In the following text I describe some happy memories of having been a young home computer user in the 1980s. Background information about the companies and the people who created them have been added.
The text may give an impression, what these machines were about and how they were used, in case you want to emulate them on today's computers.
But also, like Coldmirror said: The site is intended to remember, to chuckle, to fall asleep, to be read on long journeys or while baking a cake. Ideally all of these together at once.
The text also helped me to understand better what was going on when I was young, how things developed to today's condition, and why those machines, that were on everybody's mind back then, aren't of any importance today and are even hardly known any more.
Back then, I had these computers:
So I can tell most about these. Much later I also had an "Amiga 1200" and an "Apple Macintosh Classic" as retro-computers.
I don't have any of these machines as hardware any more. I still like them though.
Most home computers came with an external power supply. I think, it's not necessary, to mention that each time.
It wasn't necessary to cool home computers down. So they didn't have internal fans. Most of them didn't have hard disks either. Therefore, they operated in nearly absolute silence. I really miss that in modern computing.
The general history of home computers started in 1976 with the Apple I. Which didn't even have a case and was hardly usable. Then, in 1977, there was the first version of the Apple II, the Commodore PET 2001 and the Tandy TRS-80. In 1979, also the Atari 800 and 400 entered the market. But that was something, that only happened far away in the USA, these machines didn't become popular in Germany.
The industry's next step was to make home computers much cheaper, so Commodore released the VIC-20 in 1980, and Sinclair the ZX81 in 1981. I once saw a VC-20 (which was the VIC-20's name in Germany) at a friend, but I wasn't too impressed by it, with its big keyboard, but its just 3.5K of memory. And the Sinclair ZX81, which I only encountered later in the 80s, again felt hardly usable to me.
So, the story of home computers may have started earlier in the 1970s, but to me it only began, when the Sinclair ZX Spectrum became available in 1982.
This is not a scientific text, so there can be contentual mistakes. If you find any, please feel free to tell me by email.
The Sinclair ZX Spectrum was my first computer. It was released in 1982 by "Sinclair Research Limited", a British company founded by Sir Clive Sinclair (1940-2021).
The ZX Spectrum used a "Zilog Z80A" 8-bit-CPU as its processor, running at 3.5 MHz. You could buy a Spectrum with either 16K or 48K or RAM.
According to a price-list of the trading company "Vobis", in July 1983 prices were 415,- DM for the Spectrum 16K, and 539,- DM for the Spectrum 48K.
Prices on another "Vobis"-advert, that I had seen in 1983, were higher though: 498,- DM for the 16K version, and 725,- DM for the 48K model. Home computer prices weren't static, but varied quite a lot, it seems.
Besides the Z80A, there was a custom chip inside the Spectrum called ULA ("Uncommitted Logic Array"), that controlled the input and output (I/O) of the keyboard, of audio and tape and generated (together with an UHF modulator (RF modulator)) the display on a television, that had to be connected to the computer.
The Spectrum had rubber-keys. Each key had multiple meanings, which were written onto the keyboard. There were two different shift-keys: "Caps Shift" and "Symbol Shift". When you started the Spectrum, it showed in black letters on a "white", or rather kind of greyish screen:
© 1982 Sinclair Research Ltd
When you pressed a key, you entered the Spectrum's BASIC environment. The cursor turned into "", meaning, that you could enter a keyword (written onto each key) now. To load a program (from cassette), you had to execute the command 'LOAD ""' on the Spectrum.
When you pressed "j" in keyword-mode, it said "LOAD " with one key-stroke (the key "l" was reserved for the "LET"-command). The cursor then entered letter-mode and therefore turned into "". Then you pressed "Symbol Shift+p" twice to get the quotation marks, and finally had:
LOAD ""
Then you pressed "Enter" and were done. The Spectrum then waited for the cassette player to be started (by you).
It was not possible to type in "LOAD" as "L", "O", "A", "D", you had to use the special behaviour of the keys and the differrent key-modes.
Other key-modes were:
After a while you got used to these different keyboard-modes, that may seem strange at first.
Actually, the keyword-mode had the advantage, that keywords could be stored in memory as a single byte, instead of let's say 5 bytes for a 5-character-command like "LOAD ".
When you entered a valid command on the command-line, the Spectrum showed:
0 OK, 0:1
which always was somehow reassuring.
The cursor could be moved by pressing "Caps Shift" plus "5", "6", "7" or "8" (left, down, up, right).
The Spectrum had a screen-resolution of 256 x 192 pixels. This resolution could be accessed by machine code programs like games.
But when programming in BASIC, only a resolution of only 256 x 176 pixels was used. Because the last 16 pixel lines (or two character lines) were used to display the BASIC command line and error messages.
In the Spectrum's BASIC, there was a PLOT-command to set a pixel with a x-position from 0 to 255 and a y-position from 0 from 175. There were commands like DRAW or even CIRCLE, too.
There was also a "resolution" for characters (because a character needed a field of 8x8-pixels) of 32 x 22, so you could write
PRINT AT 10,10;"H"
to print a character somewhere onto the screen. It seems, the Spectrum didn't have a special text-mode. Instead also characters were rendered in graphics-mode. That was probably the reason, you could do a "PRINT AT", which wasn't possible on most other home computers.
The computer was called "Spectrum", because (unlike the ZX81) it featured colours. Exactly eight of them, that is:
Black, Blue, Red, Magenta, Green, Cyan, Yellow, White.
These colours could be assigned to the 32x22-resolution, but not to the 256x192-resolution. So you could print a character "A" in, let's say, a yellow foreground with a magenta background. And the next character in two different colours.
But you could only plot two pixels next to each other in different colours, if they were in different character-fields.
The Spectrum's screen was surrounded by a well-spaced border, whose colour could be set (to one of the eight available) with the BORDER-command. When the Spectrum was loading software from tape, the border used to flicker in characteristic colours according to the noise coming from the tape. You could see by the quality of the flickering, if the data transfer probably would be successful.
The typical storage device for the ZX Spectrum was an audio-cassette tape recorder. Originally, Sinclair didn't produce a dedicated one, but you were supposed to buy a general-purpose cassette recorder, that was easily available at the time. The Spectrum had an input (called "EAR") and an output (called "MIC"), that were connected to the cassette recorder's headphone-output and microphone-input. Standard 3.5 mm audio jacks were used. The Spectrum came with a corresponding cable.
To load or save a program of the full 48K took about 5 minutes.
The Spectrum didn't have a sound-card or sound-chip. It just had a little speaker, that was stimulated directly by the CPU to produce some kind of beeping noise. The speaker had a pleasant volume-level though, unlike the speakers built into PCs.
As the CPU was used to produce sound with the speaker, the whole computer was frozen, while the sound played.
Therefore, game-programmers used a lot of tricks to produce very short sounds, that wouldn't stop the game too long from playing. Nevertheless it was quite amazing, what results programmers were able to achieve with sound under these circumstances.
It is said, that the Spectrum also sent its sound to its "MIC"-output, from where it could have been amplified or recorded. But I never felt the need to do that, as the Spectrum's sound wasn't that exciting anyway. The tiny built-in speaker was perfectly sufficient for it.
The Spectrum came in a black carton box, on which the computer with its rubber-keys was depicted and the words "Personal Computer" were confidentially written. Inside was the computer, the power supply, the cables to connect to television and tape recorder, a shorter manual, an informative book about BASIC programmming and a tape cassette with demonstration-programs called "Horizons". One of the programs on this tape introduced you to the hardware of the Spectrum (like briefly described above):
When you switched on the Spectrum, you got into its BASIC-environment (loaded from the ROM). The Spectrum's BASIC was quite a capable language. Though, of course, programs in that language didn't run that fast. Nevertheless, games like "Pimania", "Mined Out" or even the "Monopoly"-implementation "Go to Jail" showed, what was possible.
(Actually, the BASIC game "Go to Jail" by "Automata UK", was originally called "Automonopoli". But the producers of the (analogue) board game "Monopoly" took legal actions against Automata for releasing their game for the computer. Automata's program was renamed to "Go to Jail", before it was withdrawn from the market by court order in 1984. In 1985, an officially licensed game called "Monopoly" by a company called "Leisure Genius" was released for the ZX Spectrum and other computers. It was not written in BASIC though.)
If you wanted to go further than BASIC, the Spectrum also made it possible for the user, to directly write Z80-machine code.
To do that, you had to reserve a suitable memory-area for your program first:
CLEAR 32767
Then, you poked the numbers containing your machine code into the memory addresses. The number 201 was the command to just return from a machine code program, for example. So let's poke that in:
POKE 32768, 201
Check, if it's there:
PRINT PEEK 32768
Every command including the word "USR" ran a machine code-program at a given address. You could use "PRINT USR 32768" or "RUN USR 32768". Or you could use:
RANDOMIZE USR 32768
RANDOMIZE is a BASIC-command to prepare the random-generator. It is totally useless in regard to the machine code program. It just does more or less nothing, and the "USR" does the rest. Altogether, a simple BASIC-program to get into machine-code and back could look like this:
10 CLEAR 32767 20 POKE 32768, 201 30 RANDOMIZE USR 32768
Writing assembler or machine code was as difficult as it always was, but at least the system didn't get in your way, if you wanted to to go that route. With BASIC-commands like CLEAR, POKE and USR, you could say, the system even invited you to try it.
The memory-area from 16384 to 23295 held the Spectrum's screen-buffer and the colour attributes. You could poke directly into that area to set pixels or colours. The command
SAVE "name" SCREEN$saved that memory-area to cassette tape. Many games used beautiful loading screens to make the long loading-times less boring. The image's data was directly loaded into this memory-area using a 'LOAD "" SCREEN$'-command.
The Z80A-processor could address 64K of memory altogether. The first 16K held the Spectrum's ROM. After that, there were the 48K of RAM (of the Spectrum 48K). The RAM was separated into different parts. First, there was the screen buffer and colour attributes. Then there was an area for the printer buffer and the system variables. After 10 more reserved bytes, there was the program-area of about 40K. At the top of the RAM there were another 168 reserved bytes.
So, effectively there were about 40K for the programs, be it BASIC- or machine code-programs.
At the rear of the Spectrum was an expansion slot called "edge connector". External interfaces could be directly connected to a dedicated part of the main circuit board there. Printers like the (weird) ZX Printer or the (better, but more expensive) "Seikosha GP-50s" could be connected there. You had to be careful, not to connect or disconnect expansions, while the computer was running. Otherwise there was a risk to break it.
Originally, the Spectrum didn't come with a joystick port. So most games could be controlled with the keyboard.
When companies realized, that the Spectrum could be used as a gaming platform, several third party companies like "Kempston", "Protek" or "AGF" built joystick-interfaces, that were plugged into the edge connector and added a single joystick-port. Then, Sinclair released its own joystick-interface called "ZX Interface 2". It offered even two joystick-ports and a slot for ROM-cartridges (but only very few of these game-cartridges were ever produced).
The joystick-interfaces were mapped to the Spectrum's keys as follows:
| • Protek / AGF | Cursor-keys, that is "5678" and "0" for "fire". |
| • Sinclair "ZX Interface 2" | "12345" (Player 2) and "67890" (Player 1). |
| • Kempston | Not mapped to any keys. Instead, I/O-port 31 was used. From BASIC, it could be read with the command "LET j = IN 31". |
| • Dk'Tronics | Joystick 1: "67890" (like ZX Interface 2). Joystick 2: I/O-port 31 (like Kempston). |
| • Fuller ("Fuller Audio Box") | I/O-port 127. |
So the Kempston-interface and the "ZX Interface 2" only worked, if they were explicitly supported. Many games supported the Kempston-interface, but not so many the "ZX Interface 2".
With the interfaces by Protek and AGF, you had a chance, that games supported the cursor keys for keyboard control anyway. Then you could use the joystick instead. Nevertheless, the Kempston-interface seemed to have been the most popular one.
Another expansion for the edge connector was the "Cheetah SpecDrum", which (together with an included software) turned the Spectrum into a little drum-machine. The sound wasn't so bad for the time, it sounded a bit like a lo-fi-version of the (professional, high-end) LinnDrum. But don't expect too much, either. Unfortunately, I never got it to work. Instead, the interface broke my Spectrum permanently. Maybe I just was unlucky.
A lot of games were written fo the ZX Spectrum. With such limited memory and without any sound card or sprites, some couldn't believe, that these games could be good.
But once a friend of mine, who had told his parents, he would come back early in the afternoon, returned six hours late in the evening, because we had been playing "Jet Set Willy" all the time. So you can't deny, it had to be a really good game ("Jet Set Willy" cassette cover).
I have another great memory about the adventure-game "The Hobbit" by "Melbourne House", which is another Spectrum classic ("The Hobbit" cassette cover).
I had bought the official cassette, which was rather expensive: 80,- DM, I remember. But included in the game box was also a paperback edition of Tolkien's novel, which was a rather nice idea. I played the game quite often (and meanwhile listened to the Beatles' LP "Abbey Road", as the game doesn't have any sound). Although "The Hobbit" is sometimes a bit weird, because such a huge amount of content had to be fit into such little memory, I still think, it's a great game. The typical situation happened, that I was stuck at a certain point. I couldn't get over it for a long time, maybe a year or two. (Searching the internet for the answer or downloading a walkthrough was not possible at the time. You were just stuck for a few years, unless you asked someone at a computer club or so, but that wasn't worth the effort.) One day, I visited a friend of mine who had "The Hobbit" on his C64. He was stuck at the same point in the game as I was. Together, we somehow managed to solve that certain puzzle and finished the game. I remember that experience until today, so I think, I will probably never forget it.
Some say, the ZX Spectrum used to develop heat during operation, which in the long run could possibly harm the circuits. I don't remember that. I didn't leave the Spectrum on too long anyway. In the afternoon, after I had done my homework, I switched it on, loaded a game from cassette, played a bit, maybe wrote some BASIC, and switched it off again. That usually took about an hour maybe. Then I had to learn for a test the next day, or I went out to meet some friends.
Playing "Jet Set Willy" for six hours definitely was an exception.
The Spectrum wasn't a product for eternity anyway, as in the years after 1982 computer evolution went on rapidly.
In 1983, Sinclair released the "ZX Interface 1". When it was connected to the edge connector, it lifted up the whole back-side of the Spectrum, so that the computer became a bit inclined.
It extended the Spectrum's ROM ("Read Only Memory"). It featured a RS-232 port and provided networking facilities, that allowed up to 255 Spectrums to directly communicate with each other. This feature, which sure was quite ahead of its time, was probably meant for networking in a school and wasn't that interesting for the ordinary home user.
The "Interface 1" also allowed to connect a "ZX Microdrive" (even up to eight of them) to the Spectrum. The ZX Microdrive was supposed to be something similar to a floppy disk-drive. But instead of floppy disks, it used weird small special cartridges with continous tape inside. A cartridge could hold about 85K of information. It took about 7 seconds to run one loop of the tape. So usually it took 7 seconds for the Microdrive to load or save a program, be it 1K or 48K in size. Unless reading or writing the data failed in the first loop. Then more loops had to be run, and access-time increased to 14, 21 or more seconds.
Nevertheless, compared to the 5 minutes of loading-time from cassette tape, the 7 seconds definitely were an improvement.
The "Interface 1" implemented new commands for the Microdrive inside the Spectrum's ROM. To check the content of the Microdrive (drive 1), you'd write:
CAT 1
("CAT" for "catalogue"). To load program "Hello", you'd write:
LOAD *"m";1;"Hello"
- which was the weirdest command I've ever seen. You had to press a lot of "Symbol Shift"-combinations to write it. Similar for saving:
SAVE *"m";1;"Hello"
Or for deleting:
ERASE "m";1;"Hello"
At first use, Microdrive-cartridges had to be formatted:
FORMAT "m";1;"Hello"
All of these commands were already written onto the Spectrum's keyboard, so Sinclair had already planned this, when the Spectrum was first released in 1982. I remember, the price for the "Interface 1" was 300,- DM at the time, and for one "ZX Microdrive" also 300,- DM. Together, that was quite an amount of money, to get a supposedly better storage-device (compared to the cassette recorder).
In 1984, 1985 there were the ZX Spectrum
In 1986 and 1987, Amstrad, that had bought the computer product range of Sinclair Research Limited released the ZX Spectrum
The later models fixed many of the issues, that had been generally criticized before. Compared to the original ZX Spectrum of 1982, the Amstrad models finally had these features:
To a certain extent, these features turned the Amstrad Spectrums into different machines. It could be, that the original Spectrum was meant to have these features too, but it hadn't yet been possible to produce such a machine at low cost in 1982.
About the audio sockets of the Amstrad Spectrums: While the Spectrum 48K had two audio sockets for input ("EAR") and output ("MIC"), the +2, +2A and +3 models had only one audio socket. The one of the +3 could be used for input and output using a special stereo cable.
But the audio socket on the +2 and +2A was only meant for sound output. So you couldn't just connect an external cassette recorder (or today a smartphone) to these computers, if the internal recorder was somehow broken. There are tutorials to make the +2/+2A receive external audio by modifying their hardware ( +2, +2A; also see this discussion and this video ). Without making changes to the hardware, it is also possible to use a socalled "cassette tape adapter" in the internal recorder of the +2/+2A. That is a strange device with a special cassette, that makes it possible to feed in external audio. It was normally used to connect a mobile CD-player to the cassette player in a car.
From 1989 to 1992, the British company "Miles Gordon Technology" ("MGT") produced a home computer called "SAM Coupé". It had a typewriter keyboard, used a Z80B CPU running at 6 MHz and came with 256K RAM. The SAM Coupé had its own system with graphics, sound and SAM BASIC. It is mentioned here, because it was also compatible to a 48K Sinclair ZX Spectrum.
About 12.000 units of the SAM Coupé were sold, so it remained one of the more exotic home computers.
In the last month of 1989, when already my Atari 800XL days were over and I had been happy with my Amiga 500 for two years, the company "Codemasters" released a "CD Games Pack" for the ZX Spectrum, which was a compilation of 30 games on an Audio-CD. To use it, the audio output of a CD-player had to be connected to the Spectrum's joystick interface ("Kempston" or "Sinclair Interface 2"), using a special cable that came with the CD. Then, a small special loader program had to be loaded from tape in the conventional way. After that, games could be loaded into the Spectrum by selecting one of the audio tracks on the CD. As audio on a CD was much more reliable than tape audio, it had been possible to heavily compress the audio tracks on the CD. Therefore it took less than a minute to load a game from the CD player, instead of about 5 minutes when loading it from tape. That was a remarkable technical progress, but it just remained an oddity.
Today, there's even a SD card reader for the ZX Spectrum called "DivMMC Enjoy! Pro One".
An emulator of the ZX Spectrum is called "fuse". The owner of the copyright of the Spectrum's ROMs since 1986, Amstrad, deserves credits for making this code available for legal download and use in emulators, as stated by an Amstrad official in this message.
The machine is very well documented at "worldofspectrum.org".
Also, more than hundred episodes of "The Spectrum Show" can be watched on Youtube.
Here's some more background infomation about Sir Clive Sinclair and his company:
Clive Sinclair learnt about electronics already in his early years, and published several books about the subject. His father was a mechanical engineer and had his own company. Since the time Clive Sinclair grew up in Britain in the 1950s, he wanted to invent, develop, produce and sell new and unheard-of electronic devices. And that's what he did.
In the 1970s, his company, that was called "Sinclair Research Limited" in the days of the ZX Spectrum, mainly produced electronic measuring devices and pocket-calculators.
But it also often released highly experimental products at the leading edge of technology, that were in several cases hardly usable in everyday life and didn't sell well. I've never heard of any other company, that released that many strange and eccentric products.
There had been for example a digital "Black Watch" (in the 1970s), and a mini-sized television "TV80" (an idea, which today has been realized in smartphones).
Basically, just the measuring devices and calculators, to a certain extent the ZX81, and then the ZX Spectrum were the only products, that were commerically successful.
Regarding the computer products, predecessors to the ZX Spectrum of 1982 were the just experimental ZX80 and the minimalistic ZX81, which seemed hardly usable, with its only 1K of RAM (exendable to 16K), its ridiculously tiny membrane keyboard and its flickering black and white graphics. Nevertheless, the ZX81 had (and has) its fan-base too.
Prices of the ZX81 in 1983 were 179,- DM for the 1K version, and 249,- DM for the 16K version.
There also was a very small printer for the ZX81 called the ZX Printer. It burnt its print-out into special aluminium-paper with a non-standard width of 10cm. It was also compatible to the ZX Spectrum, but also has to be regarded as one of Sinclair's weirder products.
In 1981, the time of the ZX81, Sinclair formed a cooperation with the American company "Timex". This way he could produce larger quantities of his products and even introduce them onto the American market.
At that time, the BBC wanted to produce a television series called "The Computer Programme". Different tasks were shown, that could be performed with a personal computer. To find the right computer for the show, the BBC launched a public submission for computer companies to present their products. The computer, that would win the competition, would then also be used in schools all over Great Britain.
Sinclair wanted this kind of promotion for his products. But at the time he could only submit the ZX81. The BBC chose another computer by a company named "Acorn" (which had been founded by Chris Curry, who had previously been an employee of Sinclair's and had worked closely with Sir Clive in the 1960s and 70s). Later, Acorn's computer was called the "BBC Micro". It was much more capable, but also much more expensive than Sinclair's products.
Sir Clive was upset, that his product hadn't been accepted.
He studied the BBC's requirements for a computer system. The topics included programming, graphics, sound and music, teletext, controlling external hardware, and artificial intelligence.
With the ambitious list of the BBC in mind, Sir Clive shaped the technical specifications of his next computer, which was the ZX Spectrum.
The ZX Spectrum, during development also called "ZX82", became a worldwide success.
In 1983, Clive Sinclair was knighted - and therefore became "Sir Clive". He also received a honorary doctorate of a university this year.
The future looked very promising. But unfortunately, Sinclair's products after the ZX Spectrum weren't as successful again.
In 1983, the "Interface 1" and the "ZX Microdrive" were released. The small Microdrive tape cartridges became unreliable after some time. The Microdrive wasn't accepted by software developers, and I don't know of any game that made use of it. In the end, the main storage device of the ZX Spectrum remained the cassette recorder.
In 1984 and 1985, the "ZX Spectrum+" and the "ZX Spectrum 128" were released.
Also in 1984, Sinclair (in cooperation with Hoover) released something completely different: The "Sinclair C5", a rather dangerous, electric vehicle, an electrically powered tricycled soapbox. The driver had to sit in it in a rather uncomfortable and unnatural position with his or her the arms down, because the handlebar had to be steered from under the knees.
The C5 was a commercial failure, and was dicontinued in 1985.
It is said, that from the inventor's point of view, the C5 was meant just as the first step to an electric car, like the ZX80 had being the first (and hardly usable) step to a home computer.
But with the release and failure of the C5, many thought, Sinclair had made a fool of himself.
Nevertheless, 30 years later there are electric bicycles.
Also in 1984, Sinclair released the "Sinclair QL" ("QL" for "Quantum Leap"), based on a processor of the Motorola 68000 family. Everybody thought, it would be a successor to the ZX Spectrum. But it was aimed more at the business market. It was supposed to be a business computer, that was more affordable than the IBM PC and the Apple Macintosh. As a storage device it used two built-in Microdrives. But as mentioned, the Microdrive wasn't accepted in the market. And so wasn't the QL. It was a huge failure.
In 1985, Sinclair's first wife divorced him. The married couple had three children.
Plans were made about a new, fast, games- and multimedia-oriented Sinclair computer with the code-name "Loki", that should compete with the Amiga, but these plans were never realized.
In 1986, the whole computer product range of Sinclair Research Limited was sold to Amstrad.
Sir Clive lost his leading role in the computer industry.
Production of the QL was immediately dicontinued.
In 1987/1988, Sir Clive once more released a computer, which was called "Cambridge Z88". It was a Z80-based, light-weight portable kind of "subnotebook" with rubber-keys for business use. It wasn't a bad machine, but didn't become a huge success.
Sir Clive also didn't give up inventing electric vehicles:
In 1992, there was a small electric bike called the "Zike".
In 1995 there was a small, portable electric engine called "Zeta", you could put onto the back wheel of an ordinary bicycle,
And in 2010 he showed a two-wheeled prototype for a successor to the C5, called the "X-1". It didn't reach mass production though.
In its time, the ZX Spectrum had been a huge commercial success. All in all 5 million units (including later versions and clones) are said to have been sold. That's (probably) why Sir Clive Sinclair - in spite of all experiments and commercial failures of his company - remained a rich man until the end.
From 2010 to 2017 he was married for a second time, to a former beauty queen who was 36 years younger than him.
He died on September 16th 2021, aged 81.
Sir Clive Sinclair was never interested in the ZX Spectrum as a games-machine. Instead he wanted it to be an affordable means to introduce people to the world of computing. In this regard, in my opinion the ZX Spectrum did a very good job.
Compared to other computers of the time like the Commodore 64 or the Atari 800XL, the system of the ZX Spectrum was much more comprehendable. The Atari for example, had 16 different graphics modes, with different rules for resolutions and the amounts of colours that can be used. The Spectrum just always had a resolution of 256x192, and the colours it had, could be used everywhere on the screen. It also didn't have a "text mode": Letters were just drawn, so the Spectrum was always in "graphics mode", even when displaying text. This may be a bit slower, but it was also rather convenient. There was for example a "PRINT AT" command in BASIC, so the user could just put letters, wherever he or she wanted.
To produce sound in BASIC on the Atari, there were four arguments to a "SOUND" command, that controlled the filter settings and so on. The Spectrum's sound was of course rather primitive, but in comparison it was very easy to produce a simple beep here and there.
The Spectrum's memory layout was also very easy to understand: In the first part, there was the system ROM, at 16384 the screen memory started, own machine code should be stored beginning at 32678, and 65535 was the last memory address.
There weren't any hardware sprites, and on the 48K Spectrum there wasn't any complicated bank switching to access more than 64K of memory with an 8-bit processor. This of course limited the capabilities of the computer. But it also made the system a lot more easier to understand for the user.
Even today I would recommend the ZX Spectrum's system, when it comes to introducing young people to computers.
Sinclair QL
In 1984, only two years after the ZX Spectrum and one year after the "Interface 1", Sinclair Research released a new computer called the "QL" ("Quantum Leap"), which used a Motorola 68008 processor (a 32 bit processor with an 8 bit data bus, running at 7.5 MHz) and was aimed mainly at office users. It came with 128K RAM (which could be upgraded to up to 896K) and had two built-in Microdrives for data transfer, which suffered from the problem of the unreliable cartridges. The QL's data format wasn't even compatible to that of the ZX Spectrum's Microdrives, so you couldn't even transfer documents between the two Sinclair machines that way (according to Nostalgia Nerd). It was possible to create a simple network between the QL and a ZX Spectrum with an "Interface 1" though. The QL had an expansion slot, to which a floppy disk drive interface like the "PCML Q+" or the "Miracle Systems QL Trump Card" could have been attached. So theoretically, it would have been possible to connect one or more external floppy disk drives to the QL and work around the problems with the built-in Microdrives that way. That may have invalidated the main argument against the QL, but it seems, customers didn't accept that solution. Instead, they just didn't buy the QL, so that it became a complete flop.
When looking at the TV advert for the QL, it becomes clear, what the machine was supposed to be. In the commercial, we see Sir Clive jumping over the IBM PC, the Apple Macintosh and some BBC office computer, to land near the QL.
If you had a small company and wanted to do word processing and spreadsheet calculation with a computer, before 1977 you couldn't. Unless you wanted to put a mainframe machine the size of a cupboard into your office.
Then, beginning in 1977, the first smaller computers became available, like the Apple II and the Commodore PET, which both ran at just 1 MHz. Also the Commodore 64 of 1982 just ran at that 1 MHz, and you wouldn't want to use the ZX Spectrum for word processing in an office either (it's just not designed for that).
So most office users bought the incredibly expensive IBM PC (1981) and then the just as expensive Apple Macintosh (1984). In this situation in 1984, Sinclair thought, they could produce a computer for the first time, that was suitable for office use, but was much more affordable than the other systems.
That didn't seem to be a bad idea, but others saw that market too: In 1985, Atari released the Atari ST with just that slogan: "Power without the Price". In 1984, Commodore struggled to put the Amiga onto the market, but also in 1985 released the Commodore 128, which had a special mode of 80 columns of text, that was supposed to be used with office applications.
I think, I agree with the consumers of that time: If I had wanted an affordable office computer in 1985, I probably would have bought an Atari ST or a Commodore 128, not a Sinclair QL.
It seems, Sinclair recognized the competition to the QL from IBM and Apple, but not so much from Atari and Commodore. Their machines had not been ready in 1984. But when they were released a year later, they were better than the QL, especially the Atari ST and the Amiga 1000 were.
Cambridge Z88
In 1987/1988, Sir Clive released a portable computer called "Z88" ("Cambridge Z88").
It was a Z80-based, dark-grey, rather slim and light-weight "subnotebook" (just 0.9 kg) with rubber-keys. The keys were better than the ones of the Spectrum; there weren't key-modes either.
4 standard AA-batteries powered the Z88 for about 20 hours. There also was an input to connect an external power supply.
The Z88 had a slim LCD-display featuring 6 lines of 80 characters each - enough to write down some texts when travelling on a train or a plane.
Memory could be expanded to several hundred KBytes or even up to 3.5 MBytes, using optional eprom-cartridges. There were three slots for such cartridges.
The Z88 had its own operating system and a preinstalled software suite called "Pipedream", including word processor, spreadsheet, database and also calendar, calculator and BASIC. There also was a RS-232 port.
For file-transfer, a package "PC Link II" was sold separately. It contained a PC-program on a 5.25''-disk, a Z88-program on a cartridge and a required non-standard serial cable. Wordstar and Lotus-formats were then supported too.
Think of the Z88 as a Spectrum turned into a subnotebook for business use, coming with a pocket calculator screen.
The machine made more sense than the QL of 1984, that had already been aimed at the business-market.
It had also a better concept than PC-laptops of the time, like the "Tandy 1100FD" with a weight of 3.2 kg.
As the Z88 didn't spy on you in the background, didn't hide the system from you and provided a proper keyboard and useful productivity software, it even made more sense than most mobile computers today.
But it just remained exotic.
ZX Spectrum Software Development on a modern PC (in C, Assembly, BASIC)
Today, it is possible to write programs in C or Assembly for the ZX Spectrum on a modern PC using the cross compiler "z88dk".
Its compilation from source is a complicated and ridiculously long process though. Instructions how to compile on Linux can be found here. They have to be followed closely, especially regarding the installation of the long list of prerequisites. "zsdcc" is also needed, so the "export"-variables have to be checked, before running "./build.sh", as described in the document.
Once "z88dk" is installed, the official tutorial to get a "Hello World" program compiled on the PC, and running on the virtual or real ZX Spectrum would be here.
In an older version, the file "z88dk.m4" from the compiler distribution always had to be in the directory of the source files when compiling. The tutorial didn't mention that, and it took me several days to figure it out. In version 2.2 of "z88dk" (of 2022), this doesn't seem to be necessary any more.
A C program "hello.c" would be compiled with z88dk with:
export ZCCCFG=/usr/local/share/z88dk/lib/config zcc +zx -vn -O2 -startup=0 -clib=sdcc_iy -zorg=32768 hello.c -create-app
This creates a file "hello.tap", that can then be loaded into an emulator like "fuse". The command
zcc +zx -vn -O2 -a -startup=0 -clib=sdcc_iy -zorg=32768 hello.c
would create a file "hello.c.asm" containing a listing of the generated assembly code.
Also writing Z80 assembly language on a PC is possible with the cross compiler "z88dk". Here's a simple "Hello World" program in ZX Spectrum assembly. The program's "cheating" in a way, because it calls the PRINT routine in the Spectrum's ROM for printing (I don't know much Z80 assembly yet):
.jpg){kind=link}
org $8000
jp start
string:
defb "Hello World!", 13
start:
CHANOPEN EQU $1601
PRINT EQU $203C
CLS EQU $0D6B
call CLS
ld a, 2
call CHANOPEN
ld de, string
ld bc, 13
call PRINT
ret
On my system, this code compiles to "hello.tap" (again) with the command:
z80asm +zx hello.asm
It is also possible to write ZX Spectrum BASIC code on a modern PC, and feed that code into an emulated Spectrum. The utility "bas2tap", which can be found on "World of Spectrum", reads in a text file with Spectrum BASIC code on a PC and (yet again) creates a ".tap"-file from it.
ZX Spectrum Next
In 2017, there was a kickstarter campaign to build a "ZX Spectrum Next" as a retro computing project. Those who took part in the campaign, received the machines in February 2020. So up to now, the "Next" can't be bought in regular shops.
The "Next" runs a FPGA processor in Z80-mode. It can run the 3.5 MHz of the original Spectrum, or provides "Turbo-modes" of 7, 14 and 28 MHz. It has 1 MB RAM, expandable to 2 MB. For storage, SD-cards are used. It can be connected to HDMI devices or VGA-monitors.
On the outside, the "Next" looks a bit like a modern version of the Spectrum 128. It's really beautiful. Case and keyboard of the "Next" were designed by Rick Dickinson (1957-2018), who had also designed the keyboards of the classic Sinclair computers.
The "Next" also has two DB9-ports for the old joysticks used in the 1980s.
It has three AY-3-8912-chips which produce sound in, well, AY-quality. External speakers have to be connected to the Next's 3.5mm stereo audio output.
Basically, the Next uses the old resolution (256x192), but can also produce higher resolutions, 256 colours and up to 128 hardware sprites.
It comes with its own Spectrum-style operating system called "NextZXOS".
An enhanced version of the Next called "ZX Spectrum Next Accelerated" also has an additional accelerator board and provides much more processing power (1 GHz), more RAM (512 MB) and also WiFi for networking.
The "Next" is compatible to the old machines, so it can be used to play Spectrum games. But the Next can also go beyond the old specifications.
Games in the new modes, with more colours, sprites, sound and more speed remind a bit of the Commodore Amiga.
To understand, what the Commodore VIC-20 and also the C64 were about, we first have to take a look at Commodore's previous product, the PET series.
The Commodore PET
In 1976, Commodore had bought the company "MOS Technology", that produced the 6502 microprocessor, that were used in many other computers of the time like the Apple II or the Atari 800. "MOS Technology" then also produced the other chips for the Commodore-computers.
Commodore released the "PET 2001" in 1977, the same year the Apple II and the TRS-80 were released.
The PET used a 6502-processor, running at 1 MHz. All three machines came with as little as 4K RAM, that could be expanded. The PET was much more affordable than the Apple II though. The PET ran BASIC, a version, Commodore had bought at surprisingly low cost from Microsoft.
The PET didn't have any sound at all. It didn't have any graphics modes either. But its character set (called "CBM ASCII" or "PETSCII") featured graphical symbols, that were shown on the keyboard and could be displayed on the screen. That way, pseudo graphics (also called "block graphics") could be produced.
The PET 2001 had a rather weird membrane keyboard (derived from Commodore's pocket calculators of the time) and a built-in casette tape. It came with a monochrome monitor. The PET looked somehow futuristic. It was actually quite beautiful.
The later "PET 4016" and other modells had a proper typewriter keyboard, and the casette tape wasn't built-in any more. It looked even better.
Data transfer with the cassette tape was extremely slow. But as in the base configuration there were only 4K to load, that didn't matter much.
Commodore then also built a floppy disk drive for the PET (Commodore 2040). As the PET wasn't designed to use a disk drive at all and also couldn't be expanded with a floppy controller, the disk drive had its own 6502-processor and other hardware, so technically it was a computer of its own. The PET just gave commands to the disk drive, and the drive took care of the rest and in the end provided the data.
Several PETs (in a computer lab or a school, for example) could be connected to one floppy drive.
From the PET to the VIC-20
At that time, "MOS Technology" created a custom chip for graphics and sound, called 6560, "Video Interface Chip" (VIC).
Commodore saw, what Sinclair did in Britain with the ZX80. So Commodore took concepts from its PET series to build an affordable home computer, using the VIC-chip in combination with the 6502 processor.
The result was the Commodore VIC-20, released in 1980. It's initial price was at $299 (398,- DM in Germany in July 1983 - plus 100,- DM for an optional 16K memory expansion, plus 100,- DM for a tape drive VC 1530 (Vobis list)). Like the PET, the VIC-20 used the 6502 CPU, running at 1 MHz. It had a proper typewriter keyboard, not just rubber keys. The manual was well-written, and invited the user to the world of computing. The VIC-20 was connected to a television and also featured colours. Basically 8 of them, 16 under some circumstances. It also had sound.
The biggest drawback of the VIC-20 was, that it came with just 3.5K RAM available to the user.
Like the PET series before, the VIC-20 didn't have dedicated graphics modes, so basically it only displayed text, in a resolution of 22x23 text characters (that used a 8x8 bit matrix). It was possible to get graphics with some work-arounds though:
On the one hand, the PETSCII character set contained graphical symbols, that could be used to display graphics using graphical "letters". The graphical symbols were also printed on the keys of the VIC-20's keyboard. This concept of "block graphics" had been adopted from the PET series..
The graphical characters of the PETSCII set were integrated seamlessly into the system of the Commodore machines (also of the C64). They could be found directly on the keyboard, and be entered by pressing either "Shift" or the "Commodore key" together with the corresponding key. They were shown in BASIC listings just like ordinary characters and could also be printed as such. They had their own PETSCII-values, that could be retrieved using the "ASC()"-command, and the "CHR$()"-command converted these values back to characters. So these PETSCII-characters were always available to the user to draw simple graphics to the screen.
On the other hand, the look of the VIC-20's text characters could be completely redefined. That way, it was possible to write games, that made use of just moving redefined text characters around the screen.
The characters moved in character spaces then. "Pixel exact" movement could only be achieved using advanced programming techniques, which involved redefining characters all the time.
The VIC-20 also supported a "double height character mode", in which text characters were displayed twice as high. In this mode it was possible to fill the entire screen with the 256 different characters, that were available. As these could be redefined, this made it finally possible, to separately access every pixel on the screen (there were 176x184 pixels). Unfortunately, this couldn't be realized with just 3.5K RAM.
The VIC chip was also able to generate sound, with three square wave voices plus noise. The sound had some limitations (no envelope, no filter, limited voice range), but it already was much better than what most other computers of the time could produce.
The VIC-20 had a single joystick port. It also had a slot for ROM game cartridges. Many of these game cartridges also added 16K more RAM for use with the game. As RAM was expensive, that probably raised the price for the cartridges.
In 1981, Commodore released a cartridge called "Super Expander" ("VIC-1211A"), which improved the VIC-20's BASIC, and also added 3K more RAM for general use. In fact, it added graphics modes to the VIC-20, and also BASIC commands to control these graphics such as "POINT" (plot), "DRAW", "CIRCLE" and even "PAINT". A "SOUND" command was also added. So basically, it fixed most problems with the general use of the VIC-20. But not many people owned this cartridge, and developers kept targeting the unexpanded machine. The cartridge's circuitry should have been part of the VIC-20 in the first place. But it wasn't.
There also was a cartridge called "VIC-1111" that expanded the VIC-20's memory by 16K.
The VIC-20 was the first computer which was sold in general stores, not in specialised computer shops. It was promoted by a big campaign featuring William Shatner (who had played "Captain Kirk" in "Star Trek").
The VIC-20 was a commercial success. It was produced up to 1985. Overall, about 2.5 million units were sold.
In Germany, the VIC-20 was just called "VC 20", because "VIC" could be spoken as a rude word in German. There was also the idea of "VC" meaning "Volkscomputer", which means "people's computer". It even said "VolksComputer" on the VC 20's package, so that was an official idea of Commodore.
The "8-Bit-Guy" gives an overview of what games on the VIC-20 looked like.
A modern expansion cartridge for the VIC-20, which adds 32K RAM, has 40 cartridge games built-in and makes booting to some SD-card adapter possible is called the "PenUltimate Cartridge".
An emulator of the VIC-20 (and other Commodore-computers) is called "VICE". The command to start the VIC-20 emulation is called "xvic" (run "xpet" for the PET emulation).
I didn't have a Commodore 64 (also called C64 or CBM 64).
In the 1980s, it was ominipresent though. It could be seen (and heard) in shops, in people's homes, at computer clubs and on television. In West-Germany, the logo of the magazine "64'er" kept reminding you of the machine at newsdealers. Classmates were talking about it during breaks in school. Several friends of mine had one.
Like the ZX Spectrum (539,- DM in 1983), the C64 was released in 1982. At first, it was way more expensive (1.495,- DM), but its price quickly dropped to about 795,- DM within the next year. Nevertheless, to make use of it, you also had to buy the floppy disk drive "VC 1541" for about another 775,- DM. Or at least a tape drive (model "VC 1530 (C2N)", called "Datasette" in Germany) for about 140,- DM (according to a price-list of the trading company "Vobis" of July 1983).
A price-list in the computer magazine "RUN" of December 1985 (12/85, page 131) showed the C64 at 468,- DM and the VC 1541 at 499,- DM (VC 1530: 79,- DM).
The case of the C64 was light brown and had a similar form like the white one of the VIC-20.
The C64 used a microprocessor called "6510" as the CPU, that was produced by Commodore's subsidiary "MOS Technology". It was basically a "6502", that had been slightly modified to fit into the hardware environment of the C64. It was running at about only 1 MHz.
To increase the performance, the 6510 CPU was supported by a graphics-chip called "VIC-II" and a sound-chip called "SID".
The C64 came with 64K of RAM, but only 38K could be used for BASIC-programs.
The C64 had a proper typewriter keyboard, the same as the VIC-20's, with the letters and also graphical PETSCII-symbols printed onto the keys. There weren't any "key-modes" like on the ZX Spectrum, you could just type in commands in ordinary single letters.
On the back side, there was also a slot for inserting dedicated ROM cartridges. Actually, I've never seen anybody use such cartridges with the C64. Instead, it was all about floppy disks (or cassettes). But it seems there were some useful applications for cartridges too.
The C64 had two joystick-ports. You had to be careful with these: If you plugged joysticks in or out, while the computer was running, there was a risk to break it.
The C64 had a RF-output to connect it to a television. It also had another audio-/video-output, which allowed to either connect it to a television using a composite signal or to connect it to a monochrome monitor (which led to a very clear monochrome screen). The output also produced a socalled "LCA"-signal which (by using an adapter cable) was compatible with the later S-Video-technology. That signal could be used to connect the C64 to the monitor "Commodore 1702", that showed a very clear screen with colours.
When you switched on the C64, it showed a pastel-coloured blue screen with a light-blue border. On the screen was written in the same colour as the border's:
**** COMMODORE 64 BASIC V2 ****
64K RAM SYSTEM 38911 BASIC BYTES FREE
READY.
Text resolution was 40x25 characters. The characters were quite big and could be read well. Actually, even on today's systems, I keep in mind not to configure text fonts much smaller than the ones of the C64 (the size can be checked in emulation).
By pressing "CTRL+2" on the C64's keyboard, the colour of the text could be set to "white", to raise the contrast in relation to the background (you could also press the "Commodore key" plus "6" to achieve something similar).
Data was transfered either using a special, dedicated cassette recorder (model VC 1530 (C2N), called "Datasette" in Germany).
To load, you just entered "LOAD" and it said:
PRESS PLAY ON TAPE
Loading this way was unbearably slow, though.
Or you could buy the much more expensive floppy disk drive "VC 1541", which used 5.25''-disks.
Probably the worst thing about the C64 was, that this drive also was disappointingly slow. In some cases, even the Spectrum's tape worked faster. So even with the 1541, you often had to wait 6 or more minutes to load a single game (they say, the transfer-rate was 300 bytes per second, but I think, that drive was even slower).
However, later there were programs to speed up the 1541 such as "Epyx Fast Load". These programs weren't bad, but they weren't reliable with every kind of software to be loaded.
To view the content of a disk used in the 1541, you had to type:
LOAD "$",8
"8" was the drive-number and meant the disk drive. It then said:
SEARCHING FOR $ LOADING READY.
Then you typed in:
LIST
and the content of the disk was shown.
The reason for this was the heritage from the PET. Like the PET, the C64 didn't have dedicated commands to view the disk's directory. It just had "LOAD", "SAVE" and "VERIFY". The floppy disk drive had its own processor and was a computer of its own. The C64 just talked to that "computer", and the drive took care of the rest. It returned the disk's directory to the C64 as a BASIC program. That's why you had to view it with the "LIST" command.
The floppy drive "VC 1541" had originally been designed for the VIC-20. There had been a bug in the communication between the computer and the disk drive. Commodore had decided not to fix the bug, but to use the computer's CPU to process the data coming from the drive as needed. That led to very slow disk access. On the VIC-20 that didn't matter much, as it had very limited memory, so not much data had to be processed.
But with 64K RAM it mattered, so that's why the C64's disk drive was so slow. It seems, Commodore didn't fix the problem, because they realized, that consumers were willing to spend money on additional software on cartridges (such as "Epyx Fast Load") to solve the problem.
The following command on the C64 would load the first program on the disk as machine code. You usually typed this, when you wanted to start a game from disk:
LOAD "*",8,1
The ",1" was needed to indicate, a machine code program should be loaded. It then said:
SEARCHING FOR *
LOADING
READY.
Then you entered
RUN
and started the game. Sometimes, it also started automatically, without the need to enter "RUN".
LOAD, SAVE and VERIFY were the only disk commands built into the C64's BASIC. If you wanted to delete a file from the disk, for example a file called "SOMEFILE.PRG", you had to use this construction:
OPEN 1,8,15,"S:SOMEFILE.PRG":CLOSE 1
The graphics chip inside the C64 was called VIC-II. It was the successor to the VIC-20's VIC-chip.
The C64's VIC-II featured a palette of 16 colours and a bit-mapped high resolution graphics mode with 320x200 pixels. To save memory, colours could only be set per character cell (40x25). In this mode, two colours per character cell were supported.
The 16 colours of the C64 had a somehow pastel kind-of look to them.
To draw simple graphics, the PETSCII character set was available to the user, just like it had been on the Commodore PET series and the VIC-20. The PETSCII characters could be easily printed, be seamlessly used in BASIC listings, and they were also printed onto the keys of the C64's keyboard for quicker access.
The C64 could render 8 hardware sprites, that were independent from the graphics modes and the colour cells mentioned above. So they were able to add more colours to what the general graphics mode offered. With one colour (from a palette of 16), sprites could have a size of 24x21 pixels.
There also was a "multicolor mode". In this mode, sprites could have a size of just 12x21 pixels, but all 8 sprites could have two colours (altogether - that means, all 8 sprites were limited to these two colours), plus there was a third colour for each sprite. So three colours per sprite altogether, but not three different colours per sprite.
When you have a single colour sprite of 24x21 pixels, you need 8 * 3 bits (3 bytes) to store the information of one row on the x-axis. And you need 21 of these rows. So the pure image information of one sprite uses 63 bytes.
Sprites can be moved fast around the screen. They use an area of 512x256 pixels, so parts of this area are outside the visible screen.
The C64 had a sound-chip called SID ("Sound Interface Device"), which had been created by Robert Yannes at MOS Technology, Commodore's subsidiary chip producer.
The SID had three channels, so it could produce three different synthesized sounds at once (a lead, a bass and some drums, for example). It also had a filter, similar to those used in professional synthesizers, which was quite a unique feature. To achieve the impression of a chord on just one sound-channel, often very fast arpeggios were programmed, which had a characteristic sound. The SID was probably the best and most popular sound-chip in home computers of the time. Nevertheless, its music still didn't sound anything like pop-music produced in a professional studio. Although it had its own charm.
Somehow, commercially that worked out: When you heard the melody of a certain pop-song in a game on the C64 and liked it, you still wanted to buy the studio recording of that song on LP or CD, as the SID-version always left you with the feeling, that the song could sound better, if only there were better instruments. Probably that's the reason, today there are lot of approaches, to remix many of the songs, originally composed for the C64.
Unlike other home computer sound chips, the relatively advanced features of the SID attracted a few artists, that were as well professional musicians as 6502 assembly programmers. They composed interesting musical pieces for the SID. As software companies wanted their C64 games to sound good, they often hired these artists to create the music for their games. Musicians, working with the C64's SID were:
My favourite composition of these SID-tunes is "Giana Sisters" by Chris Hülsbeck. I liked it so much, I did a recreation of it myself. Unfortunately, people say, it turned out just mediocre. I'm sorry, I couldn't do better.
To interrupt a program on the C64 - mainly a BASIC-program - you had to type "RUN/STOP + RESTORE" ("RUN/STOP" was a special key on the C64's keyboard, "RESTORE" another one).
In Commodore BASIC, often-used abbreviations were "?" for the "PRINT" command, "L╮" for the "LIST" command and "R╭" for the "RUN" command.
The "╮"-sign was a PETSCII symbol, and was produced by pressing "Shift+I" on the keyboard of the C64.
Accordingly, "╭" was produced by pressing "Shift+U".
You could skip space characters in C64 BASIC commands. So you could type
LOAD"$",8
without the space between the command and the next character.
The C64 would perform a reset after the command:
SYS 64738
The "SYS"-command started machine code programs in the memory. There also were the commands "PEEK" and "POKE".
But in general, the C64 didn't encourage the user to write his own programs (or hers - but actually, I didn't know a single girl, that owned a home computer back then, it was almost exclusively a boy's thing). With colours, sprites and the SID, the C64 had quite good technical features, but it kind of hid them from the user.
Instead there grew quite a large software industry around the C64, that provided users with games and applications in return for good money.
But to the displeasure of that industry, also a large cracking scene developed, and lots of disks with cracked games were exchanged in schoolyards.
Some young users even studied the C64 so much, that they finally managed to reveal its hidden features in their programs, or could even build their own extensions to its hardware.
Here's the technical reason, why there were just "38911 BASIC BYTES FREE": Eeight bit processors can access just 64K of memory altogether (memory addresses 0 to 65535). But the C64 had 64K RAM plus 24K ROM, which would sum up to 88K. To manage that kind of memory, MOS Technology modified the 6502, creating the 6510 processor. With the modifications it was possible to swap memory banks as needed. The addressable area remained 64K, but that area could be shared by different types of memory. In the default configuration on startup, the memory banks were configured in a way, that BASIC could access 38911 bytes of RAM (although the size of Commodore BASIC in ROM was just 8K).
However, in the default configuration there were 4 more KByte of RAM available in an upper memory block. But these 4K couldn't be used to store BASIC code.
The person responsible for the C64 was the president of Commodore, Jack Tramiel (1928-2012) (the image shows him in maybe 1982, he's the person on the left).
Jack Tramiel was born in 1928. He was of Polish-Jewish origin. During World War II, the Nazis sent him and his parents to "Auschwitz", which was the Nazis' main death camp. Tramiel was selected for working and was moved to another camp near Hanover. In April 1945, at the age of 16, he was freed from this concentration camp by US-troups.
He went to the USA and joined the U.S. Army. There he learnt how to repair office equipment including typewriters. In 1954, he founded "Commodore Business Machines International" (CBM), a company originally producing typewriters.
In the 1960s, Commodore needed money, so that the businessman Irving Gould (1919-2004) became shareholder and chairman of the board of directors. Jack Tramiel stayed president of the company.
He had a reputation of being an extraordinary tough American businessman. He had a habit of controlling every expense of the company larger than $1,000 in person. When he went on vacation, the business of Commodore just stopped.
In the 1970s, Commodore produced pocket calculators and computers like the "PET 2001" (1977).
It also became a global company with divisions all over the world. Tramiel was just president of the US-division. Although that was probably the most important one, where the general decisions about new products were made, the divisions outside the USA worked more or less independently from Tramiel.
Maybe that's why "Commodore Germany", Commodore's main distributor in Europe, had a good reputation and wasn't known for rough business practices.
A quote by Tramiel is:
Business is war. I don't believe in compromising. I believe in winning.
Probably, Tramiel's work principles had developed, when he had been in the US-army in the late 1940s.
But there are also other quotes of Tramiel:
We need to build computers for the masses, not the classes.
and
We are here to serve the customer. Give him the best product for the lowest price. The customer is always right.
When the C64 became available in the early 1980s, I somehow didn't get comfortable with the machine.
Actually, there were interesting reasons for its weak points or quirks:
So I didn't buy a C64. Many other people were quite happy with it, though.
In a way it had similar capabilities like the Apple II (the C64's sound chip was even better), but it was much more affordable.
Although Commodore's PETs with their integrated monochrome monitors had looked like office PCs and even futuristic, they had not been that capable. The C64 used a very similar processor running at the same speed (1 Mhz), but was way superior, when it came to graphics, colours, sprites and sound. So before Commodore started producing clones of the "IBM PC" starting with the "Commodore PC-10" in 1984, and before it bought "Amiga Corporation" in the same year, the C64 really was the computer with the best overall performance, Commodore could deliver. At times, Commodore could even more or less be seen as "the C64 company".
Actually, in the year 1983 Commodore also released the office computers 610 and 620. In the US, they were called "B128" and "B256". They featured an 80 columns black and white text-mode. They were based on a 6502-compatible processor called 6509 running at 2 MHz. They had 128K (610) or 256K (620) RAM, but it was somehow complicated to write programs using that amount of memory. They didn't have graphics, sprites or even colours, just black and white text (including PETSCII-symbols). They had the SID-chip though. The 610/620 were meant as pure office computers and were much more expensive than the C64. They weren't successful and remained exotic. In 1984, they were withdrawn from the market, and Commodore turned to the "PC-10" instead. For further information about the 610/620 computers read about the "the Commodore CBM-II series".
But back to the C64: The main engineers, that had designed its chips - Al Charpentier (graphics) and Robert Yannes (sound) - didn't receive appreciation for their work by Tramiel. They didn't get a bonus, instead he even cut their salaries (because in his opinion they hadn't worked fast enough). They then left MOS, the subsidiary of Commodore. Yannes founded the company "Ensoniq" in 1982, that produced professional synthesizers ("ESQ-1") and samplers (like the "Mirage", the first affordable sampler (1985)). Charpentier followed Yannes to Ensoniq.
While the PET had been popular in computer classrooms in schools (my school used several PET 4016, for example), the C64 wasn't often used in schools. It turned out, it often was stolen or it broke. Commodore created a computer "Educator 64", a C64 in a PET case, but it was too late. The market for school computers couldn't be won back.
A few years later, after Tramiel had left Commodore, there was a model "C64-II" (or "Commodore 64C", "C64C") in a more modern case.
The "C64G" had a "breadbin"-case again, but a white one.
In 1984, there also was a portable model called "SX-64". It was a box (or case) weighing 10.5 kg, holding the computer itself, the floppy drive and a CRT (tube) television/monitor with a really tiny 5'' screen. The keyboard was connected externally. So there you had it: A portable C64.
Actually, at the time the SX-64 wasn't as weird as it looks from today's perspective: There were several business computers that had a similar design with a tiny built-in CRT monitor, such as the "Osborne 1" (1981, running CP/M) and the "Compaq Portable" (1983, running MS-DOS). So at the time, there must have been a market for such machines (as inexplicably there seems to be a market for mobile computers today too).
In 1985, the C128 was released (see below), which could operate in a C64-mode, that was 100%-hardware-compatible.
An emulator of the C64 and other Commodore-computers of the time is called "VICE".
There is even a dedicated hardware for C64 emulation. It is an ARM-computer in a C64 case with a proper C64 keyboard, that runs the emulation. It can be connected to a modern HDMI television set. This product is called "TheC64 Maxi".
Today, it is also possible to write programs in C or assembly language for the C64 on a PC using the cross compiler "CC65". A simple "Hello World"-program in C called "hello.c" would be:
#include <stdio.h>
void main() {
printf("%s\n", "Hello World");
}
The program would compile with a small script like this:
#!/bin/bash export CC65_HOME="/usr/local/share/cc65" # Set environment cl65 -O -t c64 hello.c -o hello.obj
You can then use the utility program "c1541", that comes with the VICE-emulator, to create a blank C64 ".d64" disk-file called "mydisk.d64", and write the file "hello.obj" onto it:
c1541 -format diskname,id d64 mydisk.d64 -attach mydisk.d64 -write hello.obj hello.bin
Then you can launch VICE with "x64 mydisk.d64", and it will run your program.
With
export CC65_HOME="/usr/local/share/cc65" cc65 -O -t c64 hello.c
you get a file called "hello.s", that contains a listing of the generated assembly code.
It is also possible to write C64 BASIC code on a modern PC, and feed that code into the emulated C64. The easiest way to do that, is to run VICE with "x64" and then paste your code directly from the PC-clipboard into the emulator window, using F12, and then "Edit/Paste from Clipboard". You can then use the emulator, to save that program to a ".d64"-disk. But watch out for the ancient trap, that a 'LOAD "$",8' to examine the disk's directory would delete your BASIC program from memory and replace it with the disk's directory.
Alternatively, you can write a text file "program.txt" with the BASIC code on the PC. Use lower case letters (!) and line numbers for the C64 commands and (by default) for the rest of the code. Like this:
5 rem program.txt 10 print "commodore 64 basic" 20 goto 10
From this template, a ".prg" file can be created using the "petcat" command, that comes with VICE:
petcat -w2 -o program.prg -- program.txt
This "PRG"-program can then be run with "x64 program.prg".
Note: The small letters in the string in the template file are translated to capital letters in the C64 program. If you used capital letters, they would be translated to PETSCII graphical symbols (which is probably not what you want). If you also want small letters in strings in the C64 program, start the string with the control code "{swlc}". Then you can also write capital letters in the template file.
It is also possible to create a ".d64" disk-file, that contains the "PRG" program. The file names (especially the name of the "prg" file) also have to be in small letters:
petcat -w2 -o program.bas -- program.txt c1541 -format diskname,id d64 mydisk.d64 -attach mydisk.d64 -write program.bas program.prg
The program can then be run with "x64 mydisk.d64".
When I wrote this part about the Commodore 64, I remembered, how omnipresent it really had been in West-Germany of around 1984. It had been the mainstream home computer of the day. It dominated not just the market, but also had an influence on society and the public discussion about computers in general. Almost every software company produced programs for it as a main target platform. Magazines and computer television shows focussed on it. It seemed, viewers were expected to have one. To many people, the C64 was the only home computer they had ever heard of. Some people even talked disrespectfully of other systems.
All of this gave me a feeling of reluctance against the C64, that (as it seems) still stays on with me until today.
Both released in 1982, the Commodore 64 and the Sinclair ZX Spectrum (48K) were the best-selling 8-bit home computers on the market. So there was a lot of discussion, which one was "better". Here's a comparison of the two:
| Commodore 64 | Sinclair ZX Spectrum (48K) | |
| Processor | MOS Technology 6510 (6502) at 1 MHz | Zilog Z80a at 3.5 MHz |
| Text resolution in characters | 40x25. | 32x22 (plus 2 more lines for text input). |
| Graphics resolution in pixels | Up to 320x200. | 276x192 (directly accessible, 16 lines of pixels used for the command line), surrounded by a fix border. |
| Colour Palette | 16 colours (pastel-looking). | 8 colours (ordinary and bright). |
| Colours per cell | 2 in high-res mode, 4 in multicolor mode. | 2. |
| Hardware Sprites | 8. | None. |
| Scrolling | Support for smooth scrolling. | In general only hard scrolling (by character cell). |
| Sound | SID chip (3 voices, filter). | No sound chip, just a beeper. |
| Storage devices | Datasette VC 1530 (extremely slow) or floppy disk drive VC 1541 (expensive, very slow). | Any cassette tape recorder (slow) can be connected; alternatively "Interface 1" plus Microdrive (expensive, faster, but unreliable). |
| Games | Almost any 8-bit game available on floppy disk. | A huge library of games available on cassette tape. |
| Free RAM in BASIC | 38.911 Bytes. | 41.612 Bytes. |
| BASIC dialect | POKE is needed for graphics and sound. | Has commands for graphics (PLOT, DRAW, PRINT AT, PAPER, INK) and "sound" (BEEP). |
| Keyboard | Typewriter keyboard, integrated in the case. | Just rubber keys. |
| BASIC keyboard mode | Ordinary typing of letters. | Custom keyboard modes with different meanings of keys. |
| Joystick ports | Two built-in. | None by default. Possible with external interfaces. |
| Video Outputs | RF-output. Additional composite- or S-Video-output, which also can be connected to a SCART-input. | Just an RF-output (radio frequency-output). |
| Printers | Office-type printers available. | Ridiculous ZX Printer. Just a few other printers available. Some word processor programs though ("Tasword", "OCP Word Manager"). |
| Initial price in 1982 | About $600, plus about $300 for floppy drive 1541. | $280, plus $50 for a cassette recorder. |
| German Prices in July 1983 | 795,- DM, plus 775,- DM for floppy drive 1541 (or just 139,- DM for tape drive VC 1530). | 539,- DM, plus about 70,- DM for a cassette recorder. |
Why was there this discussion at all? The concepts of the two machines seem to be rather different. Wouldn't you rather compare the Atari 800XL to the C64 than the ZX Spectrum?
I think it has something to do with the timeline: If you wanted to buy a home computer in the years before 1982, let's say in 1981, you basically had two choices: Either you spent a lot of money, several thousands of dollars (or DM). Then you could buy an Apple II, IBM PC, BBC Micro, TRS-80 or Atari 800 (yes, the original Atari 800 of 1979 was very expensive too).
Or, in 1981, you could get yourself a VIC-20 or a ZX81. But the VIC-20 only had 3.5K RAM, the ZX81 even only 1K. So these machines weren't hardly usable.
Then, in 1982, suddenly the C64 and the ZX Spectrum were available. Both with colours, lots of games and a reasonable amount of RAM (64K / 48K). And consumers thought for the first time, now that's a computer that is well usable and affordable. It was the breakthrough of home computers, in a way it was a technological miracle. So these two machines made a special impression on the general public, and that's why they also were compared with each other. Although regarding the technical specifications and the general concept of the machines, the comparison didn't even make much sense.
In fact regarding price and features the Atari 800XL was much more close to the C64 than the ZX Spectrum. But it wasn't released until 1983, so it wasn't such a big surprise on the market any more as the two other computers had been the year before.
Another option in 1981 might have been the TI-99/4A. But Texas Instruments tried to control the software developers (especially the developers of game cartridges), so that not much software was produced for that machine.
Although the C64 without doubt had technical advantages, the ZX Spectrum had its own charm, which made many users like it. After all, you could do similar things with both machines, like writing BASIC programs, managing your housekeeping money or playing games. In spite of the limited graphics- and sound-capabilities of the Spectrum, some programmers were able to create marvellous little games for it.
So technically the C64 may have been the superior machine, but the Spectrum was ... different. And that's something too.
The Texas Instruments TI-99/4A was released in 1981. It had a silver-coloured metal case and a solid, black typewriter keyboard. It looked quite elegant.
To the right of the keyboard there was a slot where special cartridges could be inserted into the computer.
The TI-99/4A used Texas Instruments' own processor, a 16 bit CPU called TMS9900, running at 3 Mhz.
It came with 16K RAM. It could produce up to 15 colours and also sprites.
It had a sound chip with three sound voices and an additional noise channel. The same chip was also used in the BBC Micro, and later in the DOS based IBM PCjr and Tandy 1000. The TI-99/4A also could produce speech, that sounded similar to the device called "Speak & Spell", also by Texas Instruments.
Programs could be loaded with an external cassette recorder. The computer could be connected to a general purpose recorder, but Texas Instruments also offered a dedicated one called "TI Program Recorder“ (model PHP2700). There also was a 5.25'' floppy disk drive, but it seems, most people used the cassette recorder with the computer.
The TI-99/4A was in the upper price range ($525 at its release in 1981), but it wasn't as expensive as the Apple II, for example.
It was quite successful, until there was severe competition, when the Commodore 64 and the Sinclair ZX Spectrum entered the market in 1982.
Since Texas Instruments had won the competition against Commodore in the market for pocket calculators in the 1970s, Commodore's president Jack Tramiel considered Texas Instruments his main competitor.
With the C64 he started a price war against the TI-99/4A.
An argument for the C64 was, that software developers were free to create any program for the machine they wanted, while Texas Instruments licensed only certain developers to program their proprietary processor and the cartridges (which they even had to buy) for the TI-99/4A.
In July 1983, the price of the TI-99/4A in Germany was 585,- DM (Vobis list).
In the end, Texas Instruments had to lower the price of the TI-99/4A to even less than that of the VIC-20. Commodore won the price war and pushed Texas Instruments out of the home computer market.
In October 1983, Texas Instruments announced to stop producing the TI-99/4A. It was discontinued in March 1984.
The BBC Micros were a series of altogether nine computers by the British company "Acorn". The first one, "Model A" was released in 1981.
In 1981, the BBC wanted to produce an educational television series called "The Computer Programme". To find a suitable computer for the show, they launched a public submission for computer companies to present their products.
Acorn, lead by Christopher Curry, won the competition and produced in the following years a series of computers to the requirements of the BBC. Acorn could also sell many computers of these types for the use in British schools.
"Model B" was solidly built and had a proper typewriter keyboard. It used a 6502-CPU, running at 2 Mhz, and came with 32K RAM. Later, the models "B+64" and "B+128" had 64, respectively 128K RAM. The maximum graphics-resolution was 640x256. Depending on the graphics-mode, up to 8 colours were supported. There was a sound chip with four sound-channels.
Unlike most other home computers, the BBC Micro used an internal power supply unit.
At first release, the Sinclair "ZX Spectrum 48K" cost £ 175 (quickly dropping to £ 129), the Acorn "Model B" £ 399.
So Sinclair made lots of compromises, to keep the Spectrum affordable for a wide range of people, while Acorn built the BBC Micros without compromises for people with more money or for governmental institutions.
Which somehow made sense: Schoolkids could watch the BBC's computing examples on television. They could try to recreate the examples in their classes on a BBC Micro. If they wanted to try even more at home, the ZX Spectrum was a good choice, as it was more affordable.
There was an interesting, rare extension for the BBC Micro, the socalled "UMI-2B" Music Sequencer. It was a combination of hard- and software, to control external instruments (synthesizers and drum machines) by Midi. The UMI Music Sequencer had been built in cooperation with Vince Clarke and was often used by him to create his music.
There is also a Youtube-clip of Depeche Mode working with the BBC Micro and the UMI Music Sequencer in a studio in Paris. It was probably in 1986/1987, as the songs they were working on were from their 1987's album "Music for the Masses".
On the one hand, the "Model B" was solidly built, had a proper keyboard and a 6502-based processor like the Apple II or the C64.
On the other hand it had the British charm of the ZX Spectrum. Many Spectrum-games had been converted to the Micro. "Elite", by David Braben, had even been an original BBC Micro-game.
Overall the BBC Micro maybe could be seen as a "British Apple II".
An emulator of the BBC Micro is called "BeebEm".
In April 1984, the British company Amstrad released an 8-bit computer called "Amstrad CPC 464".
1984 was rather late for an 8-bit computer: Apple started selling 16/32 bit Macintoshs this year, and Amiga Corporation had shown their "Boing Ball"-demo at the C.E.S..
Nevertheless it seems, the "Amstrad CPC 464" became quite popular at the time.
The Amstrad CPC 464 used a Z80A-processor as the CPU, running at 4 MHz. That is, it used the same processor as the Sinclair ZX Spectrum, it just ran slightly faster. The Amstrad had 64K RAM. It had a proper keyboard, a single joystick-port and came with a built-in cassette recorder for data transfer.
Rather unusual was, that it also came with its own monitor. Either with a monochrome monitor, that produced a rather ugly "light green on dark green" screen (£ 199 complete with the computer), or with a colour monitor (£ 299 complete).
Initial prices were even higher: £ 249 (with green monitor) and £ 349 (with colour monitor).
Using a real computer monitor lead to a significantly sharper image compared to connecting a computer to a television, like it was required by most other home computers at the time.
The power supply unit was built into the monitor and supplied both devices with power, the monitor and the computer. Besides the power cable, this setup required only one more cable from the monitor to the computer, as the keyboard had been built into the computer too. That was quite clever.
The CPC 464 featured up to 27 colours, the highest resolution (with only 2 colours, which means black-and-white then) was 640x200. The colours had a rather bright and clear look to them - quite different from the C64's colour palette for example.
Its sound chip, the "AY-3-8910", had three channels, stereo. Its sound wasn't that spectacular though. It could generate either square wave sounds or noise (combined with an envelope generator). It didn't have a filter. That chip could also be found in the Atari ST-computers and in the later 128K-models of the ZX Spectrum.
The CPC 464 came with "Locomotive BASIC", which seemed to have been a decent BASIC dialect.
There were quite a number of games for the CPC 464, many titles known from the ZX Spectrum or the C64.
For some business-reason, in Germany the machine was sold as "Schneider CPC 464". The company "Schneider" was otherwise known in Germany for selling greyish-looking stereo-systems, that were cheaper, but didn't sound as good as its competitors. (Much later, in 2002, Schneider filed for bankruptcy, but the Amstrad CPC was not the reason for that.)
The dark-greyish plastic case of the Amstrad computer and its monitor wasn't that beautiful.
I remember the speaker in the Amstrad's monitor being quite bad, resulting in an overall unpleasant plasticky and harsh sound output of the computer (example). According to Wikipedia, at least later versions of the CPC 464 had a headphone jack, that could be used to replace the monitor's awful speaker with an external one.
I didn't like the machine too much. But users seemed to be quite content with it.
To me it looks like Amstrad had watched for two years, what others had done - what they had done right and where they had gone wrong - and then built a computer with resonable specifications at an affordable price without taking too much risk.
Amstrad released several more versions and improvements of the CPC 464, for example the CPC 664 (3' floppy disk drive, better keyboard) and the CPC 6128 (128K RAM).
There also was a floppy disk drive for the CPC 464, which was called "DDI-1". The CPC computers with floppy disk drives could also run CP/M, beginning with version 2.2.
CP/M ("Control Program/Monitor") was a single-user operating system, that was first released in 1974. There were versions for Intel processors and the Z80, making CP/M programs interchangable between systems to a certain extent. CP/M was quite popular in the business market, until in 1981 IBM released their IBM PC, which used MS-DOS instead.
In 1986, Amstrad also bought Sinclair's computer range including the ZX Spectrum.
From today's perspective, some people see the Amstrad CPC 464 as the true main competitor to the C64. That needs a bit of explanation: Back then there were basically two 8-bit processors: On the one hand there was the MOS 6502, which was used in the C64 (as 6510), the Atari 800XL and the Apple II. And on the other hand there was the Zilog Z80, which was used in the Sinclair ZX Spectrum (Z80A), the Amstrad CPC 464, the TRS-80 and the MSX-Computers. So there were these two sides of the 8-Bit world. The C64 was probably the main representative of the 6502-side, while the Amstrad CPC 464 could be seen as an equally powerful home computer of the Z80 side.
But I've only heard about that kind of view in recent years. Back then in about 1984, I've never heard of anybody asking himself, whether he should buy a C64 or an Amstrad CPC 464 instead. The choice was more between the C64 and the Sinclair ZX Spectrum or between the C64 and the Atari 800XL. Mainly, because software was more easily available for these machines.
The Amstrad CPC 464 may have been a reasonable computer, but in its ugly grey case and with its bad speaker it was just not "cool". Maybe I underestimated it. But that was the way it was back then.
Today I'm quite impressed, how many games existed for it. So could it have been fun to own an Amstrad CPC 464? My feeling still says "no". Not more fun than owning a grey Schneider stereo-system (instead of a real one from a company like Akai, Technics or Denon). Sorry.
My second computer was the Atari 800XL, which was released in 1983. Having seen it at a friend of mine, I bought one in 1984/1985.
The company "Atari" ("Atari Inc.", to be specific) had been founded in 1972 in California. The name was a term from the Japanese board game "Go". Nevertheless "Atari Inc." was not a Japanese, but an American company. It produced the arcade game machine "Pong". In 1977, Atari released the hugely successful home gaming console "Atari VCS 2600" ("Video Computer System"). In 1979, Atari released the arcade game cabinet "Asteroids".
Atari 800 and 400 (1979)
Also in 1979, the developers at Atari were working on a successor to the "VCS 2600". They realized, that competition against gaming consoles from home computers like the Apple II and the TRS-80 was getting stronger. So they added computer keyboards to their devices and finally released two 8-bit computers called "Atari 400" and "Atari 800". They basically already used the same technology as the 800XL. Chief-designer of the integrated circuits was Jay Miner.
The Atari 400 and Atari 800 were solidly built and looked rather bulgy. Probably because they basically were 1979's gaming consoles (think of the design of the VCS 2600) with computer keyboards added on top. Compared to the 400 (which already had a decent size), the 800 was even much bigger and heavier (this video gives a good impression of the Atari 800).
The names "400" and "800" were originally indicating, that the machines were supposed to come with 4 and 8 KB memory. But at their release in 1979, it was already possible to provide both of them with 8K RAM.
The 400 only had a membrane keyboard, that was rather bad, maybe only a slightly better than the one of the Sinclair ZX81. There was one cartridge slot, a SIO port for a disk drive or a printer and only a RF video output. The Atari 800 on the other hand had a full typewriter keyboard, even two rather solid cartridge slots, another video output in addition to the RF one, and was overall more expandable. The right cartridge slot of the Atari 800 was only rarely used.
Both machines had four joystick ports.
The internal hardware followed the concept, that the CPU (6502, "Sally") was supported by several socalled "custom chips" ("Antic", "CTIA" and "POKEY"). That way, the performance of the machines regarding graphics, sound and I/O could be distinctly increased. In 1979, the Atari computers may have been the first ones to use such custom chips. Programmers, that wanted to develop more demanding programs or games for the Ataris, had to study the behaviour of this special internal setup in detail.
The BASIC interpreter for the 400 and 800 came on a cartridge, that had to be inserted into a cartridge slot. Atari wanted to buy "Microsoft BASIC" for their machines, but it wasn't possible to fit it on an 8K cartridge. A company called "Shepardson Microsystems" then created a custom BASIC version for Atari.
In the following years the RAM of the Atari 800 was raised to 16K and then 48K. The Atari 800 was rather expensive: 2.995,- DM in 1981 for the computer, and about another 2.000,- DM for the floppy disk drive "Atari 810". With these kind of prices the Atari 800 was more in the range of the Apple II (while the IBM PC still was even more expensive at the time).
In the U.S., Atari also managed to a certain extent to establish their 400 and 800 as school computers.
In 1981, prices of the Atari 400 (with 8K RAM) dropped from about 630 to about 500 US-Dollars. It cost about half of the Atari 800 then. A RAM expansion to 16K could be bought separately.
In 1983, Atari released refurbished models of their 8 bit line. Only very limited changes were made to the internal hardware. There were the 1200XL with 64K and the 600XL with 16K RAM. The relation of the model numbers 1200 to 600 were the same as 800 to 400 had been before.
But it turned out, that the 1200XL had technical problems and was too expensive. Customers even turned to the old 800 again, which had become more affordable in the meantime. So the 1200XL was replaced by the Atari 800XL, which then became commercially successful.
Atari 800XL and 600XL (1983)
The Atari 800XL was much smaller than the huge Atari 800 had been. It was also much more affordable, so that it competed against the Commodore 64. I remember prices of about 750,- DM for the Atari 800XL and another 750,- DM for the disk drive Atari 1050 in maybe 1984. Prices dropped even more later.
The Atari 800XL looked quite compact and elegant, with its beige surface and its brown-greyish typewriter keyboard. Somehow it looked a bit prosaic and "grown-up". Although its built-in slot for ROM-cartridges, its two joystick-ports and the silver-coloured range of special keys on its right side ("Reset", "Option", "Select", "Start", "Help") still showed its gaming heritage.
Like the Atari 400 and 800, the 800XL used a slightly modified version of the 6502-processor as the CPU (called "Sally"), running at 1.77 MHz on PAL machines, and at 1.79 MHz on NTSC machines. That was quite a bit faster than the C64's processor 6510 of 1982, which ran at about just 1 MHz. The Atari 800XL came with 64K of RAM (only 16K RAM in the Atari 600XL).
The usual screen of the Atari for BASIC or "Atari DOS" was very light blue (close to white) writing on a light blue background. These colours didn't have the pastel touch of the C64's, but looked quite clear instead.
Inside the Atari, graphics were produced by two custom chips called CTIA/GTIA and ANTIC, that worked together with the 6502-CPU. ANTIC generated the main graphics and text ("playfield graphics"). CTIA/GTIA provided colours and created sprites ("player/missile graphics"). CTIA was the older version with a palette of 128 colours. GTIA was built into the Atari 800XL and could provide a palette of 256 colours.
To load and save data, you could either buy the very slow dedicated cassette recorder "Atari 1010".
Or the floppy disk drive "Atari 1050".
In contrast to Commodore's VC 1541, disk access on the Atari 1050 was sufficiently fast.
I would say, loading a game that used all memory of the 800XL took about 20 seconds, maybe. I always was pleased with the 1050. There was nothing to complain about it, actually. Besides, that it was quite expensive. It cost about the same as the computer itself.
The disks used in the Atari 1050 could store about 130K of data (socalled "enhanced density"; the 1050 could only read and write single sided). The disk drive came with a disk containing "Atari DOS", in various versions (version 2.5 was quite popular).
The Atari computers used a special protocol called "SIO" ("Serial Input/Output") to communicate with the 1050 drive. Today, it is possible for retro-enthusiasts, to connect a real Atari 800XL to a PC using a special cable, and then make the PC emulate just the Atari's 1050 floppy drive ("sio2pc").
There is also a modern floppy emulation in hardware called "SDrive-MAX", which lets you connect the small modern controller "Arduino" to the Atari to emulate several Atari floppy drives. Actually, the Atari's "SIO" port can be seen as an early ancestor of today's USB. When connecting the "SDrive-Max", the Atari can also power the Arduino through the SIO-port.
There's also a socalled "Ultimate Cart", which is a modern cartridge that can be put into the Atari's cartridge slot, so that an SD card can be used as an Atari cartridge replacement. That's also an interesting approach, but I think, the "SDrive-MAX" is even more powerful, as the full features of several drives 1050 are emulated.
The "Option"-key on the right side of the 800XL had a special meaning: If you pressed it a start-up, the Atari's BASIC wasn't loaded from ROM.
Instead, the disk in the floppy drive was searched for a machine code program (probably a game) to be loaded automatically.
If in this case, the disk drive was not available, the Atari ran an internal program called "Self Test" instead, where you could test the Atari's hardware, like the memory or the sound-chip. Which was quite useless in the situation, and probably not what you wanted.
If the disk drive was connected, but there wasn't a disk inserted, the Atari quickly printed the words "BOOT ERROR" onto its blue default-screen over and over again:
BOOT ERROR BOOT ERROR BOOT ERROR BOOT ERROR BOOT ERROR BOOT ERROR
This was quite a strange behaviour, if you think about it. Users that weren't familiar with it, probably would have thought, the computer was out of order. The Atari hardly gave them a clue, what was going on.
If the "Option"-key was not held down on start-up, the drive was ready and a disk was inserted, the Atari would enter into BASIC. Some message about the BASIC-version would be printed on the screen, and then just:
READY
In Atari BASIC, you could use "?" for the "PRINT" command, and "L." for the "LIST" command.
"L. 100, 200" would list the BASIC code from line numbers 100 to 200.
If the disk with "Atari DOS" was present in the drive, you could enter
DOS
at the BASIC-commandline and get into Atari DOS. It showed a menu, from which you could select DOS-commands. For example you could choose "A" to view the directory of the present disk.
From Atari DOS, you could often return to BASIC by choosing "B". Atari DOS says, that "B" means activating the cartridge, but it seems, even in the Atari 800XL, BASIC is seen as a cartridge (it had been on cartridge in the Atari 800).
Unfortunately, Atari BASIC didn't have a dedicated command for displaying the directory of the present floppy disk (so you couldn't even do something like 'LOAD "$",8' on the C64). If you wanted to view the directory of the disk in the floppy drive, you either had to start Atari DOS and select "A", then go back with "B" (as desribed). Or you had to write a small program onto each floppy disk, that would achieve the task. Such programs were often named something like "CATALOG.BAS" (so now you know, what that command is for, if you see it on an Atari disk somewhere - back then, I saw it often, but didn't know, what it meant).
A rather nice solution for the problem was posted in this forum discussion.
Early versions of "Atari DOS" had already been developed before the release of the 800XL in about 1980, when the Atari 800 had been a quite expensive computer, that was also supposed to compete against the IBM PC and its PC DOS (MS-DOS) of 1981. So the Atari 800 (and also the 800XL) had been designed for both: For playing games and for running serious business applications. (Which again reminds of Jay Miner's following design, the Amiga, which also could be used for both tasks well.)
The Atari 800 had a slightly different ROM than the 800XL. If you wanted to play some of the older games, that had been developed for the Atari 800 on the 800XL, you had to load a program called "Translator" first, before you could load the game itself. On today's software emulators, you'd have to select the 800's ROM ("OsB") instead of the 800XL's ROM ("OsXL") to make these games work.
The original Atari 800 featured a palette of 128 colours (the Atari 800XL even of 256 colours). That many colours on a computer in 1979 were a really amazing technical achievement.
Cycling through all possible colours (256 or 128) was often used to produce nice rainbow effects.
Due to memory limitations, the Atari 800XL offered several graphics modes for different purposes. When booting to BASIC, a text mode with a resolution of 40x24 characters was used, that could only display 2 colours at once, so just plain text on a background of another colour (which makes two altogether). The characters used a matrix of 8x8 pixels, but most characters had a height of only 6 pixels. The characters looked quite nice, more "solid" and "professional" than that of the ZX Spectrum for example.
It was possible to switch to other graphics modes, also from BASIC. The highest resolution of the Atari was 320x192 pixels (in "graphics mode 8"). But in this resolution it was again only possible to display 2 colours (which in praxis is only 1 colour, as the other one is the background colour).
16 colours from a palette of 256 were possible in "graphics mode 9" (GTIA only), which had a resolution of only 80x192.
Some graphics modes reserved the main part of the screen for graphics, but also displayed a kind of "window" at the bottom for text. That text part could be switched off (in BASIC) by adding a "+16" to the "GRAPHICS" command. So often, something like "GRAPHICS 3+16" could be found in BASIC code.
For the ordinary "playfield graphics" games often used a graphics mode of 4 colours (from a palette of 256) in a resolution of just 160x192.
Often, the graphics of text characters were redefined to represent game objects.
The Atari 800XL could also generate sprites, that were independent from the rest of the graphics. As the "Atari 800" already could do this in 1979 (it probably was the first home computer that could), and the term "sprites" was not yet common, Atari called this technique on their 8-bit-computers "Player/Missile Graphics".
There were four "players" and four "missiles". Both of them were 256 lines tall, so both could fill the entire screen vertically. The "players" held horizontal information of 8 bits, while the "missiles" only of 2 bits. When you think of "Space Invaders", the space ship on the bottom was created as a "player", and the laser shots it fired as "missiles".
The pixels of the "players" and "missiles" could be scaled just horizontally. That way they appeared larger on the screen. But their horizontal image information still remained 8 bits / 2 bits, so their pixels became more visible (lower resolution).
At the same time, on the vertical axis there was enough image information for (up to) the whole screen (256 bits). Sometimes these circumstances led to kind of weird looking graphics. For example, the resolution of the player's "player" in the game "Pogo Joe" had a size of 28x48 pixels, with vertically 48 truely different pixels, but horizontally only 7 different pixels, that were scaled up by factor 4 (to 28 pixels).
In general, "players" and "missiles" only featured one colour from the palette. But two "players" could be combined into one to get three colours (one from each "player", plus a third colour mixed of the other two).
The four "missiles" could be combined into a fifth "player". In "Pac-Man" on the Atari 8 bit for example, there were 4 ghosts (= "players"), and Pac-Man as the fifth object, which was probably a fifth "player" combined from the four "missiles" (or one of the ghosts was). The "missiles" weren't needed as shots, as there wasn't any shooting in Pac-Man.
The Arcade game "Pac-Man" wasn't an original Atari game. It had been released by Namco/Midway in 1980. But Atari produced a ported version for their "VCS 2600" under licence from Namco and often bundled it on a cartridge with their console. The VCS version was in a way different from the arcade game, and today people say, it's been a bad portation. But back then, as a child, I liked it and was happy, when I could play it in stores or at friends. It had a characteristic sound, and for Atari, it was a commercial success.
The "Player/Missile Graphics" on the Atari 800XL were independent from the "playfield graphics". That also meant, they were independent from the graphics mode. They always operated with a resolution of "320x256". As described, with the limitation of horizontal information of just 8 bits (2 bits for "missiles") and the possibility of horizontal scaling.
Also the colours of the "Player/Missile Graphics" were independent from the playfield and the graphics mode. So, while the playfield was for example limited to only four colours (from a palette of 256) due to the graphics mode, the four "players" added four more colours.
It was also possible to make the graphics chips change the colour of the screen per horizontal scanline. This was done by writing interrupts to the ANTIC's "display list" (in assembly language). That way, additional rainbow effects could be created on the playfield.
In detail, that worked as follows: The vertical position of the scanline beam was stored. Then, ANTIC told the CPU to go to sleep for a few milliseconds, until the horizontal end of the scanline was reached. Just in that moment, when nothing was drawn, the CPU woke up and wrote the previously stored value of the beam's vertical position into the colour register. The graphics chips then read that register and used its colour value to draw the next scanline of the screen.
The result was a kind of "rainbow". But it couldn't be just a rainbow: If the colours were set carefully, it could also be a sunset or the different colours of daylight. This effect was for example masterly used in Activision's game "The Great American Cross-Country Road Race". In this game, the player took part in a race across the American continent, which took several days. With the Atari's large colour palette, night, sunrise, daytime, snow, fog, sunset and other natural processes could be simulated beautifully.
The 800XL had a sound-chip called POKEY (which also controlled data-I/O and the keyboard). It produced four sound voices based on square waves, that each could play a different synthesized sound. Some distortion effects could be added. Two voices could easily be used to produce a chorus effect by slightly detuning them (see the BASIC example below). POKEY also had a high-pass filter. The Atari's sound was a bit simpler and "cleaner" than that of the C64. But to my taste, it didn't sound bad.
The song "Passionately" or the intro music of the games "Ballblazer" or "Mr. Robot and His Robot Factory" may be good examples of POKEY's sound.
Actually, often classical pieces of music were used in computer games. That was a good idea, because it was of very high quality, and also its copyright had already expired. It could give a game a somewhat "noble" touch, that I liked. Played in a yet unheard POKEY-version, the music even sounded fresh and interesting.
For example, the Atari's internal "Self Test" and the game "Spelunker" used Mussorgski's "Pictures at an Exhibition", a ("Frogger" like) game called "Preppie" used Anton Dvorak's "Humoresque", the game "Mountain King" used "In the Hall of the Mountain King" and "Anitra's Dance" by Edvard Grieg.
The melody of Grieg's "In the Hall of the Mountain King" had also appeared in "Manic Miner" on the ZX Spectrum; while its successor "Jet Set Willy" played a simplified version of Beethoven's "Moonlight Sonata" in its intro.
So in a way these computer games even introduced us kids to classical music. The first time, I heard these tunes, was in a game. Sometimes, later I even was surprised, when I found out, that these tunes originally hadn't been composed for the game, but were cultural heritage.
There weren't so many outstanding musicians working with the POKEY as with the C64's SID.
In this regard, the name "Kemal Ezcan" comes to my mind. He created music and sound for a number of Atari games.
Today, he calls himself "Yoda Zhang" (having been married to a Chinese wife (as I read in a forum)). Recently, he has created a couple of browser games, that still remind me of the Atari. And again, especially the sound of the games is pretty good. Check out "Bomber Jack" for example (the game also has music; try reloading the website, if you don't hear it).
When not pressing the "Option" key at startup, the Atari entered its BASIC, which was loaded from internal ROM (just like on other computers like the C64 or the ZX Spectrum). Atari BASIC had more commands than the C64's BASIC. There were for example commands like "PLOT" and "DRAWTO" to create simple graphics, and the command "GRAPHICS" to switch between the different graphics modes. There was also a "SOUND" command to create sound and music, accessing the sound chip POKEY.
But as POKEY had some more advanced features like a high-pass filter, four arguments had to be passed to each "SOUND" command.
Here's a simple example how to get that "chorus"-effect of two slightly detuned POKEY channels, which was often heard in games, in Atari BASIC:
10 REM Initialize sound: 20 GOSUB 500 30 REM Produce chorus sound using channels 0 and 1: 40 REM First argument is the sound channel, 50 REM second argument is the pitch, fourth argument is the volume: 60 SOUND 0, 253, 10, 9 70 SOUND 1, 255, 10, 9 100 REM Create a pause to hear the sound: 110 FOR I=0 TO 1500: NEXT I 120 REM Stop the sound: 730 GOSUB 500 140 END 500 SOUND 0, 0, 0, 0 510 SOUND 1, 0, 0, 0 520 RETURN
But the different graphics modes and the socalled "display lists" of the Atari's ANTIC-chip weren't so easy to understand, and when you wanted to program "Player/Missile Graphics", which was also possible in BASIC too, you finally had to get into "PEEK" and "POKE"-commands.
So the Atari's BASIC wasn't bad, it was much better than the C64's, but it still wasn't as inviting as the ZX Spectrum's. Because the system itself with its graphics- and sound-chips was more complex. There was a book called "De Re Atari" which gave a detailed explanation of the Atari's system, but with its many diagrams and tables it was also a bit intimidating to be honest.
The Atari 800XL could also be programmed in other, more advanced programming languages than BASIC. There was for example "Atari Pascal", "Action!" (an Atari-specific compiled procedural language sold on cartridge by the company "Optimized Systems Software" (OSS) since 1983) and of course assembly.
But in the end, the Atari's system was over the head of most users, also because in 1983 there wasn't the internet to retrieve specific information. So most people just pressed "Option" at startup, skipped Atari BASIC and just started their games or other programs. Which probably was, what the designers of the Atari had anticipated, when they had built in that unusual function of the "Option"-key.
There was a printer "Atari 1027", that could only print text (20 letters per second) and unfortunately often wore out after a while. And there was a plotter "Atari 1020", that used paper, that was horizontally smaller than letter-size, and looked like Atari's cassette recorder ("Atari 1010"). I didn't have any of these.
Nevertheless, all these peripherals (including the floppy drive 1050) were designed in the same colours as the 800XL itself (beige, grey-brown, silver), which looked rather nice together.
In 1983, Atari also suffered from the socalled "video game crash" and was close to bankruptcy in early 1984. It was then restructured by Jack Tramiel as "Atari Corporation". Its main interest was the 16 bit Atari ST-line then. But Atari still had the 8 bit technology too. In 1985, Atari designed different cases (and keyboards) for these machines, that looked similar to the greyish ones of the ST line. The most popular of these computers was the Atari 130XE.
Inside it still had the architecture of the Atari 800XL, but it had the new look and 128K RAM. But the XE-line didn't change much on the market.
By then, it was just another offer in the lower price range, while Atari's main product was the ST-line.
After all, the Atari 800XL had been the main competitor to Tramiel's C64 in the years before. So it must have been a weird situation for Tramiel, having to sell the Atari XL/XE now, competing against his own C64.
It's interesting, that also Commodore gave the C64 a new look in 1986 ("C64C").
Since 1988, a program called "SpartaDOS" was developed, that - unlike Atari DOS - used a command-line prompt similar to MS-DOS. Later versions of the program were called "SpartaDOS X".
The Atari 8 bit computers were produced from 1979 to 1992. There had been the models Atari 400, 800 (1979), 5200 (gaming console, 1982), 1200XL, 600XL, 800XL (1983), 65XE, 800XE, 130XE (1985), XEGS (gaming console, 1987). It seems, about 3 to 4 million units have been sold worldwide.
Overall, the Atari 800XL was a quite similar machine to the Commodore 64. Both had:
You could write programs in Atari BASIC, do some calculations or even run more serious applications.
There was a word processor called "AtariWriter", released by Atari itself in 1982, that became quite popular. Like BASIC on the Atari 800 (1982), it came on a cartridge, so it was available without loading times (which in that case was actually quite useful).
And there even was a version of "VisiCalc" for the Atari, which was the most-used spreadsheet-application of the time.
But most people used the Atari 800XL for games. There were improved versions of a lot of games originally created for Atari's successful console "VCS 2600". But also many other titles were available.
Overall, a few more game-titles may have been developed for the C64. But there sure were enough games to play on the Atari 800XL.
When I think of it today, the Atari 800XL/130XE may have been the ideal 8 bit computer.
Its processor ran quite a bit faster than that of the C64 and the Apple IIe (1.79 MHz on the Atari compared to about 1 MHz on the other two). The Atari had solid, colourful graphics and a decent system font. It had a nice keyboard and the common joystick ports. The SIO-port to external devices was even ahead of its time. The Atari could be used for calculations and programming as well as for playing games. It had a reliable and fast floppy disk drive. And it wasn't too expensive.
I also like the sound of the POKEY chip: Everybody was talking about the C64's "SID" chip and its filter, but I didn't like the harsh and dirty sound it produced much. It may have been able to emulate an electric guitar (in the game tune of "Wizball" by Martin Galway for example), but I'm not much interested in that. I also didn't like the sound of the AY chip of many other computers either (and I'm not too happy that that one reappeared in the ZX Spectrum Next). So, when it comes to synthesized music and game sound on a home computer, even today I can't think of many sound chips that produced nicer results than the humble POKEY chip of the Atari.
In 1987, Tramiel's Atari Corporation used its 8 bit line technology once more to release a gaming console called Atari XEGS ("Atari XE Gaming System"). It was grey and looked a bit like an answering machine.
The typical buttons "Reset", "Option", "Select" and "Start" were round and of pastel colours. They could be found on top of the machine in the front.
Inside, the Atari XEGS was just an Atari XE with a memory of 64K RAM (like the 65XE). As a gaming console, the XEGS was meant to run cartridges. A few dedicated ones were developed, but it could also run cartridges created for the XE/XL computers.
In the "deluxe set", the XEGS came with a special keyboard, that turned the machine into a XE computer again. The XEGS also had a SIO-port, so that again a floppy disk drive could be connected.
The XEGS wasn't a commercial success.
It was interesting in a sense, that it showed, that the Atari 8 bit computers had been gaming consoles with added computer functionality from the beginning.
Today, there's an interface called "Trans Key II" to connect a PS/2 keyboard to the XEGS. And with things like the "Ultimate cartridge" or the "SIDE2 cartridge", the XEGS may be more fun today than it was back then.
Software emulators of the Atari 800XL are called "atari++" and "Atari800".
For some strange reason the emulator "Atari800" doesn't get the Atari 800XL's colours right. Although the used colours don't look too bad either and the rest of the emulation is nearly perfect.
In "atari++", the "BREAK" key of the Atari, that is used to break the execution of BASIC programs (or "LIST" commands), is emulated by the "F8" key on the PC keyboard.
Today, it is also possible to write programs in C or assembly language for the Atari on a PC using the cross compiler "CC65". A simple "Hello World"-program in C called "hello.c" was already shown in the chapter about the C64. For the Atari, it would compile with a small script like this:
#!/bin/bash export CC65_HOME="/usr/local/share/cc65" # Set environment cl65 -O -t atari hello.c -o hello.obj
You can then use a small utility program called "franny", to write the file "hello.obj" to an Atari ".atr" disk-file containing AtariDos 2.5 with:
franny -A dos_2_5.atr -i hello.obj -o HELLO.OBJ
Then you can launch an Atari emulator with "dos_2_5.atr", select "L", type in "HELLO.OBJ" and run your program. With
export CC65_HOME="/usr/local/share/cc65" cc65 -O -t atari hello.c
you get a file called "hello.s", that contains a listing of the generated assembly code.
It all works very similarly to the process for the C64, and also the same compiler is used. It shows, that both machines used the same processor (variations of the MOS Technology 6502).
It is also possible to write Atari BASIC code on a modern PC, and feed that code into an emulated Atari. To do that, first make a backup of your BASIC file on the PC. Then convert it to ATASCII using the Python script "aac.py". Then use "franny" to write the resulting file as "PROGRAM.BAS" to an Atari ".atr" disk-file. Start an emulator, telling it not to hold the "Option" key at startup (to get into BASIC). Then run this command from Atari BASIC:
ENTER "D:PROGRAM.BAS"
This should load the code into the emulated Atari.
And this would be the other way round. That is extracting an Atari BASIC file from an "atr"-image file on Linux, using atari++, atrcopy and aac.py:
LOAD "D1:PROG.BAS"
in BASIC to load the BASIC program.
LIST "D1:PROG.LST"
in BASIC to save the listing of the program to the DOS disk.
atrcopy dosc.atr list
on the Linux shell.
atrcopy dosc.atr extract PROG.LST
You should have a file "PROG.LST" in your directory now.
aac.py -u PROG.LST prog.lst
to get a version "prog.lst" of the BASIC program, that can be read on the PC.
Back then, I didn't know it, but it all started with the Apple II.
(To be exact, it started with the "Apple I", that was sold as a raw (populated) circuit board in 1976. You can find out more about the "Apple I" here.)
The Apple II had been designed by Steve Wozniak and was first released in 1977.
It used a 8-bit 6502-CPU, running at about 1 MHz. It was solidly built, had a proper typewriter keyboard and a joystick port. In 1977, it came with 4K RAM, which over the years was expanded to 64K or later even to several MB. With a separate RF-modulator, it could be connected to a NTSC-television.
An audio cassette recorder could be used for data storage. Soon there also were 5.25'' floppy disk drives.
"Integer BASIC", also written by Steve Wozniak, was loaded from ROM at start-up.
There were several graphic modes. The Apple II featured 16 colours, but it originally didn't send out a real color signal, but a pixel pattern that produced certain color artefacts on NTSC televisions. So when using the Apple II with another monitor or a PAL television, not colors were shown, but these patterns of pixels. For the European market, there was an extension card to add PAL color support to the Apple II.
The graphics system including colours was improved in later models.
Even the first Apple II models came with eight expansion slots.
Although the joystick port looked like a standard DB9-port, it couldn't handle digital joysticks like for example the Competition Pro. Instead, a special analogue joystick had to be connected.
The keyboard of the first Apple II could only be used to type upper-case characters. The software, too, could only handle upper-case characters at first.
When I think about it, maybe that was the reason, other home computers like the C64 or the Atari 800XL mostly used upper-case characters in their environments too. But maybe that was just an oddity of BASIC.
The Apple II's hardware was an open system, the documentation of its design was published. In the early models only standard-components (like the 6502-chip) were used. Therefore, a lot of clones of the machine were produced, especially in the Eastern Bloc. The ROM code of the Apple II was closed, legally protected software though, and Apple sued many producers of clones for copyright violations. Notable Apple II clones were called "Franklin Ace 1000" (1982) or "Laser 128", "Laser 128EX" (around 1986).
Original prices of the computer in 1977 were $1,298 (4 KB RAM) and $2,638 (up to 48 KB RAM).
According to the Vobis list of July 1983, the price of an original Apple II in Germany was 2.385,- DM then, plus 1.285,- DM for a disk drive or 2.698,- DM for a double disk drive (original Apple models), plus another 350,- DM for adapters (including a RF modulator) to connect the computer to a television (otherwise you had to buy a monitor).
Vobis also offered an all-in-one set including a clone of the Apple II (48K), a green 12 inch monchrome monitor, a compatible Siemens floppy disk drive (individual price 775,- DM) and a printer, at a price of 3.750,- DM altogether.
As the Apple II was highly expandable, you could always upgrade the computer to what was possible at the time.
The "Mockingboard" for example, was a soundcard, that added the "AY-3-8910" to the Apple II, a sound chip with three channels, that was also used in several other home computers.
Concerning sound, Apple always had to be a bit careful, as there also was the music company "Apple" of the Beatles, and there was an agreement, that "Apple Computer" was allowed to use its company name, as long as it would stay in the computer industry, and not reach out to the music-industry (I don't know, how they later handled "iTunes" in this regard).
In 1983, Apple released a revised model of its computer called "Apple IIe". The "e" in the name meant "enhanced". It came with 64K RAM, could also handle lower-case characters and mended most other imperfections of the first years. For example, built-in support for color graphics in PAL models (Europe) was added. By using a special switch on the down-side of the keyboard, a local character set (for example a German one) could be activated as an alternative to the default US ASCII character set.
An expansion card was sold separately, that made 80 columns of text per line possible on the Apple IIe, and also expanded the memory to 128K.
The Apple IIe had a few smaller revisions during the years and was produced until 1993.
In 1986, there also was the "Apple IIgs", but it was a totally different, 16-bit machine with a graphical user interface, although it also had an 8-bit "emulation mode" to become compatible with the older models.
The original Apple II-models were especially popular in American schools.
The Apple II was also popular as a games machine. A lot of titles were produced, for example in 1980 the first (still black-and-white) graphics adventure game called "Mystery House", the first game of the company "Sierra" (Roberta and Ken Williams), or in 1985 the strategy game "The Oregon Trail".
The Apple II was also designed to be used as an office computer. The first spreadsheet-programm "VisiCalc" (1979) was produced for the Apple II. An integrated office suite called "AppleWorks" (combining word processor, database, and spreadsheet) for the Apple II was released in 1984.
At my school, there was a teacher of maths and physics, who tried to be funny. He used to point out, that the Apple II was a far superior machine to the at the time much more common Commodore 64. So when he had graded physics tests, he entered the class with the bad tests in a Commodore plastic bag, and the good ones in an Apple plastic bag.
I didn't feel addressed, as I was a ZX Spectrum user.
His opinion wasn't correct either: Both, the Apple IIe and the C64 used a 6502-type processor (C64: 6510), running at just about 1 MHz. With its colourful hardware sprites and the SID-chip, the C64 was probably superior, when it came to graphics and sound. And it was available at less than half the price: About 1.600,- DM for a C64 with floppy drive, compared to about 4.000,- DM for an original Apple II (48K) with floppy drive and television adapter in 1983. The Apple IIe was more expandable, but ordinary people like pupils or teachers didn't need that.
So you couldn't really say (objectively), that the Apple II was superior to the C64. Maybe that teacher just liked it better (subjectively). But the extremely high price of the Apple II put it out of reach of most people anyway. At these prices people just couldn't be expected to own such a device.
The differences between the 6502-based 8-bit home computers weren't as big as they were perceived in the 1980s. Apple II, Commodore 64, Atari 800XL, BBC Micro - they all could do similar things. With only slight variations. What differences do you expect from machines all using a 6502-processor, running at 1 or 2 Mhz, connected to televisions, accessing 64K RAM, cassette tape or maybe 5.25'' floppies? It's basically the same idea.
I think, it was Steve Wozniak who originally had this idea. Without doubt he was a genius.
On the other hand, those systems were built around the CPU. It were processors like the 6502 or the Z80, that made affordable home computers possible. And the processors were created a few years before the computers, the 6502 in 1975. So you could say, the idea of those processors already included the idea of a home computer. And there had been an early attempt to build one by Commodore in 1976, called the "KIM-1". Nevertheless it was Wozniak who came up first with a really usable one. His setup inspired all later approaches.
I've never seen a real Apple II. They weren't displayed in the ordinary shops. Probably Apple already had its own specialized distributors back then. But there wasn't one in my vicinity. You had to search for it. I didn't do that, as I knew, I couldn't afford Apple products anyway.
The Apple II computers were mainly popular in the United States. In Europe, they were actually quite rare. In Germany, the Commodore 64 was used instead. In Great Britain, also the BBC Micro was popular.
In the (fictional) TV-series "The Big Bang Theory" (season 4, episode 2, 2010) Sheldon Cooper talks to Steve Wozniak, who is sitting at a table in a restaurant. Wozniak offers to sign Sheldon's 1977 Apple II, but when Sheldon tries to fetch it from his apartment quickly, he stumbles, and the computer breaks.
After the huge success of the C64 in 1982/1983, Commodore (led by Jack Tramiel) thought of producing another machine for the lower price range. And another one addressed at business users.
Costs should be reduced by integrating the hardware for several tasks into one chip, especially for graphics and sound. So Commodore's subsidiary MOS Technology developed a chip according to this specifications called TED ("Text Display").
When Tramiel left Commodore in 1984, the new management made other decisions than originally planned, and the result of the developments were three computers using the TED-chip called Commodore 116, Commodore Plus/4 and Commodore 16.
The 116 had rubber keys and just 16K of RAM. It was meant as an entry level home computer and was supposed to compete especially against the Sinclair ZX Spectrum in Europe. The 116 was supposed to be sold at $49.
Just 16K RAM in the year 1984 was quite low though and wasn't accepted well in the market, as RAM had become more affordable in the years before.
The Plus/4 was meant as a business machine. It also had a small case, but a typewriter keyboard and 64K RAM. The 116 and the Plus/4 were smaller than the C64.
The Commodore 16 had the same features as the 116 (including the very low amount of just 16K RAM), but came in the familiar "breadbin" case of the C64 and VIC-20, with typewriter keyboard. It was meant as a successor to the VIC-20. Its case was black, so it looked like a black C64 - but it wasn't.
All three machines used the MOS Technology 7501/8501 processor. The TED chip could make it run at about 1,77 MHz, which was quite a bit faster than the C64's processor.
The graphics resolution was up to 320x200 with a palette of 121 colours (C64: Just 16 colours).
The BASIC was improved in comparison to that of the C64. Now there were commands like "DRAW" to create graphics on the screen and better commands for floppy disk access.
But the 16, 116, Plus/4 computers didn't have hardware sprites. And the sound produced by the TED was much simpler than the one of the C64's SID: There were just two square wave voices (that could also produce noise).
The Plus/4 came with a business software suite called "3-plus-1" on a hard-wired ROM chip inside the machine. But the software wasn't that good, that it convinced business users of the qualities of the machine.
The 16, 116 and Plus/4 computers were not compatible with the C64 (or the VIC-20). So it couldn't be made use of the huge software base, that by then had been established for the C64.
The computers also had different, non-standard Mini-DIN-joystick-ports. The reason was to reduce size on the small 116, but all three computer models came with these ports (backside of a Plus/4). Commodore produced a joystick that worked with these ports, but it just remained exotic. In the end, customers had to buy a special adapter to be able to use a standard C64-joystick.
For data transfer Commodore datasettes and floppy disk drives could be used, but again not all connectors were compatible. A VC 1541 floppy drive could be connected, but was as slow as it ever was.
Many game programmers started developing games for just 16K again, to make the games run on all three machines, not just on the Plus/4.
The computers were offered at a higher price than originally planned. At one point the Plus/4 cost about the same as the C64, but wasn't the better machine.
The machines didn't have a good reputation at (us) youngsters either. The general feeling about them was: "Oh, a Commodore computer?" "Yeah, but it's not a C64." It really was quite disappointing, to us it felt like Commodore had created a cheap imitation of their own brand.
Maybe Commodore hadn't communicated well, what they wanted the machines to be.
The Commodore 16, 116 and Plus/4 computers were a commercial flop and were discontinued in 1985/1986.
I remember the last ones being sold off cheap at the grocery store "Aldi".
Commodore concentrated on producing and selling C64s again.
The "xplus4" command of the VICE emulator starts emulation of a Commodore Plus/4.
When the "Plus/4" failed, Commodore realized, they needed another successor to the C64. One, that (unlike the "Plus/4") had to be compatible to the C64. And they needed it quickly. So they developed the Commodore 128, which was released in 1985.
The C128's keyboard and case actually looked quite nice and modern, a bit like a slim Amiga 500.
Inside, it was a computer that was a C64 - and also something else. Which made the C128 a rather complicated machine.
Unlike most other computers, the C128 had not just one, but two different graphics chips:
Therefore, the C128 could operate in three rather different modes:
When you switched on the C128, the machine was in the first mode mentioned. It showed a pastel-coloured dark grey screen with a light-green border. On the screen was written in the same colour as the border's:
COMMODORE BASIC V7.0 122365 BYTES FREE
(C)1985 COMMODORE ELECTRONICS, LTD.
(C)1977 MICROSOFT CORP.
ALL RIGHTS RESERVED
READY.
In this mode the computer used more RAM than the C64. But also the BASIC was improved.
Now, there were BASIC-commands for creating graphics, like "GRAPHIC" (to select the graphics mode), "COLOR" and "DRAW", and even a "SPRITE"-command for controlling sprites.
There were also commands for creating sound and music, like "SOUND" and "PLAY".
The joysticks could be queried using the "JOY"-command. Non-blocking key-reading had already been possible before with the "GET"-command.
There also were better BASIC-commands for accessing the floppy disk drive, like "DIRECTORY", "DLOAD" and "BOOT" (which replaced the earlier 'LOAD "*",8,1').
There was a special (locking) key called "40/80 DISPLAY" in the top row of the keyboard, used for switching between 40 and 80 columns of text. When you pressed that key and did a reset afterwards (the C128 had a proper "Reset"-button on its right side), the C128 switched between the first and the third mode mentioned above. So that the RGBI-graphics-output was used, and you didn't get any signal on the composite-output any more.
At the end of this chapter will also be explained, how switching to 80 columns mode can be realized in the emulator "Vice".
Ideally, you had the monitor "Commodore 1084". Then you could connect both graphics outputs of the C128 to that monitor and switch comfortably between both signals by pressing a special key on the monitor (Later, I had that monitor for my Amiga 500 and didn't understand what that key was for. Now I know).
When you were in the BASIC-environment, either in the first or in the third mode, you could type the command:
GO 64
or
GO64
which was the same thing (because space characters could be omitted in Commodore BASIC, as mentioned above).
This would switch the C128 into the second mode and turn it into a C64. You could also hold the "Commodore"-key on start-up to achieve this. In this mode, the C128 was fully compatible to a C64. It basically "was" a C64 then, which used the ordinary composite graphics output.
Inside, the C128 mainly used the processor MOS Technology 8502/8580. You could set the processor speed (even in BASIC) with the commands "FAST" to 2 Mhz and with "SLOW" to 1 Mhz. In 80-columns-mode that worked as expected. But in 40-columns-mode, you lost the display signal to the composite output again, if you used the "FAST"-command. Because the VIC-II-graphics-chip was limited to the processor running at just about 1 MHz. But you didn't get an error, the screen just went blank. You could even type in "SLOW" again, to get the display back.
So using processor speed of about 2 Mhz just made sense in 80-columns-mode. Which was only used by people who had a suitable monitor.
The C128 did not only have two graphics chips, it also had two different processors:
Surprisingly, besides the 8502 there also was the processor Z80A inside the C128, mainly for switching the computer into its different modes.
When pressing the "Commodore"-key at start-up, the Z80A took care of booting the 8502-processor and the rest of the system into C64-mode.
A side effect of having a Z80A-processor in the C128 was, that it was also possible, to use the operating system "CP/M".
In 80-columns-mode, it could be loaded from a floppy disk called "CP/M-Plus" (Version 3.0), that came with the computer. Technically, the 8502-processor was switched off then, and the Z80A processor took over. So that was a fourth mode, in which the C128 could operate.
But in the end not too many people were interested in running CP/M. In 1985, it was already rather outdated, as on PCs it had been replaced more and more by "MS-DOS" since 1981.
As the C128 came with a complete set of compatible hardware of the C64, it also had the SID-chip.
Commodore finally also fixed the speed problem of the floppy disk drive by releasing the updated drives Commodore 1570 (5.25'' single-sided), 1571 (5.25'' double-sided) and 1581 (3 1/2').
Unfortunately, in C64-mode the old speed problem occured again.
In this mode, the machine also fell back to just 64K RAM and to the original C64-V2-BASIC.
In 80-columns-mode more screen data had to be stored. The limit of 128K memory was soon reached, and there were also problems with the communication-speed between the active processor and the VDC-graphics-chip. So compromises concerning graphics had to be made. In 80-columns-mode, the C128 mostly displayed lots of text in relatively high resolution, but with just a few colours and not too many graphics. There also weren't any hardware sprites in this mode. As a result, in this mode the screen looked more like the ones of business-computers of the time.
So programmers hardly developed any games for 80-columns-mode. But in the 40-columns-modes the C128 wasn't faster than the C64 (just about 1 MHz, if you didn't want to lose the display). Therefore, programmers just kept developing C64-games and reached the users of both machines with them.
About 90% of the people who had a C128 just used it as a C64. Maybe only about 10% used the 80-columns-mode. And just about 1% used the CP/M-mode.
In 1986, there also was the "C128D", a desktop-computer with a separated keyboard, that looked more like an Amiga 1000 or 2000.
The Commodore C128 was the last 8-bit-computer. In fact, in 1984 Commodore had already acquired the 16-bit technology of the Amiga. In 1985, Commodore released the Amiga 1000 as well as the C128. But it took Commodore until 1987 to release the Amiga 500 as a home computer model of the Amiga series. In the meantime, there was a lot of competition by the Atari ST-range of computers, which had reached the market earlier. So, commercially 1985 to 1987 were difficult years for Commodore, and the C128 was meant to by-pass the time, until the Amiga home computer was ready to be released.
The C128 was produced until 1989, and 4 million units were sold. So, commercially it was a success. But mostly due to being compatible to the C64.
But when you think of it: The original C64 really wasn't that great, when it came to doing work, such as writing longer texts for college.
You then also better used a dedicated monitor instead of a television set.
For productivity, the C128 offered
So it was much more suitable for productivity than the C64. You just didn't have fancy graphics and games, when you used these features. But you could use the other modes for that. So you could use the C128 for:
Actually, altogether that doesn't sound like a bad computer.
The program "VICE" can fully emulate the C128 (emulator command: "x128"), including its 80-columns-mode.
To switch to 80 columns mode in the emulator "VICE", do the following:
In 1984, several changes happend in the computer-industry:
Apple had gone beyond its "Apple II"-range of computers and had built the "Macintosh", a computer, that used the Motorola 68000 as its CPU (a 32 bit processor with a 16 bit data bus). It presented a "Graphical User Interface" (GUI), that was controlled with a mouse.
A remarkable commercial for the Macintosh, directed by Ridley Scott, was shown during the break of the Superbowl and became famous. Its topics were individual freedom and the novel "1984" (by George Orwell). In the end it said, 'On January 24th, Apple Computer will introduce Macintosh. And you'll see why 1984 won't be like "1984"'.
The first Macintosh had just 128K RAM, a disk drive for 400K-disks and a strange built-in 9'' black-and-white monitor. It was offered for 2.495 $. Later models with more RAM and a hard disk could cost even several thousand dollars more.
As announced in the commercial, the Macintosh was introduced by Steve Jobs on January 24th 1984 on the (winter) "Consumer Electronics Show" (C.E.S.) in Las Vegas.
In Britain, Sir Clive Sinclair tried his luck, to build a computer based on the Motorola 68008 with the "QL". He managed to produce such a machine, but it wasn't a commercial success.
Also in January 1984, just a few days before the C.E.S., Jack Tramiel suddenly had differences about how to lead Commodore with Irving Gould, the primary shareholder and chairman of the company, who had cooperated with Tramiel since 1966. As a result of the dispute, Tramiel left Commodore.
Strange enough, in July 1984 Tramiel bought the consumer division of Atari, including the video console- and computer product range. He completely restructered the company and refounded it as "Atari Corporation".
Jay Miner (1932-1994) (the image shows him in 1990) had been the chief designer of integrated circuits at Atari since the late 1970s. He had been responsible for the chips inside Atari's successful products, such as the gaming console "Atari VCS 2600" or the "Atari 800" and "Atari 800XL".
In about 1982, Miner felt, the time had come, he could construct a computer of the next generation. But at Atari he wasn't given the opportunity to do that. So he and a few others left Atari, and Miner became vice-president of a small company called "Hi-Toro", that he founded, together with Dave Morse (David Shannon Morse, 1943-2007).
A capital of 7 million dollars was raised. Main investors were a group of dentists from Florida.
At Hi-Toro, they started designing the new computer with the code-name "Lorraine" (which was also the name of the wife of Dave Morse, the company's president), based on the Motorola 68000-processor.
To hide, what they were really doing from the other computer-companies in their vicinity in the Silicon Valley, and to raise some money for their company, they also produced accessories for already existing computer-systems, such as joysticks.
The people working at Hi-Toro were very creative and highly motivated to build the best computer they could imagine. Often they were paid with shares of the company and had lots of freedoms at work.
When things finally took shape, the computer's name was changed to "Amiga" and the company's name to "Amiga Corporation".
"Amiga" was quite a strange name for a home computer, as it meant "female friend" or "girlfriend" in Spanish. Nevertheless, the name sounded good and was generally accepted.
In 1983, also Atari had invested in Amiga, before a crash of the sales of video consoles (like the "VCR 2600") happened that year.
To gain more investors, the developers of Amiga had to present on the "Consumer Electronics Show" in January 1984, what they had done. The chips weren't ready still though. So they built a prototype of the Amiga, made of a larger amounts of main-boards and various other electronics (that didn't even look like a computer). This prototype was able to run an early version of the "Boing Ball" demo, showing a huge white and red ball bouncing around on the screen. The demo caught the attention of the audience.
Jay Miner, who had already invested in Amiga Corporation personally before, mortgaged his private house to be able to put more money into the company. He believed in the company's value, because he knew, it was working on a good product.
But in spite of all efforts, in the summer of 1984, Amiga Corporation ran out of money, so it had to be sold.
It seemed, the only offer came from Jack Tramiel, the new owner of Atari. He just wanted to buy the Amiga-technology at low cost and then use it in his own company.
A loan of 500,000 dollars was given to Amiga Corporation, that had to be paid back to Atari within a month, otherwise Atari was allowed to take over Amiga (as far as I've read, the loan had been part of an earlier agreement between Atari and Amiga, that Tramiel had found in Atari's documents, when he took over the company). Tramiel also offered one dollar per share to buy Amiga Corporation, which was rather disappointing to the people, who worked at Amiga and therefore held company shares.
But short before time ran out, to pay back the loan to Atari, there was another offer by Irvin Gould of Commodore: 4.25 dollars per share. The people at Amiga agreed to that price. So, Amiga Corporation and the Amiga technology were sold to Commodore for all in all 27 million dollars, and the loan of 500.000 dollars was repayed to Atari in time. Also the dentists got their return of investment.
This is basically what happened, though I simplified the story a bit. If you care about details like contracts or lawsuits during the process, there are other resources to read.
Of course, Commodore made a lot of changes to Amiga. The wild times of "Hi-Toro" were over. But a spark of the spirit of the first developing years always remained inside the Amiga-products.
At the CES in the summer of 1985, Commodore was finally able to present the "Amiga 1000". It was a desktop-computer with 256K RAM. It had signatures of Jay Miner (who can be seen as the "father of the Amiga"), other developers and even Miner's dog "Mitchy" inside its case. That was funny, but there already had been similar case-signatures inside the first Apple Macintoshs.
It seems, it never was quite clear, if the Amiga was supposed to be a home computer or an office-computer. It could perform both tasks well.
Commodore wanted to sell the Amiga 1000 as a business computer. But there were problems with the distributors and computer shops, because in the years before, Jack Tramiel had treated them very badly, when he had sold the VIC-20 and C64 through ordinary stores.
So Commodore went through two very difficult years and even had to be restructured.
But finally, in 1987, Commodore managed to release two new Amiga models, the business-type desktop-computer "Amiga 2000" and the home computer "Amiga 500". On the inside, both had similar specifications.
It seems, Commodore sold the Amiga 500 like their other home computers before. I bought mine in a department store. The Amiga 500 finally was a huge success and became Commodore's main product for the next years.
The Amiga 500 used a Motorola 68000 as its CPU, running at about 7 MHz (7.09 MHz on PAL, 7.16 MHz on NTSC). It had 512K of RAM. It had a proper keyboard with ten function-keys, separate cursor-keys and a number-block. It had two ports for the conventional digital joysticks. One of these ports was usually used for the mouse, that came with the computer.
The Amiga had a built-in floppy disk drive for 3.5''-disks, that could hold 880K of data.
Back then, it was a huge improvement to be able to store now 880K on the Amiga's floppies, when before you could only store 130K on the Atari 800XL's floppies.
The speed of the Amiga's drive was alright. It took maybe half a minute to load a game of a few hundred KBytes.
The Amiga's floppy drive was actually very similar to a standard PC-floppy drive (it is also possible to modify PC floppy drives to work in an Amiga; there may be even a solution using a special cable).
But the Amiga's floppy controller is very special. Even today's emulators can't make its software counterparts work faster. So they still have to load with the same speed from virtual disk files as the Amiga 500 did from real disks. This isn't a big problem any more though, as recent emulators let you save the whole memory-state of the Amiga. So you can load a game once to its actual start, and save the memory-state there. After that you can totally recall the game's start inside the emulation, whenever you feel like playing it.
The real Amiga can also use standard 3.5'' PC-floppy disks in its Amiga floppy disk drive. In PC-format, these disks can hold up to 1.44 MB. They can be modified in a simple way, so that the Amiga can format them to its 880K-format: Such a disk has two small gaps, one on each side. If you open one, it means the disk is write-protected. If you close the other one, by sticking a piece of tape on it, it means, the disk can only be used for 770K on a PC. In this mode, the Amiga can use it for its 880K.
The Amiga can never handle more than 880K. So there's no way, to get the full 1.44 MB. Nevertheless you can use these kind of disks with the Amiga. Which is good (if you have a real Amiga), because that may be the only kind of floppy disk, you still can purchase somewhere.
The Amiga had various graphics modes. Some of them allowed to use 4096 colours. Screens with high resolutions and many colours tended to flicker though. (Later, there was a third-party device called "flickerfixer" to improve the Amiga's screen-quality in high resolution modes. Often, it also featured a socalled "scandoubler", that allowed to connect an Amiga to a PC-type VGA-monitor.)
The sound-chip of the Amiga was called "Paula". It also controlled disk-I/O. It could produce four different sounds at once. These weren't synthesized sounds, but 8-bit samples.
So there could be lead- or even voice-samples on one channel, sampled synth-string-chords on the second, a bass, sampled from the DX7, on the third and sampled drums or effect-sounds on the fourth. All limited to what was possible with 512K of RAM, that is.
The result was chiptune-music, that sounded quite characteristic. It turned out, that this kind of music was created best with a "piano-roll", that moved vertically in a 4/4-beat, and showed all four sound-channels at once. The first software, that realized that, was the "Soundtracker" ("Ultimate Soundtracker") by Karsten Obarski. Many other "trackers" followed, the "ProTracker" was quite popular. The files created by these trackers included the song-data and the used samples. They had the suffix ".mod" and were called "modules".
On the back side of the Amiga was a large serial port. And a parallel port, which was useful, because a printer could be connected there. I could use a "NEC Pinwriter P20" with the Amiga, although originally that was a PC-printer.
In 1991 I considered buying a "Macintosh Classic", but gave up the idea, because I would have needed to buy a special printer too (like the "Apple StyleWriter II"). So I bought only the Pinwriter instead and used it with my trustworthy Amiga 500 for many years.
When I had bought the Amiga 500 as a disciple in about 1987, I hadn't thought of using it for serious word processing at all. I had wanted to use it for games, sound and programming. Then, when I used it for word processing during my studies in the early 1990s, I just was pleased, that it could do that too. I didn't know, that the fact that home computers could also be used for office purposes had been an achievement of the Motorola 68000-era just starting in the mid-1980s. Before that, it had been difficult.
It would have been far from ideal to do word processing on a C64 with its 8 bit processor running at just 1 MHz. The Sinclair QL (68008 processor) and the Commodore 128 (still 8 bit, but with a special 80 columns mode) may have been the first affordable computers, that were to some extent suitable for office use. Then, the Atari 520ST and the Amiga 500 became available and could do that without problems. But that couldn't be taken for granted.
In the years before, students probably would have used typewriters. I still wrote my first paper for university on an electric typewriter. That could be done, but it was extremely uncomfortable.
There also was a port to connect a second 880K floppy drive externally. To use more than one drive with the Amiga was a good idea, as it was much more convenient than having to switch floppies often.
It was also popular, to expand the Amiga 500's memory to 1 MB. The Amiga 2000 already came with that amount of RAM. But that, the desktop case and the separated keyboard were the only differences to the Amiga 500. I was surprised, that the professional office-model with a four times higher version-number basically didn't offer any more computing power than the modest home computer Amiga 500. But of course that was only the base configuration, and the idea of the Amiga 2000 was to put additional expansion cards into the desktop case lateron.
Like the early home computers, the Amiga 500 had to be connected to a television. To do that, a separate "TV modulator" called "Amiga 520" was sold. Back then I didn't realize, there also was a different cable, that could directly connect the Amiga's video out to a SCART input of a television (without the "Amiga 520"), and even produce a better screen-quality that way.
Commodore also offered monitors like the "1084S". I was quite surprised, that this "monitor" was a kind of television without a tuner. You could connect a video cassette recorder, that typically had a built-in TV-tuner, to the 1084S' SCART-input and watch television that way. I had bought the 1084S to achieve a better screen quality when writing texts. Instead it turned out, I had bought just another television. Well, after all it did alright together with the Amiga 500.
It must have been in the winter of 1986, in the time before christmas, when I saw an Amiga for the first time displayed in the window of a shop in a big city. It wasn't even for sale. Maybe it was an Amiga 1000, I'm not sure. It ran the game "Defender of the Crown". I had never seen such lively graphics. The water around the British Island in the game seemed to shimmer a bit like the real sea. Fire flickered and casted mysterious shadows. Trees and bushes looked realistic, and even the walls of the castles looked like built of real stones. I knew immediately, I would buy that computer, if I could afford it.
Soon after the Amiga 500 was available in a local department store, I bought it for 999,- DM. That didn't even seem expensive, considered that it came with a built-in floppy drive, while the Atari 800XL plus its drive 1050 had cost way more just a few years ago.
I remember, switching on the Amiga 500 for the first time. It showed a white screen with an image of a hand, holding a blue disk, labeled "Amiga Workbench V1.2". At regular intervals, the empty disk drive made a noise, as if it was waiting for something. I didn't know for what, yet. It was a magical moment.
The program, that showed the image of the hand, was stored in ROM and was called "Kickstart". My Amiga 500 had Kickstart 1.2. Later, version 1.3 was popular for use with the 500. The Amiga 1000 could load different Kickstart-versions from disk, as it still was experimental. In the Amiga 500, the Kickstart was burnt firmly into ROM instead.
From the screen with the hand, you could directly insert a disk with a game or an application into the drive. Or you could insert the "Workbench"-disk. The "Workbench" was the operating system of the Amiga including its graphical desktop. "Desktop" and "Workbench" mean pretty much the same, if you think about it. I had Workbench 1.2.
The workbench had a blue background. The icons and letters were white, black and yellow. The mouse cursor had the shape of a red arrow. When you clicked on disk-icons, corresponding windows opened. Directories on the disks were shown as drawers. The icons for files and directories were stored as ".info"-files on the disks. You could manage files using the graphical desktop. The operating system of the Amiga also could do multi-tasking.
It is said, that memory-management wasn't that good in AmigaOS though. It was optimized for speed, not for stability. So it could happen, that the system crashed, especially when you tried to run several programs at once (multi-tasking). Then you suddenly saw a black screen with blinking red writing on it, saying "System Failure" or maybe "Guru Meditation", which was a funny line, implemented in the early days of Amiga Corporation. When you saw that screen, you had no choice, but to reboot the machine.
To do that, you pressed both Amiga-keys together with the Control-key, which could be found on the left, in the middle part of the keyboard. Very early models of the Amiga 500 had a Commodore-key instead of a left Amiga-key, so then you pressed the Commodore-key, the Amiga-key (on the right) and the Control-key to reset the machine. That was basically the Amiga-version of the key-combination "Ctrl+Alt+Del" that was used to reboot MS-DOS PCs.
On the Amiga Workbench-disk, there was also the CLI ("Command Line Interface"), the shell. It provided a command-line, to enter AmigaDOS-commands. When managing files, I hardly ever used the GUI and the icons, I always opened the shell and typed in commands. To switch to the disk in the internal disk drive, you'd write:
cd df0:
To show its content, you'd use "dir". To copy a directory to the external drive, you'd write:
copy my_directory to df1: all quiet
To copy all files in "my_directory" to the external drive, you'd write:
cd my_directory copy #? to df1:
(Note: To get the character "#" in German keyboard-layout in the emulator "fs-uae", go to the settings on the left side of "fs-uae-launcher" and set "Host Taste Backslash" to "Amiga Key 0x2B" there. Thanks, Frode!)
To get information about the drives and its contents, you'd use "info".
The shell of AmigaDOS was not case-sensitive.
But to where could you copy, if you didn't have a second disk drive? Well, there was also a RAM-disk, which was installed by default, when loading the Workbench. It was accessed as "ram:". The problem was, its size depended on the available RAM.
That meant, if you had a default 512K-Amiga, you couldn't copy a disk of 880K to the RAM-disk. With 1 MB of RAM, it probably still wouldn't work, as the AmigaOS also had to be held in RAM. Now you probably see, why a second disk drive made sense with the Amiga.
In directory "c" on the Workbench-disk (accessed with "cd df0:c" or "cd c:"), the available AmigaDOS-commands were stored.
In directory "s" was the file "startup-sequence", that was automatically executed, when the Workbench was loaded.
There also was a "Trashcan", like on the graphical desktops of other operating-systems. I never was fond of these.
The Workbench came with a simple text-editor called "ed". Slightly similar to "vim" on Linux, it had an insert-mode and a command-mode. To save text, you went to command-mode pressing "Escape", then entered "x" to save. To leave without saving, you also went to the command-mode, but entered "q" there. That worked quite well. Sometimes, you had to edit the "startup-sequence". So you did:
cd df0: cd s ed startup-sequence
To leave the shell, you entered:
endcli
In the drawer "Prefs" on the Workbench-disk was an icon "Preferences". You could use that program to edit some interesting system-settings, for example the colors of the GUI. They were stored in the file "df0:devs/system-configuration".
In "df0:devs/keymaps/", the maps for the keyboard-layout was stored. "d" was the file for the German keymap. They were activated with the "setmap"-command, usually found in the "startup-sequence".
Back then, this was all very new and unfamiliar. Actually, I didn't understand, what these things really meant, until I had to use Windows and Linux later. On the Amiga 500, in most cases the Workbench was skipped, and another program-disk was inserted, when Kickstart was loaded at start-up.
The Amiga also came with AmigaBASIC. It could be found somewhere on a diskette called "Extras", that was an accessory to the Workbench. You didn't need to mess with AmigaBASIC, even less than with the Workbench. Unless you really wanted to.
The role of BASIC in the system had quite decreased, since the times when you switched on the ZX Spectrum or the C64 and immediately entered their BASIC-environments.
The developers at Commodore Amiga didn't even write their BASIC-interpreter themselves. They just bought an existing dialect. From a small and at the time not very well-known company called "Microsoft".
Part of the user experience of the Amiga 500 was the behaviour of its internal 3.5'' floppy drive. When you switched on the Amiga, Kickstart was loaded from ROM, and the internal drive started clicking. When you inserted a disk, the drive started running, because something was loaded, and that also made characteristic sounds (which are sometimes today even emulated, because it is part of the experience). When the first parts of the program were loaded, usually something happened on the screen, but the drive kept running, because more data was loaded.
Although technically the drive couldn't be considered "slow", loading could take quite a while. Loading Workbench 1.3 from the standard disk took 1 minute and 10 seconds. And if you had just a standard Amiga 500 (with just a single drive and 512 KB RAM), you couldn't just use the Workbench with other disks. Because, first you had to set up a RAM-disk and copy the commands, you wanted to use. That took even more time, and you had to plan first, what you wanted to do.
I remember creating a custom version of the Workbench disk, which set up a usable RAM-disk automatically and which booted directly into the CLI (= shell).
With just a single drive, you also had to do a lot of disk swapping. It sure would have been useful to have a second, external floppy drive with the Amiga 500 (Commodore model A1010, A1011). Such a device made things much more convenient.
So, although in 1987 the Amiga 500 was probably the most powerful home computer on the market, with its processor running at about 7 MHz, and with the floppy drive being the only storage device, it used to take its time.
You often just had to wait until the Amiga had stopped loading or until it had finished some calculations. And when working with the Workbench and its CLI, you often had to think about what you wanted to do and where you wanted to store which data in advance. These two things combined sometimes lead to something, I would call a "meditative state of mind". For example, back then, people waited patiently for over a minute for the computer to load just a few hundred kilobytes of data. Today, they wouldn't have that kind of patience any more.
On a rainy Sunday afternoon, I would make myself a cup of tea, switch on the Amiga, and play a game of Chess ("Chessmaster 2000") or Golf ("Leader Board"). Or one of the larger games such as "Barbarian (by Psygnosis)", "The Bard's Tale", "J.R.R. Tolkien's The Lord of the Rings, Vol. I" (Interplay) or "Sherlock: The Riddle of the Crown Jewels" (an Infocom Text Adventure). That kind of activity just made me feel good.
I remember, in "The Bard's Tale" the floppy drive often ran, when you entered a building for example. That was, what you were used to. But at only very few occasions, data had to be loaded, because something very special had happened in the game. You noticed that, because the floppy suddenly started running without obvious reason. That was always rather scary - and sure was part of the fun.
The Amiga could use different types of memory:
For example, a stock Amiga 500+ (see below) could maximally be expanded to 2 MB Chip-RAM, 8 MB Fast-RAM and 1.8 MB Slow-RAM, so to 11.8 MB of memory altogether.
Lots of games and applications were written for the Amiga 500. It seemed, people really wanted it to be the next C64. Maybe the Amiga was, in a way, but it still was quite a different system.
Actually,
reminded not so much of the C64, but more of the Atari 8-bit computers, Jay Miner's previous design.
Especially the sample-based sound chip "Paula" of the Amiga had nothing in common with the C64's "SID" chip, which in a way was a small multitimbral three-voice subtractive synthesizer.
After all, when the Amiga-project was started in about 1982, hardly anybody would have believed, the technology would one day go to Commodore.
In 1989 Commodore released an optional 20 MB hard disk called "Amiga 590 Hard Drive Plus". It had to be connected to the expansion slot on the side of the Amiga 500 and also included a memory expansion of up to 2 MB. This device was quite expensive though, and I never saw one in real life.
An emulator of the Amiga-computers including the Amiga 500 is called "fs-uae". On Windows, "WinUAE" may even be more popular still.
If you use a Raspberry Pi3, you may want to take a look at the project "Amibian".
In 1983, sales of video game consoles like the "Atari VCS 2600" broke in.
Atari had lost consumers' trust, because Atari's version of "Pac-Man" and the game "E.T." had been really bad (well, my younger self actually liked that "Pac-Man"). Besides, there were too many other consoles on the market like "Colecovision", "Intellivision", "Fairchild Channel F" and others. There also was strong competition of the home computers like the Commodore 64 or even the VIC-20. Atari also wasn't able to control third party companies producing cartridges for the VCS 2600. Therefore the supply of games increased, so that cartridges in general became rather cheap.
All that lead to what was called the "video game crash of 1983", that especially affected the market in the USA.
The 8-bit computer-range of Atari didn't sell well either. So Atari had huge losses. At the time, Atari was a subsidiary of Warner Communications. Because of the losses, Warner wanted to get rid of Atari.
In January 1984, Jack Tramiel left Commodore, the company he had founded 30 years before. He had had a dispute with Irvin Gould, Commodore's primary shareholder since the 1960s.
Maybe Tramiel realized, that computers like the Apple Macintosh, that was introduced at C.E.S. at the same time, would soon replace the 8-bit machines like the C64, and then Commodore would have the same problems, Atari had then.
He thought, he was better off on his own, without Gould. He wanted a family business, something he'd make successfull and then hand over to his three sons.
He saw the situation at Atari and thought, he had chances to take over that company.
Warner didn't know, what to do with Atari. So Warner sold Atari's consumer section to Tramiel in the summer of 1984.
The deal between Warner and Tramiel was rather complicated, I think, I don't have to describe the details here. In the end, Tramiel took over Atari and payed Warner an appropriate amount, though mostly not in cash, but in promissory notes and shares.
A number of 240 million was mentioned, but I don't think, Tramiel ever paid this amount.
Tramiel fired most leading employees at Atari and reduced its staff to just a fraction. He sold, what he couldn't use, and refounded the company as "Atari Corporation" (he first had founded a new company using his own name, and now renamed that company to "Atari Corporation"). Tramiel also injected 30 million dollars into the new company.
At the same time, also "Amiga Corporation" was close to going broke. Tramiel offered just a small amount of money, but was outbid by Gould's Commodore. Maybe Tramiel had pokered too high (or in this case too "low", that is).
As Tramiel's Atari needed a new computer-model on a Motorola 68000-basis, but didn't get the Amiga technology, Atari had to develop such a machine on its own.
Lead designer of the new "Atari ST" product-line became Shiraz Shivji. He had previously worked for Commodore and had now followed Tramiel to Atari, together with some other executives and engineers. At Commodore, Shivji had also been involved in the development of the C64.
Atari hurried up and could present the "Atari 520ST" already in 1985. As a lot of ideas had been taken from the Apple Macintosh, some people called the new Atari a "Jackintosh".
It used the "TOS" (some called "Tramiel operating system") and a desktop called "GEM"-DOS ("Graphics Environment Manager") by "Digital Research".
There were many adversities, but nevertheless the Atari ST product-line became a success, especially the "Atari 520ST" and the "Atari 1040ST".
The commercial slogan for the Atari ST - compared to the Apple Macintoshs - was:
Power Without the Price
As it had been Tramiel's Commodore 64 against Jay Miner's design of the Atari 800XL before, it now was Tramiel's Atari ST against the Commodore Amiga, again designed by Jay Miner.
Back then, I didn't know any of this background. But I instinctively went for Jay Miner's designs and avoided the products of Jack Tramiel, regardless of the label "Commodore" or "Atari" on them.
Today I think, in his lifetime Jay Miner was probably a nicer person.
Nevertheless, the fact, that Jack Tramiel did things his way, and that he succeeded on his way, has to be acknowledged. I'm not sure, if I would have had the courage and the will at the age of 55 to enter the sinking Atari in the situation of 1984, to create a completely new product-line of computers from scratch and make Atari successful again. That's cool, somehow. He really must have been a tough guy.
As the developers of the Atari ST had been in a hurry, when they had created the machine in 1984/1985, they didn't manage to design a sound chip that was as interesting as the SID chip of the C64 had been. So they just used a conventional Yamaha YM2149F sound chip, which was very similar to the AY-3-8910.
They knew, that that chip didn't sound too good. Therefore they also added an internal Midi-interface ("Musical Intstrument Digital Interface", introduced in 1983) to the Atari ST computers. That way, external sound modules like the Roland MT-32 could have been connected to the ST. The Midi-interface also hardly raised production costs.
Such an interface is usually used to control external electronical instruments, like synthesizers and drum-machines. So the built-in Midi-interface and the surprisingly good timing of the Atari ST computers made them especially popular among musicians and music studios. Professional music software was created for the system, notably the programs "Cubase", "Notator" and "Band in a Box". Recent (and much larger) versions of these programs still exist today ("Notator" developed to "Logic Pro").
It's a bit weird, that Atari STs were much more popular for professional music production with Midi than Amigas, as external Midi-interfaces would also have been easily available for Amigas.
In the late 1980s, Jack Tramiel handed Atari's business to his son.
After Tramiel had left, the game console "Jaguar" was meant to bring Atari back into the console business, after Atari had stopped development during the video game crash of 1983 (the VCS 2600 was produced until 1992). But in the meantime development had moved to Japan, where Nintendo had successfully placed the "NES" ("Nintendo Entertainment System") onto the market. Atari failed to get back into that business.
Also Atari's handheld game console "Lynx" couldn't compete with Nintendo's "Game Boy" range of products.
Atari's new computer "Falcon 030" wasn't successful either. In 1995, Tramiel's son suffered from a heart-attack. Tramiel went in and sold Atari. In 1996, Tramiel retired. He died in 2012, aged 83.
Someone still holds the trade name "Atari", but it's not present on the market with new, exciting products any more.
An emulator for an Atari ST-computer is called "Steem".
In 1981, IBM released the IBM PC, with an Intel 8086 (8088) processor running at 4.77 MHz, which used a version of MS-DOS (IBM PC, model number 5150, 1981).
This system became the standard for office computers. It cost several thousand dollars and had hardly any multi-media capabilities. Back then, it didn't have a hard disk either. These systems were way too expensive for the home user.
Office computers usually displayed more columns of text than home computers. 80 columns was a standard for office computers, while the ZX Spectrum for example displayed 32 columns of text, the C64 40 columns. But the more columns, the smaller each character was displayed on the screen. So on ordinary televisions less columns meant bigger and therefore more readable characters. To display 80 columns sharply, you had to use a special computer monitor. At the time there were only CRT (tube)-monitors, either processing composite- or RGB-signals.
At first, PCs had either a MDA graphics card ("Monochrome Display Adapter") connected to a monochrome monitor, that was built with phosphor somehow (displaying either only black and green, only black and amber or only black and white colour, depending on what phosphor was used). Such monochrome monitors with MDA graphics could produce rather clear and sharp letters, but no graphics.
Or early PCs had a CGA graphics card ("Colour Graphics Adapter") connected to a colour monitor, a setup with usually four colours, but not that clear and sharp letters as displayed by a monochrome monitor.
In 1984, the EGA ("Enhanced Graphics Adapter") was introduced, that could produce 16 colours.
From 1987 on, EGA was replaced by VGA ("Video Graphics Adapter"), which displayed a higher standard resolution of 640x480 pixels. "Super VGA", also introduced in 1987, used a resolution of 800x600 pixels.
In 1984, IBM tried to build a more affordable version of the office PC for the home market, called "IBM PCjr" ("PC Junior" that is, nickname "Peanut").
It had a CGA graphics card, that could display 16 colours and used the sound-chip "Texas Instruments SN76489", also found in the TI99/4A and the BBC Micro, which offered three square wave voices plus noise.
The PCjr still was too expensive though (more than 1,200$ without monitor), and also had several drawbacks: Originally it came with a rubber keyboard and didn't have enough RAM for this kind of system (64K, 128K). Although it ran DOS, it wasn't fully compatible with IBM's PC platform, so special program versions were required. Expansions (like a memory expansion) had to be attached to the side of the computer and had to be powered by an additional external power supply.
So although there were some games for that system, like especially "King's Quest", the PCjr wasn't a commercial success and was discontinued only one year after its release.
Also in 1984, the company Tandy (Radio Shack, from Texas) could build a computer, that realized the idea of a PC for the home user better. It was called the "Tandy 1000". It had the 16 colour graphics and the sound chip, also used in the PCjr. Sound was called the "Tandy sound system". The Tandy 1000 had a real typewriter keyboard (which was a bit small though). It helped to reduce costs, that unlike in IBM's products, the chip set was built onto the motherboard.
The Tandy 1000 also came with a programm called "DeskMate", which was a small graphical user interface with little applications such as wordprocessor, spreadsheet, calendar, drawing- and music program. So DeskMate was a bit like early versions of Microsoft Windows.
At first, the Tandy 1000 was also quite expensive, but became more affordable over time.
Many games were programmed for the special sound- and graphics capabilities of the Tandy 1000.
In addition to the three square wave channels, the sound chip of the Tandy 1000 also had a digital to analog converter, so it could play sound files (similar to .wav-files) to a certain extent. But as an Intel processor was used, that ran at just about 5 MHz, you can imagine, that these sound files had to be of rather poor quality. In most cases it sounded quite awful. For example the games "Gauntlet II" and "Ghostbusters II" made use of that kind of sound. You can search Youtube for these games together with "DOS" to get an impression of how the PC's DAC-chips of the time sounded. Awful. With the help of a modern PC, it may be possible to optimize the quality of single sound files on the Tandy, so that they don't sound too bad then. But then you can also just use the modern PC to play the file.
The Tandy 1000 used an external power supply and didn't have or need a fan. So like other computers of the time (Amiga 500, Atari ST), it operated absolutely silently, if you didn't have a hard disk drive.
In 1986, the Tandy 1000 had a market share of 9.5% of all home computers sold in the US. I haven't ever seen one in Germany though, so it must have been successful especially in America.
In 1987, the "Tandy 1000 EX" was released, which interestingly looked similar to the Amiga 500 or the Atari 520 ST. So the keyboard was integrated into the computer, and there was a (5.25'') disk drive on the side. It also had two integrated joystick ports on the side. It came with 256K RAM and used an Intel 8088 CPU, running at 7.16 MHz. So it really was a MS-DOS based home computer.
In 1988, there was the "Tandy 1000 HX", which looked similar to the EX, but came with a 3.5'' floppy disk drive. It could boot MS-DOS 2.11 from ROM.
Its price without monitor was similar to the Amiga 500.
Nevertheless the 16 colour CGA DOS-style graphics and the three square waves of the Tandy 1000 (which in most cases just sounded like three of those PC-speakers) were still rather inferior to what the Amiga had to offer at the time.
Also Commodore produced office-type PCs (that is clones of the IBM PC), beginning with the PC-10 I in 1984. Other models were called "PC-5", "PC-30" or "Colt". The names weren't in a specific order. Commodore had a license from Intel, to produce the 8088-processor at its chip subsidiary MOS Technology. The PC-1 of 1987 came in quite a small case, which looked a bit like the one of the Amiga 1000. The PC-50 III of 1991 was a 386SX-25 PC.
Also Atari Corporation, managed by Jack Tramiel and his sons, released a PC called "Atari PC1" (1987). There were models from PC1 to PC4. In 1989, Atari released another PC line called "Atari Business Computer" ("ABC"). In 1991, there was an "Atari ABC286". The latest model of that line was the "ABC386".
In the early 1990s, the PC platform with VGA graphics, Soundblaster card and Microsoft Windows 3.1 and later Windows 95 replaced the earlier Tandy 1000 models as well as all other home computers including the Amiga, leaving just Apple with its MacOS as the only competitor.
In 1992, Commodore released the Amiga 1200. It used a true 32 bit Motorola 68EC020-processor (a slightly reduced version of the 68020), running at 14 MHz, which was distinctly faster than the previously used 68000. It came with 2 MB of RAM and cost 899,- DM. Kickstart-versions were 3.0 or 3.1.
For graphics, the Amiga 1200 (also called "A1200") used the socalled "AGA"-chipset ("Advanced Graphics Adapter"), which wasn't compatible with the chipsets of the older Amigas. But when you started the A1200 and pressed both mouse-buttons at once, a plain grey screen with a menu was shown called
Amiga Early Startup Control
where you could select several graphics-chipsets, including the "OCS" ("Original Chip Set") and "ECS" ("Enhanced Chip Set") of the older Amigas, so you could run older Amiga-programs and games this way on the Amiga 1200 still.
The AGA-chipset offered up to 256 colours at once, finally from a palette of full-colour (16.78 million).
The A1200 still was best connected to a television or the monitor 1084S. It could use the same SCART-cable as the Amiga 500.
The mouse and the internal floppy disk drive remained unchanged from the Amiga 500.
The sound-chip Paula, too. But the faster processor and the bigger RAM of the A1200 allowed to make more advanced use of the four sample-channels of the Paula-chip. There was a popular soundtracker-program called "OctaMED", that was even used in several a bit more professional music-productions of the time.
The A1200 came with housing for a 2.5'' IDE hard disk drive, including a connector and controller.
More than a decade later I bought an A1200 second hand on ebay, that had a 40 MB hard disk. That changed the way, the computer was used, quite a lot. But it also made some noise.
The A1200 supports the standard 2.5'' hard disk-format, that was used in PC-notebooks. So you could build in rather large hard disks, of up to 4 GB it is said. (I wouldn't push it to the limit though and go for just about 500 MB maybe). That large amount of disk-space didn't make much sense anyway, as the programs for the Amiga were rather small, so a nice set of programs fit well into 200 or 500 MB.
The A1200 used a kind of slimline case and felt less solid than the Amiga 500. It was significantly smaller too.
The Workbench had developed well though. For the A1200, Workbench 3.0 or better Workbench 3.1 was used. It had become a real operating system by now, that could be installed on the hard disk, if you had one. With the faster processor of the A1200 and a hard drive, the workflow was quite nice. But also became somehow similar to that of a PC.
The A1200 had a PCMCIA-slot (Type II). It could be used to expand the A1200 with Compact Flash cards. This provided capable storage devices.
Other PCMCIA cards provided ethernet to get the A1200 online. There was also software like browsers and ftp-clients.
In 1992, the "Aminet" was established, an online-repository of freely distributable Amiga software, that still exists today.
Although the Amiga 1200 wasn't a bad computer, it wasn't destined for success.
In 1992, "Windows 3.1" was released and started Microsoft's worldwide success. People bought 386 PCs at the time, and Windows seemed the way to go. 1993, a company called "id Software" released the 3D-first-person-shooter "Doom" and changed the gaming industry.
Although there was an attempt with "Alien Breed 3D" by "Team 17" to create such a game on the Amiga 1200 too, it seemed, its architecture wasn't designed for 3D games.
There weren't too many games, that made use of the special features of the A1200 (like the AGA-chipset). Most developers wanted to stay compatible with the Amiga 500.
Commodore made mistakes and finally went bankrupt in 1994.
Two months later, Jay Miner passed away, due to kidney failure at the age of 62 (he had been a dialysis-patient for a long time). It's rather sad, but according to Wikipedia his wife died soon after him the same year too.
In 1995, the company "Escom" (and its subsidiary "Amiga Technologies GmbH") tried to produce and sell the A1200 once again, but failed. In 1996, Escom went bankrupt, and the production of the A1200 was discontinued.
About that time, still the program "WHDLoad" was written. It often made it possible to write the contents of Amiga 500-disks as image-files to the hard disk of an Amiga 1200 and run older games from there.
A1200-programs themselves usually didn't need WHDLoad, as they came with their own installation-routines or could just be copied to the hard disk.
From today's perspective, the Amiga 1200 probably would have deserved more success. It's quite interesting, that you could build in a hard disk of several gigabytes and attach either CF cards or ethernet or even WiFi with PCMCIA to it. The OS boots fast and can handle these devices. The computer runs fast enough to run smaller office applications with word processor (like "Beckertext II") or spreadsheet. Email is also possible. Simple web browsing too. And being an Amiga, it can also run nice 2D-games. So the A1200 already had features, that are appreciated today.
In 1992, maybe people just didn't realize it. I remember someone from Escom presenting the A1200 on television. In its slim and small case, it looked rather fragile compared to the large and solid PC towers and even to my Amiga 500. And what? Connect to a television or to a similar composite monitor, not (without additional hardware) to the standard VGA-monitor of the time? At this point, it seemed rather old-fashioned.
But as mentioned above, the A1200 had quite interesting features.
Nevertheless, I don't suggest, getting an A1200 today (unless you really want to). In the meantime, there have been almost 30 years of development. There's nothing, a modern PC can't do as good or even better (faster) than a 1992's Amiga 1200.
Some say, today, a Raspberry Pi3 running the emulation "Amibian" would be a good replacement, as the emulation is accurate and runs even faster than the original machine. On desktop PCs, "fs-uae" also doesn't leave much to be desired. It's all there.
There had been a few more Amiga models:
The "Amiga 500 Plus", released in 1991, had been an enhanced version of the Amiga 500, featuring the "ECS"-graphics-chipset, recent Kickstart/Workbench (version 2.04) and 1 MB RAM.
It became a problem, that many games for the Amiga 500 with its Kickstart 1.2/1.3 wouldn't run on Kickstart 2.04. Additional hardware to switch between Kickstart versions were sold. Customers weren't happy.
Also in 1991, Commodore released the "CDTV". It was a console in a black Hi-Fi style case, based on an Amiga 500 (Kickstart 1.3, but already "Big Agnus" (ECS) of the A500+) with a single speed CD-ROM drive, that could also play Audio-CDs. The CDTV was originally just meant as a stylish device to show the contents of so-called "Multimedia CD-ROMs" on a television set using a (wireless) remote control.
In 1991, mainly floppy disks had been used that could store 880KB. When suddenly CD-ROMs appeared, that could store 750MB each, there was the idea, that a new type of digital content could be created, a combination of texts, sounds, music and short video-clips. So "Multimedia CD-ROMs" were produced, that often showed educational content or even held a whole encyclopedia.
Later, I had some CD-ROMs of that kind for my "Windows 95" PC ("Funk and Wagnalls Encyclopedia", "TIME Magazine Almanac", "Microsoft Encarta World Atlas", and so on - but these were later, not for the CDTV). And today this type of content is, what internet browsers mostly display (even this website could be seen as content of such type). But in 1991 with an offline 68000 CPU at just about 7 Mhz and a single speed CD-ROM drive - and even with a fan, which the CDTV unfortunately also had - it wasn't so much fun. Which is strange in a way, because from a graphics and sound perspective, many earlier Amiga games with their high speed, continous music and smooth scrolling of the whole screen should have required much more performance from a computer than just displaying some text and playing some sound here and there.
Anyway, the CDTV didn't sell well, and Commodore then also offered an external computer keyboard, a mouse and an external disk drive, to use the CDTV as an Amiga again. But it remained a commercial failure, as people just kept on using their "real" Amigas.
In 1992, the "Amiga 600" was released. It replaced the "Amiga 500 Plus", that had only been sold for about six months.
The Amiga 600 was built into the somehow fragile slimline-case of the A1200. It was even smaller, as the separated number keypad had been dropped.
At first, it was planned to name it just "A300", because although it looked more like the A1200, it was seen as an affordable version of the "Amiga 500 Plus".
Internally, A600 and A500+ basically had the same hardware specifications, regarding graphics chipset (just ECS, no AGA in the A600), memory (still 1 MB RAM) and Kickstart version (2.05 on the A600).
They also both still used the original 68000 processor of the early Amigas, not the faster processor of the A1200.
Still many older games and applications wouldn't run though, as they weren't compatible with the newer Kickstart version.
Unlike the A500+, the A600 had a PCMCIA-slot (Type II) and housing for a 2.5''-hard disk like the A1200.
With the Kickstart version, that was originally shipped with the A600, it wasn't possible to boot from these devices though. The problem was fixed in later versions (Kickstart 3.0, 3.1, on ROM chips).
In its time, the A600 was a commercial failure.
But today, retro enthusiasts who restore and collect old computers appreciate the small form factor of the A600 and find its PCMCIA- and hard disk slots convenient to build in modern storage devices.
In 1993, there also was a gaming console "CD32", based on the A1200. Like the "CDTV" before, it wasn't a great success.
Then there were the large desktop Amigas "Amiga 3000" and "Amiga 4000", using 68030-processors and upwards.
They were highly expandable. Hardware like the expansion bus "Zorro III" or the "Picasso"-graphics cards were produced for them. This was all a bit above the head of the common Amiga 500-user.
Later, AmigaOS 3.9 and 4.1 were created by other companies.
"AROS Research Operating System" is a free operating-system, compatible to Workbench 3.1, running on several systems.
It can also be used in the emulator "fs-uae", which can emulate the Amiga 500 as well as the Amiga 1200, including its hard-disk.
With the last "Amiga 1200" being produced in 1996, the age of home computers had ended.
The home computers described above were the most successful ones, at least here in Germany.
There were many other attempts, that weren't as successful.
I didn't mention for example:
Some others, I didn't 't mention either, but maybe I should have:
Instead there's chapter about the Amiga 1200, although in its time it may have been one of the less successful ones too. But I wanted to show, there had been efforts to take the Amiga-technology beyond the Amiga 500. Besides, the A1200 and its Workbench 3.1 are well emulated today, and the prices of the original hardware have risen quite high in recent years.
Apple, Atari and Commodore were American companies. Acorn, Amstrad and Sinclair were British.
It seems there weren't similar companies from Germany and France, which is a bid sad, I think. After all, the German engineer Konrad Zuse had invented the computer in the first place.
Of course, West-Germany was a big market for home computers, and there were lots of user-communities and "computer clubs". There were also smaller companies, producing peripherals, books or software, like "Data Becker" with their word-processor "Textomat" and their database-program "Datamat" for the C64 and Amiga for example.
But it didn't lead to a large industry like in Britain or in the USA. Maybe the German industry couldn't handle it. Studying computer science had a relation to the old German engineer-tradition. It took rather long and required you to become more or less a mathematician (it still does).
Being entertained by home computers was a lot of fun for the consumers. But for the responsible people in the companies it must have been like a ride on a tiger. After the C64 was introduced in January 1982, Jack Tramiel as the president of Commodore must have been the man of the moment. But just two years later he had to leave the company, that he himself had founded.
Sinclair released the ZX Spectrum in April 1982. In 1984, Sinclair failed with the QL. In 1986, everything had to be sold to Amstrad. That was it.
Releasing a home computer successfully must have been like having a number-one hit in the music-charts, even better. Later, Bill Gates and Steve Jobs became something like pop-stars.
But such a smash-hit could keep up a company for just two years. Or maybe a few years more, if the technical progress was a bit slower. Then the next top-hit had to be found. Or the company went broke and the jobs were gone. The pressure must have been enormous.
In 1984, Commodore found a next top hit with the Amiga (but couldn't release the Amiga 500 until 1987). That kept the company up until about 1992. The next hit didn't happen. In 1994, Commodore had to file for bankruptcy.
Atari almost went broke in 1983/1984. Tramiel totally restructured the company. Not much more than the name remained. From there, Tramiel managed to build the Atari ST from scratch. Then he retired, and another top hit didn't happen. So Atari was sold and was withdrawn from the market.
Apple also had severe problems in the 1990s (and was only saved from bankruptcy by Microsoft's investment in 1997). It had a come-back with the iMac in 1998. Once more, they had a hit.
In the early 1990s consumers withdrew nearly all of their spending capacity from home computers and instead put it into the PC-platform using Microsoft's Windows operating system ("Windows 3.1", "Windows 3.11", "Windows 95"). Making Bill Gates the richest man in the world. He had the biggest hit of them all.
As a teenager in the 1980s, I wasn't much of a programmer. I liked the "Sinclair BASIC" of the ZX Spectrum, but didn't manage to write much more than just small programs. I may have typed in a few programs from computer magazines. Also I didn't know anyone who was able to take the step from BASIC to assembly language, that was common for advanced programmers back then. I had a book with a listing of the Spectrum's ROM, but wasn't able to learn assembly language from it either. So I in a way I admired people that could write larger BASIC programs or even write in assembly language.
The BASIC on my next computer, the Atari 800XL, was rather slow and uninviting. You better held down the "Option" key and got rid of it. The situation was similar on the Amiga 500: Get rid of BASIC, by not inserting its dedicated disk. "Microsoft QBasic" on DOS-PCs of the 1990s was a bit more interesting.
In the 2000s I learnt how to use Linux and finally got into writing code, mainly in Perl and Python.
Then, to me the Linux world seemed something completely different than the ZX Spectrum's BASIC.
But recently, I rewrote one of my Python scripts in it. The script just did a few calculations, manipulated an array of strings and printed the results. And it wasn't too big (but not too small either). So I thought, well, let's see, if the ZX Spectrum (48K) would have been able to do that.
During developing (using the tool "bas2tap" as mentioned above), I got further into Sinclair BASIC, than I ever had in the 80s. And it was a nice experience to remember the ways, you had to write it.
So how does it compare to Python?
Although many constructions seem to be similar, several things are missing. There aren't classes, objects or even proper functions with local variables and return values. There aren't hashes (dictionaries), arrays of arrays (AoA) or while-loops.
So in a way Sinclair BASIC has a lot less language elements than Python. Which isn't surprising, as the whole interpreter is stored in just 16K of ROM. While the Python directory on my hard disk has a size of about 200 MB. Which is more than 12.000 times as much code.
I notice, that it really was a marvellous achievement, that the developers of the Spectrum ROM code could fit the BASIC interpreter (including even the Spectrum's character set) into just 16K of memory. They must have been really excellent programmers.
100 REM while loop 1 110 ... 120 IF ... THEN ...: GO TO 140 130 GO TO 100 140 REM after while loop 1
100 FOR i = 1 TO 1000 110 ... 120 IF ... THEN GO TO 150 130 ... 140 NEXT i 150 ...
100 FOR i = 1 TO 1000 110 ... 120 IF ... THEN NEXT i 130 ... 140 NEXT i
PRINT "First one: ";a$;" Second one: ";b$
So in the end, you can write code, that is in a way rather similar to the one in Python or C.
That means, basically it's the same thing. Sinclair BASIC on a Spectrum isn't comfortable to write, with required line numbers and rubber keys. I'm getting the same results much faster in Python today, so now I can be much more productive.
Back then, it also may have been a good idea, to write everything down on paper first, and only type the already completed program into the computer.
But my impression is, with what I've learnt in my later years about Perl, Python and C, I think, I would have mastered the Spectrum's BASIC back then too. So although Linux and these languages were very unfamiliar to me at first, it wasn't a mistake to learn them, and Sinclair BASIC (or the BASIC of other home computers like the C64 or the Atari 800XL) wasn't something completely different.
After all, when the ZX Spectrum home computer was released in 1982, computer scientists had already been writing C code on Unix machines for several years.
BASIC was just more basic. And (on an 8-bit processor running at just 3.5 MHz) it was more slow.
| No sound at all |
|
Silence. |
| Sending a square wave signal |
|
The processor can just send a primitive square wave signal, either from a cassette output (TRS-80), or from the expansion port ("user port", PET). |
| Loudspeaker |
|
A plain loudspeaker built into the computer. As the computer's processor can send "power" and "no power" to the speaker, a single square wave voice can be produced. Usually, the volume of the sound can't be controlled. |
| TIA (1977) | Atari 2600 ("Television Interface Adaptor", also creates the graphics) | 2 voices; pulse waves or white noise. |
| POKEY (1979) |
|
4 square wave voices; distortion; high-pass filter. |
| VIC (MOS Technology; 1980) |
|
3 square wave voices plus noise. |
| SID (MOS Technology 6581, 6582; 1982) |
|
3 voices with a selection of 4 different waveforms per oscillator (triangle, sawtooth, pulse wave, white noise); low-pass, high-pass and band-pass filter; amplitude envelopes; oscillator sync. |
| AY-3-8910 (General Instrument) |
|
3 square wave voices, 1 noise; can be forced to primitvely handle PCM data (digital audio). |
| SN76489 (Texas Instruments) |
|
3 square wave voices, 1 noise. |
| TED (MOS Technology; 1983) |
|
2 square wave voices (can also produce noise). |
| YM2151/YM2164 (Yamaha) |
|
8 voices, 4 operators FM sound chip. |
| OPL2 (Yamaha YM3812) |
|
9 voices, 2 operators FM sound chip, pseudo sawtooth waves. |
| Paula (MOS Technology 8364; 1984) |
|
4 PCM sample sound channels (8 bit); 2 channels mixed to the left, the other 2 to the right; independent control of frequency and volume for each channel. |
| MT-32 (Roland; 1987) |
|
Semi-professional sound-module using sample-based synthesis (like in the professional "Roland D-50" synthesizer). |
| OPL3 (Yamaha YMF262) |
|
18 voices 2 operators FM sound chip (also with 4 operator mode), additional waveforms to the OPL2; stereo. |
It seems, the early Apple Macintoshs used their processor in combination with the general-purpose I/O-chip "MOS Technology 6522" to produce a single voice of PCM sound (8bit, 22050 kHz).
| 1954 | Jack Tramiel founds "Commodore Business Machines International" (CBM). The company produces typewriters. |
| 1958 | Clive Sinclair (18) publishes several books about electronics. |
| 1961 | Sinclair Radionics Ltd. founded. |
| 1966 | Irving Gould becomes main-shareholder of Commodore. Cooperation with Jack Tramiel. |
| 1969 | Operating System "Unix" developed for large computers. |
| 1970 | "The Beatles" break up. |
| Moog "Minimoog" monophonic analogue synthesizer released. | |
| 1971 | Atari Inc. founded. |
| Commodore produces pocket calculators. | |
| 1972 | Atari releases arcade video game "Pong". |
| Sinclair produces pocket calculators. | |
| Programming language "C" developed by Dennis Ritchie. | |
| 1974 | Operating System "CP/M" developed by "Digital Research Inc." (Gary Kildall). |
| "ABBA" wins the Eurovision Song Contest. | |
| Helmut Schmidt German chancellor. | |
| 1975 | 6502-processor released by MOS Technology (Commodore). |
| 1976 | Zilog Z80-processor released. |
| "Jean-Michel Jarre - Oxygène" released. | |
| 1977 | Apple II (first version) released. |
| Atari VCS 2600 released (circuit-design by Jay Miner). | |
| Commodore PET-2001 personal computer released. | |
| "Star Wars" (I) in cinemas. | |
| Jimmy Carter US-president. | |
| 1978 | "Superman" (I) in cinemas. |
| "Jean-Michel Jarre - Equinoxe" released. | |
| "Dallas" TV series first season. | |
| 1979 | Atari 800 released (circuit-design by Jay Miner). |
| Motorola 68000-processor released. | |
| "Fairlight CMI" music computer (sampler) released. | |
| "Alien" (I) in cinemas. | |
| Margaret Thatcher prime minister of the United Kingdom. | |
| 1980 | Sinclair ZX80 released. |
| Commodore VIC-20 released. | |
| "Star Wars II: The Empire Strikes Back" in cinemas. | |
| "Magnum P.I." first season. | |
| John Lennon (40) shot in New York City. | |
| 1981 | Sinclair Research Ltd releases the ZX81. |
| Cooperation between Sinclair (UK) and Timex (USA). | |
| Texas Instruments TI-99/4A released. | |
| Acor wins BBC-competition. BBC Micro released. | |
| IBM PC using PC DOS 1.0 (= MS-DOS 1.14) released. | |
| Roland "Jupiter 8" polyphonic analogue synthesizer released. | |
| "Depeche Mode - Speak & Spell" released. | |
| "Jean-Michel Jarre - Magnetic Fields" released. | |
| Ronald Reagan US-president. | |
| 1982 | Sinclair ZX Spectrum released. |
| Commodore C64 released. | |
| Main-developers of the C64 leave MOS Technology (Commodore). Some go to Ensoniq. | |
| Jay Miner leaves Atari and founds "Hi-Toro" (Amiga Corporation, together with Dave Morse). | |
| Roland "Juno-60" polyphonic analogue synthesizer released. | |
| "Tron" in cinemas. | |
| "E.T. the Extra-Terrestrial" in cinemas. | |
| "Depeche Mode - A Broken Frame" released. | |
| "ABBA" break up. | |
| "Neue Deutsche Welle" music in Germany. | |
| Helmut Kohl German chancellor. | |
| 1983 | Video Game Crash: Atari loses a fortune. |
| Sinclair "ZX Interface 1" and "ZX Microdrive" released. | |
| Clive Sinclair is knighted and receives an honorary doctorate. | |
| Price war between C64 and TI-99/4A. | |
| Atari 800XL released. | |
| Apple IIe released. | |
| Yamaha "DX7" FM Synthesizer released. | |
| "Star Wars III: Return of the Jedi" in cinemas. | |
| "WarGames" in cinemas. | |
| "Depeche Mode - Construction Time Again" released. | |
| Soviet Lt.-Col. Stanislaw Petrow decides against starting World War III, when the nuclear defense system shows launching of American missiles. | |
| 1984 | Apple Macintosh introduced at C.E.S.. |
| Amiga Corporation shows the "Boing Ball"-demo at C.E.S.. | |
| Sinclair QL released. | |
| January: Jack Tramiel leaves Commodore after a dispute with main-shareholder Irving Gould. | |
| March: TI-99/4A discontinued. | |
| Summer: Jack Tramiel buys Atari's consumer product range from Warner and restructures it as "Atari Corporation". | |
| Amiga Corporation out of money. Atari Corporation (Jack Tramiel) is outbid by Commodore. Commodore buys Amiga. | |
| Commodore SX-64 (a portable C64) released. | |
| Amstrad CPC 464 released. | |
| IBM PCjr and Tandy 1000 released. | |
| "GNU Project" started. | |
| Emu-Systems "Emulator II" sampler released. | |
| "Terminator" (I) in cinemas. | |
| "Depeche Mode - Some Great Reward" released. | |
| "Miami Vice" first season. | |
| 1985 | Commodore Amiga releases Amiga 1000. |
| Atari 520ST released. | |
| Commodore 128 released. | |
| Atari 130XE released. | |
| IBM PCjr dicontinued. | |
| Programming language "C++" released by Bjarne Stroustrup. | |
| "Free Software Foundation" founded by Richard Stallman. | |
| "Back to the Future" in cinemas. | |
| 1986 | Atari 1040ST released. |
| Apple IIGS released. | |
| C64-II (C64C) released. | |
| Commodore struggles on the market. | |
| Amstrad buys Sinclair's computer product range. Sinclair QL discontinued. | |
| "Aliens" (II) in cinemas. | |
| "Depeche Mode - Black Celebration" released. | |
| Nuclear catastrophe in Tchernobyl (Ukraine, USSR). | |
| 1987 | Amiga 500 and Amiga 2000 released. |
| Atari XEGS 8 bit gaming console released. | |
| Tandy 1000 EX (DOS based home computer) released. | |
| Roland "MT-32" midi controlled sound module for computers released. | |
| Programming language "Perl" published by Larry Wall. | |
| Roland "D-50" digital synthesizer released. | |
| "Depeche Mode - Music for the Masses" released. | |
| 1988 | Cambridge Z88 released. |
| Tandy 1000 HX released. | |
| "Magnum P.I." ends (8 seasons, 162 episodes altogether). | |
| 1989 | Jack Tramiel hands business of Atari Corporation over to his son. |
| "Depeche Mode - Violator" released. | |
| Commodore 128 discontinued. | |
| "Miami Vice" ends (5 seasons, 112 episodes altogether). | |
| George Bush sen. US-president. | |
| The Berlin Wall falls. | |
| 1990 | John Major prime minister of the United Kingdom. |
| The "German Democratic Republic" ("GDR", East Germany) dissolves. | |
| 1991 | Amiga 500 Plus released. |
| Programming language "Python" published by Guido van Rossum. | |
| Operating system kernel "Linux" developed by Linus Torvalds. | |
| "Terminator 2: Judgment Day" in cinemas. | |
| "Dallas" ends (14 seasons, 357 episodes altogether). | |
| The Soviet Union dissolves. | |
| 1992 | Amiga 1200 and Amiga 600 released. |
| Microsoft Windows 3.1 released. | |
| 1993 | Microsoft Windows for Workgroups 3.11 released. |
| Apple IIe discontinued. | |
| 1994 | Commodore files for bankruptcy. |
| Jay Miner dies (at the age of 62). | |
| 1995 | Microsoft Windows 95 released. |
| Escom tries to sell Amiga 1200. | |
| 1996 | Amiga 1200 discontinued. |
| Jack Tramiel sells Atari Corporation and retires. | |
| 1997 | Microsoft saves Apple from bankruptcy. |
| 1998 | Apple iMac released. |
