RAK’s OS Adventures Double Bill – Atari TOS & SymbOS

Author’s Note – Since I didn’t write an article two weeks ago, I’ve decided to try to make up for that somewhat by writing a set of short articles on a couple of operating systems that I’ve recently acquainted myself with. Hope you enjoy!

Historical Operating Systems – Atari TOS

As I’ve indicated before ([1], [2], [3]), the mid-1980s transition from 8-bit to 16-bit processors in personal computers marked a turning point, from computers which would boot up into a BASIC interpreter to ones which would incorporate full graphical user interfaces, and towards development of multimedia programs such as WYSIWYG word processors, desktop publishing and graphics manipulation suites. A lot of the competitors in the preceding 8-bit computer wars had been lost to market pressures, leaving only a few primarily American and British competitors to fight it out in the late 1980s.

Atari was one of these competitors. Atari was one of the few console-producing companies to survive the Great Video Game Crash of 1983, despite two games, the poor-quality port of Pac-Man with its flickering sprites and the turgid, over-produced E. T. the Extra-Terrestrial, on the Atari 2600 console being partially responsible for the crash. Nevertheless, the crash of the video game console market led Atari’s parent company, Warner Bros., to sell off Atari.

Atari’s computer division was responsible for the Atari 400/800 computers which had been strong competitors against the Apple II and Tandy Radio Shack TRS-80 computers, and had maintained sales against the Commodore 64. With strong sales in Europe, where the Commodore 64 had a harder time competing against the low-price British computers such as the Sinclair ZX Spectrum, this part of Atari’s operations was still desirable.

One man to whom Atari looked particularly desirable was Jack Tramiel. Tramiel had founded Commodore International in the 1950s, first repairing and selling typewriters, then moving onto calculators, but soon branching into the nascent computer market. Commodore’s PET and VIC-20 models had been highly successful, but it was the Commodore 64 which would forever cement the company’s name in computer history. However, Tramiel’s market tactics involved quickly dropping the price of the Commodore 64, effectively underpricing every other computer in the North American market. Tramiel’s tactics would end up making the Commodore 64 the single-most successful home computer model ever produced and wiping out many of the competitors.

It has been suggested that one of Tramiel’s main targets was Texas Instruments, who had almost ruined Commodore during the 1970s, and the continuous price-dropping of the Commodore 64 did end up wiping Texas Instruments from the home computer market in the process. Nevertheless, this aggressive price strategy did come at the cost of profits, and in 1983, Jack Tramiel was unceremoniously booted from the Commodore company.

All of this made Atari a rather tempting acquisition for Tramiel, who bought Atari in 1984. Atari originally had their eyes on the Amiga, but after losing a protracted battle with Commodore for the Amiga chipset after Tramiel’s takeover of the company, Atari had to develop a replacement. The computer they developed was the Atari ST, or Sixteen/Thirty-Two (a reference to the Motorola 68000’s 16-bit external bus and 32-bit internal architecture). Arriving on the market two months before the Amiga, the Atari ST would be competitive in the late-1980s, particularly in Europe, where the home computer market moved more slowly towards the adoption of IBM PC-compatibles, but as with the Amiga and most other competitors, lose ground into the 1990s.

The operating system of the Atari ST was a licensed version of the Digital Research GEM environment, known as Atari TOS. Originally shipped on floppy disc, it was later incorporated onto a ROM chip, similar to the contemporary RISC OS for the Acorn Archimedes. Unfortunately, though, Atari TOS feels like the most sparse of the operating systems of the late-1980s 16/32-bit microcomputers, even considering that it is on a ROM chip rather than the floppy discs used for most other contemporary operating systems. While RISC OS had a powerful BBC BASIC interpreter that could be used to build graphical applications, and AmigaOS and Mac OS at least came with calculators or clocks, Atari TOS really only came with the basic OS utilities and not much more.


At least graphically, Atari TOS matched up to early versions of AmigaOS and RISC OS – although these graphics would remain into the 1990s.

What’s more, Atari TOS was a single-tasking operating system right until the shipping of the MiNT-based MultiTOS package in the 1990s; this at a time when AmigaOS had possessed pre-emptive multitasking since 1985, when Mac OS had got cooperative multitasking in 1987, when RISC OS had cooperative multitasking since its release and when Microsoft Windows was giving even the x86-based MS-DOS computers multitasking capabilities. Given that I have complained about the slow adoption of multitasking by Mac OS and the cooperative multitasking model of RISC OS, I think it stands to reason that I do not regard Atari TOS as an especially sophisticated operating system, even by the standards of the 1980s.

It’s a shame, in a way, since the Atari ST itself seems to have been a sophisticated, yet inexpensive computer that could compete well against the Macintosh and Amiga in terms of pure grunt. Plenty of games seem to have been cross-developed between the Amiga and Atari ST, the common element of a Motorola 68000 microprocessor helping matters. The Atari ST was also the first platform for which the Allegro game library was developed (hence the name of the library, originally derived from Atari Low-Level Game Routines). The operating system, on the other hand, did not match up to the sophistication of the computer it ran on, and while Mac OS and the early versions of Microsoft Windows led to greater development down the line, AmigaOS was a sophisticated operating system ahead of its time and RISC OS is still being developed today for a processor architecture which is very much alive, Atari TOS doesn’t really seem to have any particular historical significance beyond the fact that it was there.

RAK’s OS Adventures – SymbOS

From there, I move onto an operating system which is very much more sophisticated than its hardware requirements would indicate. The 8-bit computer wars of the early 1980s typically revolved around two types of processors. The MOS Technology 6502 was used in the Commodore 64, the Apple II and the BBC Micro, among others, and could indirectly trace its lineage to the Motorola 6800. The Zilog Z80, on the other hand, was used in the Tandy Radio Shack TRS-80, Sinclair ZX Spectrum, the Amstrad CPC and PCW ranges and the MSX computers and uses an extended version of the instruction set architecture of the Intel 8080.

In certain regards, the MOS 6502 is a superior processor; while Z80 processors typically ran at a clock rate of approximately 4MHz in most early-1980s microcomputers, MOS 6502 processors could be competitive at a clock rate four times slower than that. In other regards, the Z80 has the advantage; with more registers and a more sophisticated instruction set, it is a more friendly processor to program for. Both processors are still being sold today, mostly for embedded systems; Zilog still exists and has a strong market selling Z80 and Z80-compatible processors for embedded systems, some quite a bit more powerful than the 1980s models (RS Components currently has a Zilog Z180 running at 20MHz for €11.40 for an individual processor), while Western Design Center, established by a co-holder of the 6502 patent, still sells 6502-compatible processors.

No matter how popular these processors may have been in their day, they seem rather old-hat even by the standards of the Intel 80286, the Motorola 68000 or the ARM2, let alone today’s multi-gigahertz Intel processors. Nevertheless, sometimes sophisticated programming can trump raw power, as seen with SymbOS, an operating system for Z80-based computers.

SymbOS is the project of a German programmer, known by the pseudonym Prodatron, and is developed for the Z80-based Amstrad CPC, MSX and Amstrad PCW computers. The Amstrad computers were particularly popular in Germany, where they sold in numbers comparable to the Commodore 64 in that market, while the MSX computers were developed as a standard for Japanese microcomputers. These computers typically came with 64 or 128kB of RAM and Z80 processors clocked at 4MHz – only moderately powerful for 1984 or 1985, when these computers were released.

All of that makes SymbOS look especially impressive when considering what you get out of it. SymbOS provides full pre-emptive multitasking, of the same level of sophistication as AmigaOS, developed for a computer with twice as much RAM as standard and a processor a generation older. It provides a full GUI system, similar in appearance to that of Microsoft Windows 95, and similar in sophistication to Atari TOS or Mac System 6 – and more sophisticated than Microsoft Windows 1.0. It can run sound and video applications. It can support 128GB FAT32 hard drives – when Microsoft Windows couldn’t do this until Windows 98.


It even looks pretty nifty as well.

Yes, the OS was first released in 2006, twenty years after the release of the computers it runs on. Yes, it’s not the only graphical user interface for an 8-bit computer – GEM was originally developed for the Commodore 64. But it’s got pre-emptive multitasking, sound and video capabilities and support for large hard drives, all on an 8-bit processor originally developed in the 1970s, and on as little as 128kB of RAM (although more is suggested for best performance). It makes me wonder – if this sort of operating system can be developed for an 8-bit computer, would it be possible to improve some of the 32/64-bit operating systems of today if there were generally more proficient programmers developing them?

Making Hardtack: A Random Interlude

Author’s Note: I recognise that my update schedule has been somewhat disrupted in the last month. I wasn’t even especially busy, but I lacked motivation and sufficient material for a post. This post is itself blatant filler, but indicates what I’ve been up to recently. When I get back to my technical interests, I do intend to return to my usual update schedule.

I have rarely sought to limit myself to a small range of interests. Aside from the obvious technological, scientific and mechanical pursuits that I regularly talk about on this site, I have also taken forays into subjects rather more diverse, such as politics and history, literature, cinema and art. One subject which I have perpetually taken an interest in is militaria, from the weaponry and equipment used by soldiers to the tactics used to win a war.

One of the most important aspects of successfully fighting a war, and one often glossed over by many media sources, is logistics; namely, making sure that your army has sufficient supplies to sustain itself, to fight and to manoeuvre. Theatres of war have been won or lost over logistics, the most famous examples being Napoleon’s invasion of Russia crashing to a halt as the Russian scorched earth strategy worked to deny La Grande Armée the resources of its conquered lands, and the subsequent invasion of the Soviet Union during the Second World War by the German military which was halted at Leningrad, Stalingrad and Moscow by similar scorched-earth tactics.

In a modern combined-arms military, logistics is crucial to ensure that all of the components of an army can move in concord. That doesn’t mean that logistics wasn’t important in the past, and one of the most critical parts of logistics now and in the past is ensuring that all personnel have enough food. As has often been expressed, “an army marches on its stomach”. However, military operations come with their own particular set of challenges when it comes to providing food. One of the chief challenges is food preservation, problems arising from matters such as the general deficiency of refrigeration in a modern army and the complete lack of refrigeration in the past, to the conditions that an army may be fighting in, such as fly-infested swamps and humid jungles.

Correspondingly, military rations often have to be designed with these limitations and challenges in mind. A type of military ration which persisted for many centuries up until the modern day was hardtack, a type of hard-baked biscuit or cracker designed to be extremely resistant to spoiling and resistant to breakage. Unfortunately, the very steps taken to ensure that the hardtack crackers resisted spoiling also made them unappetising and difficult to eat, especially when eaten for long periods of time, as was regularly the case before modern ration design starting with the C-rations of the United States during the Second World War and continuing today with the MRE (Meal, Ready to Eat) rations of the present-day United States military.

I was interested in seeing what these hardtack biscuits would actually taste like, and so I decided to bake a batch of them. The recipe that I used dates back to the American Civil War, although similar rations were reportedly in use with the British Expeditionary Force in the First World War and beyond. The recipe is extraordinarily simple, consisting of nothing else but flour, water and a small addition of salt for flavour. The simplicity of the recipe might have itself contributed to the popularity of hardtack with logisticians of the past; it is very easy to scale up the production of hardtack, even without machine tools or anything more sophisticated than a good oven and a few simple kitchen utensils.

The recipe is as follows:

420g/≈3 cups plain flour

250ml/≈1 cup water

2 teaspoons salt

1) Preheat the oven at a temperature of 180°C/350°F.

2) Add the water and salt to a bowl.

3) Add the flour in small quantities, stirring the flour into the water-salt mixture. As the flour is added, a thick paste will be produced. Keep adding flour until you cannot stir the mixture any more.

4) Knead the paste, which should be slightly sticky to the touch. If the mixture continues to be sticky after kneading, add more flour until you have an elastic, but smooth mixture remaining in the bowl.

5) Place the mixture onto a flat surface and roll out into a flat sheet approximately 1cm/½in thick.

6) Using a pizza cutter, cut the sheet into squares of dimensions approximately 7cmx7cm/3inx3in. Using a skewer, poke holes in the top of the squares in a grid of four holes on each side equally spaced. [See below for picture indicating how this should look.]

7) Place the squares on an ungreased cookie sheet. Place into the oven and bake for 30 minutes. Flip the squares over and bake for another 30 minutes.

8) [OPTIONAL] For enhanced preservation, continue to bake squares at 120°C/250°F for an additional 30 minutes.

The picture below shows two batches of hardtack that resulted from this recipe. The front nine biscuits were a preliminary first batch, which as can be seen are rather uneven in size and thickness. Most of these biscuits were too thick, and the result was that while the external shell of the biscuits was as it should have been, their centres more closely resembled a bread roll like a baguette; this may well be more appetising than the harder shell, but would be less inclined towards proper preservation, as fungal moulds could thrive on the moisture still contained in the mixture.


The back nine biscuits, on the other hand, were more evenly sized, had the proper thickness and altogether more closely resembled what hardtack was meant to be like than the first batch. The most obvious characteristic of the second batch was that the purported hardness of the recipe was not exaggerated by the sources I had used. A soldier with poor dentition would find it difficult, if not impossible, to eat a hardtack biscuit without resorting to things such as leaving them to soak in water or coffee, a process which would itself take quite some time, as the hard shell resists water penetration to the slightly softer core. It is also extraordinarily dangerous to attempt to eat hardtack with one’s incisors; using the premolars or molars is suggested.

Once I got past the preliminary difficulty of actually eating the hardtack, I did notice that the flavour wasn’t unpleasant. It tasted like a water cracker, except with a harder outer shell, and would be a sufficient and not-unpleasant sort of food for a short period, such as a modern-day hike. It is, however, a monotonous sort of food, which would be rather less pleasant over the weeks that a soldier of the First World War or American Civil War, for instance, may be forced to eat it, let alone the consistency that would be expected on a naval journey during the time of Admiral Nelson. As survival food for somebody conducting a short activity outdoors, it would be sufficient even today, but as food for a working soldier, well, let’s just say that I’d take an MRE or its equivalent any day of the week!