IBM, dominance, and big tech

Recently, popular Apple blogger John Gruber has been on a mission to explain why, exactly, tech companies like Apple don’t need any stricter government oversight or be subjected to stricter rules and regulations. He does so by pointing to technology companies that were once dominant, but have since fallen by the wayside a little bit. His most recent example is IBM, once dominant among computer users, but now a very different company, focused on enterprise, servers, and very high-end computing.

Gruber’s argument:

It wasn’t too long ago — 20, 25 years? — when a leadership story like this at IBM would have been all anyone in tech talked about for weeks to come. They’ve been diminished not because the government broke them up or curbed their behavior through regulations, but simply because they faded away. It is extremely difficult to become dominant in tech, but it’s just as difficult to stay dominant for longer than a short run.

Setting aside the fact that having to dig 40 years into the past of the fast-changing technology industry to find an example of a company losing its dominance among general consumers and try to apply that to vastly different tech industry of today is highly questionable, IBM specifically is an exceptionally terrible example to begin with.

I don’t think the average OSNews reader needs a history lesson when it comes to IBM, but for the sake of completeness – IBM developed the IBM Personal Computer in the early ’80s, and it became a massive success. Almost overnight, it became the personal computer, and with IBM opting for a relatively open architecture – especially compared to its competitors at the time – it was inevitable that clones would appear.

The first few clones that came onto the market, however, ran into a problem. While IBM opted for an open architecture to foster other companies making software and add-in cards and peripherals, what they most certainly did not want was other companies making computers that were 100% compatible with the IBM Personal Computer. In order to make a 100% IBM compatible, you’d need to have IBM’s BIOS – and IBM wasn’t intent on licensing it to anyone.

And so, the first clones that entered the market simply copied IBM’s BIOS hook, line, and sinker, or wrote a new BIOS using IBM’s incredibly detailed manual. Both methods were gross violations of IBM’s copyrights, and as such, IBM successfully sued them out of existence. So, if you want to make an IBM Personal Computer compatible computer, but you can’t use IBM’s own BIOS, and you can’t re-implement IBM’s BIOS using IBM’s detailed manual, what are your options? Well, it turns out there was an option, and the company to figure that out was Compaq.

Compaq realised they needed to work around IBM’s copyrights, so they set up a “clean room”. Developers who had never seen IBM’s manuals, and who had never seen the BIOS code, studied how software written for the IBM PC worked, and from that, reverse-engineered a very compatible BIOS (about 95%). Since IBM wasn’t going to just hand over control over their platform that easily, they sued Compaq – and managed to find one among the 9000 copyrights IBM owned that Compaq violated (Compaq ended up buying said copyright from IBM).

But IBM wasn’t done quite yet. They realised the clone makers were taking away valuable profits from IBM, and after their Compaq lawsuit largely failed to stop clone makers from clean-room reverse-engineering the BIOS, IBM decided to do something incredibly stupid: they developed an entirely new architecture that was entirely incompatible with the IBM PC: MCA, or the Microchannel Architecure, most famously used in IBM’s PS/2.

In the short run, IBM sold a lot of MCA-based machines due to the company’s large market share and dominance, but customers weren’t exactly happy. Software written for MCA-based machines would not work on IBM PC machines, and vice versa; existing investment in IBM PC software and hardware became useless, and investing in MCA would mean leaving behind a large, established customer base.

The real problem for IBM, however, came in the long run. Nine of the most prominent clone manufacturers realised the danger MCA could pose, and banded together to turn the IBM PC into a standard not controlled by IBM, the Extended Industry Standard Architecture (with IBM’s PC-AT of the IBM PC renamed to ISA), later superseded by Vesa Local Bus and PCI. Making MCA machines and hardware required paying hefty royalties to IBM, while making EISA/VLB/PCI machines was much cheaper, and didn’t tie you down to a single, large controlling competitor.

In the end, we all know what happened – MCA lost out big time, and IBM lost control over the market it helped create entirely. The clone makers and their successful struggle to break it free from IBM’s control has arguably contributed more to the massive amounts of innovation, rapid expansion of the market, and popularity and affordability of computers than anything else in computing history. If the dice of history had come up differently, and IBM had managed to retain or regain control over the IBM PC platform, we would have missed out on one of the biggest computing explosions prior to the arrival of the modern smartphone.

To circle back to the beginning of this article – using IBM’s fall from dominance in the market for consumer computers as proof that the market will take care of the abusive tech monopolists of today, at best betrays a deep lack of understanding of history, and at worst is an intentional attempt at misdirection to mislead readers.

Yes, IBM lost out in the marketplace because its competitors managed to produce better, faster, and cheaper machines – but the sole reason this competition could even unfold in the first place is because IBM inadvertently lost the control it had over the market. And this illustrates exactly why the abusive tech giants of today need to be strictly controlled, regulated, and possibly even broken up. IBM could only dream of the amount of control companies like Google, Apple, and Amazon have over their markets and users, and we are far, far beyond the point where the “market” can even dream of taking care of them.

There are other examples in technology that unfolded similarly to the story of IBM. The browser wars are sometimes cited as examples of tech giants not needing any regulations or oversight either, but just like how the clone makers could only beat IBM by breaking IBM’s control, browser makers – specifically, Mozilla – could only challenge Microsoft’s Internet Explorer because for all its countless faults, Windows is a fairly open platform, and Windows users were, at the very least, free to use and set as default any browser they wished.

Imagine if Microsoft had done an iOS, and had not allowed Windows users to use anything but Internet Explorer (or Internet Explorer skins). Do you honestly believe Mozilla would’ve stood a single chance in hell of breaking Internet Explorer’s stranglehold on the browser market?

There are countless examples from the past, in all kinds of industries, where monopolised or otherwise strangled markets had to be forcibly fixed by governments or the legal system. Take, for example, the strangehold the Wright Company and the Curtiss Company had on the airplane industry at the onset of World War I in the US – so tight was their control, that the production of new airplanes had all but stalled, and technological development in aviation in the US had come to a standstill. The United States government was forced to step in and force the industry to set up a patent pool every manufacturer was forced to join, thereby dislodging the market.

IBM lost its dominance exactly because the market became more open, less controlled, and easier to enter. By using IBM’s story as an argument, you’re arguing that we need more openness in the market, not less – and much like IBM 40 years ago, Apple, Google, Amazon, and others are not going to relinquish that control willingly.

History has taught us that if we want more innovation, more competition, and more choice, we need to curtail the big technology companies, and I think we’re well past the point where forced break-ups should be on the table.

Because I really wouldn’t want to live in a world where the IBM PC was the only game in town.


  1. 2021-07-17 5:55 pm
    • 2021-07-18 12:16 pm
  2. 2021-07-17 7:29 pm
    • 2021-07-17 8:14 pm
      • 2021-07-18 3:22 am
        • 2021-07-18 12:34 pm
        • 2021-07-18 8:47 pm
  3. 2021-07-17 7:54 pm
  4. 2021-07-17 8:15 pm
    • 2021-07-18 3:06 am
      • 2021-07-18 11:38 am
  5. 2021-07-18 3:58 am
  6. 2021-07-18 9:23 am
  7. 2021-07-18 11:25 am
    • 2021-07-18 7:11 pm
      • 2021-07-19 1:27 am
      • 2021-07-19 3:27 am
        • 2021-07-19 4:15 am
          • 2021-07-19 7:23 am
        • 2021-07-19 2:54 pm
          • 2021-07-19 3:21 pm
      • 2021-07-19 7:31 am
  8. 2021-07-18 6:32 pm
    • 2021-07-18 7:23 pm
  9. 2021-07-18 8:51 pm
    • 2021-07-19 3:38 am
      • 2021-07-19 3:48 pm
        • 2021-07-20 3:13 am
  10. 2021-07-19 12:14 am
  11. 2021-07-19 11:15 am