Esoteric Operating Systems: RISC OS

Note: I hope this to be the first in a series of analyses of some of the more esoteric operating systems still in existence. This won't be particularly good as regards legibility, so I hope you won't judge me too hardly on that, for once.

RISC OS, or "Why the hell are people still making this?"

The story of RISC OS begins in 1987, with the British computer company, Acorn. The microcomputer revolution of the late 1970s and early 1980s had served Acorn Computers well, and where many of the competitors in the marketplace had fallen, Acorn still stood strong. The company had obtained a large amount of success with its BBC Micro, an educational computer which had been created in conjunction with the British Broadcasting Corporation's Computer Literacy project. However, with the success of Apple's Macintosh and the Commodore Amiga, Acorn moved towards their own graphical user interface, designing a computer to run their new operating system.

What they came up with was the Acorn Archimedes. The Archimedes was a very powerful computer for its time, using the then-new ARM processor architecture which produced far more calculation per megahertz than the competing Motorola 68000 used in Apple and Commodore's computers. Obviously, a useful operating system had to be built in order to take advantage of the power that the ARM processors provided.

Acorn came up with RISC OS. A co-operative multitasking operating system working from a GUI, it beat the efforts of Apple, who wouldn't move to preemptive multitasking until 2001, or even co-operative multitasking until 1988, and made a mockery of the nascent Windows operating system from Microsoft. Stored in ROM on the motherboard, it was also a very efficient operating system, quick to boot up and to respond.

However, despite a powerful computer and a relatively advanced operating system, Acorn would get nowhere near the amount of success with the Archimedes as it did with the BBC Micro. The x86 platform already had a large amount of the lucrative business market, resulting from years of IBM backing, and Apple had a more elegant design which was popular among designers and artists. Apart from the education sector, the Archimedes found little success, and ended up being relegated to a niche product.

By the time of the release of Windows 3.0, the future was looking increasingly promising for Microsoft. x86 PCs were becoming cheaper and cheaper, with a large amount of competition in the sector, while other computer companies, despite potentially superior computers, suffered from lack of recognition. By 1998, being one of the longest-surviving of the computer companies from the 1980s, Acorn closed down, beaten by a far more popular and far more powerful software platform from Microsoft.

But while the computer division of Acorn had failed at the hands of the all-sweeping Intel x86 platform, the business had more success in the processor division. The ARM processor had one great advantage over most of its competitors, something which made it extremely popular: It used far less power than its competitors. This was a feature which made - and makes - it particularly favoured in embedded systems, where low power consumption is imperative. Your mobile phone will almost undoubtedly use an ARM processor of some kind. Almost all MP3 players use ARM processors, as do the Game Boy Advance and DS handhelds. The processor architecture has found a home in most parts of the consumer electronics industry, from handheld devices to routers and hard drives. (Yes, I know this paragraph is terrible.)

With the success of the ARM processor architecture, it proved to be only a small task to reprogram RISC OS for the newer ARMs. A company named RISCOS Ltd. took up the task of maintaining the operating system for the newer computers, made by Castle Technology Ltd.

The question I have is: Why would you commercially resurrect such an old operating system? To be brutally honest, while RISC OS may have been effective and powerful for the 1980s, there's little to endear it to modern computer users. The OS still has co-operative multitasking, which proved itself obsolete in the mainstream market with the release of Windows 95, and in the computer science industry many years before that. It's tied into a very small line of computers from Acorn and Castle Technology, the former being long obsolete, and the latter being extremely expensive for what you get. Without the ability to implement file extensions, or spaces in file names, naming conventions are extremely odd.

But all of that pales in comparison to the absolute insanity of one the programming languages used in programming the operating system. Two of the languages are perfectly rational and sane: 46% is programmed in C, and 44% is programmed in ARM assembly language. These programming languages are consistent with other operating systems built on the ARM architecture; they aren't made to be portable across to personal computers. It's the third language which is madder than a hat-maker in a bathtub full of mercury: 10% of the operating system is written in BBC BASIC.

Yes, that's right - BASIC, a programming language most commonly known for its ubiquity on 1980s microcomputers and angering Edsger W. Djikstra. What's more, Acorn even originally programmed part of the GUI in it, and while I have to have respect for programmers that could program any sort of working graphical interface in BASIC, I feel strongly that it's the completely wrong language for operating system programming. All of these quirks and more contribute to an operating system which seems to be completely inadequate for modern tasks.

And yet, despite these apparent faults, RISC OS has one major factor in favour of its continued existence. It is probably the single most efficient operating system still being made for personal computers. Still provided on a 4MiB Flash ROM, RISC OS is still amazingly quick, and there's a reason why the operating system has so much assembly language content. In fact, it has to be one of the few operating systems where the more modern version actually runs faster than the older one on systems with the same hardware - and at this point, I point at Microsoft accusingly. Boot-ups in half a minute, immediate shutdowns, an amazing response all around.

The efficiency stretches to the applications as well, where each RISC OS application is far more efficient than its equivalent on Windows, Mac OS X, Linux or the other UNIX-based operating systems being used today.

On that alone, RISC OS deserves to be commended. In an era when operating systems are becoming larger and larger, this is an operating system that has improved its efficiency over the years. Startup takes about fifteen to twenty seconds, and as soon as the GUI is online, it's ready to accept input. Even a clean Windows XP install takes at least a minute to start up on modern technology, and the GUI doesn't immediately accept input. The same is true of Mac OS X and Linux. As an advocate of a quick-booting, quick response operating system for netbooks and educational purposes, RISC OS shows some characteristics which operating system manufacturers should take account of, even if the operating system is fundamentally outdated.

Thanks for that, I learned a lot of things I didn't know about RISC OS. I've never used it myself, but I remember seeing it running on a RISC PC around 1994/5, and the guy who was demonstrating it had video playing in a window, and was moving the window around with the mouse. It might not sound like much now; at the time seeing FMV in a window was pretty cool, and the way it continued to play as he moved the window, without a blip: I was amazed.

I had no idea about the BASIC thing, either. It seems like a crazy thing to do when writing an operating system. Thanks again, I look forward to reading more.

Wow! and I thought most Linux OSes were obscure.

Acorn were particularly well-known for powerful, advanced computers. Unlike the more popular Commodore 64s and ZX Spectrums, the BBC Micro actually had an operating system separate from its BASIC interpreter (the same one that they programmed part of RISC OS in), called Acorn MOS (Machine Operating System). There was also a special interface, the Tube, which allowed a second processor to be installed in a box to the side - including an 80286 which allowed for compatibility with MS-DOS, and a National Semiconductor 32016 which even allowed the BBC Micro to use a version of UNIX. That wasn't something you could do with your garden variety C64 or Spectrum.

Also, apparently, RISC OS was never intended to be the operating system of the Archimedes. Acorn had a development cycle much like Apple in those days, except they were more forward-looking and more ambitious. They had a UNIX-like operating system in the pipeline, like Sun Microsystems' SunOS and Silicon Graphics' IRIX. It was called ARX, but failed because of cost issues.

As mad as many of the components of RISC OS are, I have a feeling that Acorn Computers would have failed even quicker if they'd used a UNIX-like operating system on the Archimedes. That would have locked them into a completely workstation-based market, and probably wouldn't have even worked in the education market, not unless they wanted their users to learn the ridiculously truncated acronyms that make up most UNIX commands. As much as I like a UNIX-like interface myself, I know from experience that it isn't easy to get into.

I bought my new Archimedes 310 in 1988, brilliant machine, many fond memories.

wow that was shocking


Reply to Thread

This thread is locked