It’s great, if you’re a manufacturer, to feel like you’ve got control over a market. If you’re the only outfit making something, or at least if you control who can make that thing, you can minimise some of the traditional pains of technology manufacturing. Compatibility is less of an issue if you get to approve everything, you control the features and the documentation, and all of this adds up to fewer tech support calls, each of which costs you something to service. And, of course, if you’re licensing the design, you get to make money on things built by other people.
We’ve long become used to being able to pick up parts for computer systems from more or less any manufacturer, throw them in a box, and have the result be a useful machine. While Apple are still using Intel processors – and even, potentially, if they stop doing that – you can buy upgrades for any computer from your favourite Google result. Take notice of only a few fairly straightforward compatibility codes, throw them in the right slots, and mainly, they’ll work. That’s now so normal that it’s easy to overlook what a staggering feat of standardisation it represents.
It had always been IBM’s intention that the original PC, from which we derive almost all modern personal computers (including Apple Macs) should be an open platform. The system was designed, in mere weeks, by people who were computer enthusiasts at a time when being a computer enthusiast meant knowing how to solder, and it would have been natural for them to want to create something that other people could add to. The first PC, IBM’s famous model 5150, had five expansion slots. The company published excellent, low-level technical documentation, with the idea that industry would create an ecosystem of compatible devices. Industry enthusiastically did, and the result, via a few chicanes, was the fantastic selection of computer hardware that we now enjoy.
Arguably, the ferocity of competition in computer hardware has led to a lot of the advances we’ve seen since IBM’s 1981 great. Much of this has only been possible because, on paper, anyone can make anything and compete with anybody. OK, if you want to build a device and put the USB logo on it in 2021, you’ll find there are a few hoops to jump through, but it’s doable by even fairly small companies. It’s not clear whether we’d have seen the same meteoric improvements in graphics cards – which effectively created the modern approach to colour grading – and CPUs without competition between Nvidia and AMD, and Intel and, er, AMD.
And now, to swerve this discussion abruptly back on course, we have raw camera formats, most of which are highly proprietary, at least in their most popular incarnations. Keeping them that way allows the companies which developed them to ensure compatibility, certifying implementations and making sure everything works in a consistent and reliable manner, bringing untold benefits that are (in the words of many) greatly valued by users. It also creates a kind of (multi-) vendor lockin which stifles innovation, obviates at least some benefits of a free market, complicates workflows, and in the end leads to raw material recorded by the world’s best-known line of on-camera recorders being incompatible with the world’s best-known grading software.
It would probably have been quite an attractive idea for IBM to keep control of the PC ecosystem and charge licensing fees, and in fact the company did intend to retain the sole option to build the base system. By dint of huge effort, Compaq was able to break down the walls of that garden by reverse-engineering the system’s firmware in a way that would stand up to legal scrutiny, and as a result Compaq made not only the first PC compatible but also the first portable PC. Whether the opening-up of base PC systems was essential to the creation of such a vibrant ecosystem is lost to history at this point. Certainly there was already signs of a healthy market in PC hardware before Compaq’s little coup d’etat, which at least shows people like that sort of thing.
It’s not clear whether the modern computer hardware market would have existed without IBM’s speculative approach, but there are a lot of other examples; we know that cellphone providers like to lock handsets to their system, much as many jurisdictions have begun to oppose the practice as anticompetitive. Either way, in the meantime, the film and TV industry has more or less sleepwalked into a situation where we routinely record high-value moving images in any one of a few ways that are very much locked in, even deliberately encrypted.
Given the fierce rivalry between the incumbents, appealing to their better nature seems like a long shot, so we might just have to wait for someone to win.