIC interpreter for the IBM PC (since every microcomputer in those days had to have a BASIC interpreter built in, and it later became IBM ROM BASIC), so he was asked to recommend someone that could provide an operating system. Gates suggested Digital Research, since at the time they had an informal truce with them not to compete. IBM lowballed DR with a fixed-cost universal license of $250,000 for CP/M for the PC. They refused, since they very recently made a similar deal with Adam Osborne and when the Osborne 1 sold very well they were filled with regret. Negotiations dragged on, since DR failed to appreciate the kind of pressure that the IBM PC team was under. Meanwhile, Microsoft sniffed a business opportunity while DR stalled, and, breaking their truce with DR they found another OS that could be easily ported to the IBM PC: Seattle Computer Products' QDOS, renamed it Microsoft DOS and pitched it to IBM instead. Gates also offered to turn over all Microsoft's operations to the IBM PC ahead of its public launch. That basically sealed the deal for Don Estridge, as that was exactly what he needed at the time.
Numquam ponenda est pluralitas sine necessitate.
(Score: 3, Informative) by canopic jug on Saturday June 08 2024, @04:41PM
Gates also offered to turn over all Microsoft's operations to the IBM PC ahead of its public launch.
That's improbable but even if he did make that offer he did so knowing that the DoJ, had a case against IBM since 1969 [justice.gov], got around to forcing IBM to give up either software or hardware by the 1980s. As mentioned, IBM kept the hardware business and then got out of software for a few decades.
Money is not free speech. Elections should not be auctions.
(Score: 2) by stormreaver on Saturday June 08 2024, @01:09PM (1 child)
And we are all still suffering for it to this day, even if you don't use a Microsoft operating system.
And we are all better off for it. Reverse engineering the BIOS allowed for competing clones that dramatically lowered the price of a PC. Without this, the computing landscape would be completely different even today.
(Score: 5, Insightful) by owl on Saturday June 08 2024, @02:07PM
And, to a huge extent, forcefully created the relatively standardized market that became the X86 architecture. Compare modern windows PC's (each almost identical at the hardware API level) with the current Cell phone ARM hardware environment (where every one is a completely different, incompatible, set of hardware and hardware API's).
Without the rise of the clones, and the subsequent battle for every clone to be compatible (meaning run DOS and the major commercial applications for the IBM PC) we'd likely still have a cell phone like environment in the PC world (consider this is exactly what we had in the "PC market" prior to the rise of the clones -- An Apple II was very different to an Atari 800 which was very different to a C64, etc.).
(Score: 5, Insightful) by HiThere on Saturday June 08 2024, @01:19PM (1 child)
No. Apple, or possibly DEC, kick-started the industry. I don't think IBM would have succeeded if it hadn't been so widely copied. And prior to that there were the S100 computers.
The first personal computer used in the company I worked for was and Apple ][+. If IBM hadn't come along, some descendant of CP/M would probably have dominated. For awhile we used MP/M, but with the switch to 16-bit chips, the administrators decided to go with IBM. It wasn't a really bad choice, though I think that computers built around the M68000 had better hardware and nearly equivalent software. What gave IBM the win was two things: 1) "nobody ever got fired for chooseing IBM", and 2) there were lots of similar systems being sold (not just Compaq), so the software market moved in that direction.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 0) by Anonymous Coward on Saturday June 08 2024, @01:34PM
+Insightful/Informative
(Score: 0) by Anonymous Coward on Saturday June 08 2024, @07:02PM
(Score: 5, Insightful) by theluggage on Saturday June 08 2024, @01:46PM (4 children)
Creating any new computer system and getting it to market deserves some kudos, true.
...but let's be realistic - the personal computer industry was already booming and IBM were late to the party. The IBM PC was just an incremental development of the already-established 8080/Z80 CP/M systems. The processor was a stop-gap 16 bit extension of the 8080 (when there were chips with proper 32-bit internals available) chosen because it was easy to port 8080 code to and DOS was just a bought-in clone of CP/M. It wasn't even "open" in the modern sense - the idea that anybody could write software or make hardware for it might have been revolutionary for IBM, but it was already SOP in the personal computing world, and nobody could make a 100% compatible without IBM's proprietary BIOS ROM. I guess it had a nice keyboard, esp. for journalists used to bashing away on IBM typewriters.
The IBM PC's huge impact on the industry cannot be dismissed - but all that is really down to having "IBM" in large, friendly letters on the front, being marketed by a company with an existing dominant position in business equipment, by and for people with nice suits instead of black turtleneck sweaters. Apple screwing up the Apple III also helped. Even then, it might not have broken out from the corporate, IBM-customer "niche" and become ubiquitous if someone hadn't managed to clone the BIOS & MS hadn't hung on to the OS rights.
The actual hardware was nothing more than an OK "me too" entry in the pantheon of 8088/86 systems running CP/M-86 or DOS.
The IBM PC platform might have won the hardware wars (at least until mobile took off), but if you look at what modern PCs are actually doing, it's smothered with the fingerprints of Xerox, AT&T, Apple, Berkeley et. al.
I always wonder what would have happened if the Xerox management - who you'd think would be big enough and ugly enough to lock horns with IBM, and already had their smartly-suited foot in a lot of corporate doors - had actually marketed what they had in the labs, rather than effectively selling it to the turtlenecks at Apple...
(Score: 2, Interesting) by Anonymous Coward on Saturday June 08 2024, @04:05PM (2 children)
"I always wonder what would have happened if the Xerox management - who you'd think would be big enough and ugly enough to lock horns with IBM, and already had their smartly-suited foot in a lot of corporate doors - had actually marketed what they had in the labs, rather than effectively selling it to the turtlenecks at Apple..."
Xerox had a cash cow with the xerography patent.
They realized when the patent ran out, the Japanese would eat their lunch with low cost copiers,
so the "paperless office" was invented.
They still intended to do business the way they always did, with leased word processors.
The copier mindset never left the company, the Japanese ate their lunch, and every part of
their computer business flopped from the purchase of SDS all the way down to the 820 CP/M machine.
(Score: 2) by looorg on Saturday June 08 2024, @06:06PM (1 child)
... and we are still waiting. They have still the make the idea an exclusive reality. So much paper around the paperless officer ...
(Score: 0) by Anonymous Coward on Saturday June 08 2024, @10:24PM
(Score: 3, Insightful) by Nuke on Saturday June 08 2024, @06:57PM
TFA didn't say it started the industry, only changed it. The fact that the hardware and software said "IBM" on it made it acceptable to buinesses and corporations. I worked for a large corporation when the IBM PC came out and before then my group had a Commodore PET that we actually had to keep secret from the company IT department or it would have been confiscated. But when the IBM PC came out, that was OK, but only with the excuse that it could emulate a terminal to the company mainframe.
Home computers were then progressively replaced by the cheaper IBM PC clones because you could pirate the company software like WordPerfect at home, it being on floppies in those days. Before that I had a CP/M machine at home.
So it was a major change of direction, the largest in IT history that you can put a date to.