One of the phrases that give IT directors nightmares is “legacy systems”. These are old, old, old systems, some of which date back two decades or more, that companies invested heavily in, put all their data into, and are now stuck with. It’s cheaper to keep up the old system than it is to transfer it to a new system…even if the new system is a buggy, crashy nightmare that has to be somehow integrated with modern computers, which is kind of like figuring out how to stuff a Model A engine next to the power plant of a Ferrari.
There is one exception, though, and you unknowingly use it all the time: IBM’s OS/2.
Why? Simple, really: OS/2 doesn’t crash.
OS/2 is one of those oddities of computer history, an operating system invented by IBM back when IBM actually made consumer computers. And it’s stable because of how it was designed.
This is why your computer crashes, as a rule: a program somewhere, somehow, is doing something it shouldn’t. This is because your Windows box is what’s called an “open architecture” computer: components are swappable and upgradeable. Compare this to your Mac: it’s designed to run on computers you can’t even open without specialized tool kits. It’s possible to run OSX on a non-Apple PC, but it’s a pain in the ass.
The problem with open architecture is that this swappable and upgradable nature means that you have to write code for every possible component you can think of. This is why your printer freaks out when you first plug it into a new computer, and forces you to download and configure a driver. OS/2 doesn’t have these problems, largely because it quickly become a corporate OS, running checkout lanes and ATMs, and even New York’s Metrocard system.
That stability is valuable for a pretty simple reason: who wants to see the Blue Screen of Death when they’re trying to pull some cash out of the ATM.
In short, even in computers, just because something’s old doesn’t mean it doesn’t work.
25 Years of IBM’s OS/2 [Time]