The CoCo

There's plenty of other Web sites offering a comprehensive history of the Color Computer (The CoCo Chronicles has an extensive narrative and links to a bunch of good sites), so I won't replicate that here. Instead, this is a opinionated overview of the CoCo's suitability as a gaming machine (less than perfect is the realistic overall verdict), how some great programmers overcame some of those limits, some terminology for people new to the machine, and a comparison of some competing platforms and titles sold at the time.

The evolution of CoCo gaming and the competition (as seen through the eyes of Donkey Kong) The arcade original: Addictive four-screen game perhaps best-known for introducing Mario (then called Jumpman) to the gaming world. Notice the screen is vertically oriented, compared to the horizontal screens of home versions, a common problem programmers dealt with in various ways. First generation: The early two-screen CoCo game Monkey Kong (above) was better than the horrific Atari 2600 version of Donkey Kong (below), but not much. The red, blue, yellow and green graphics are one of the CoCo's "official" four-color 128x192 medium-resolution modes (PMODE3). Second generation: The Tom Mix game Donkey King (above) set the standard for all subsequent CoCo games. It had all four screens, unlike most home conversions, and was considered superior in nearly all aspects to the official Donkey Kong for the Colecovision (below), a machine far more advanced than the Atari 2600 that largely owned much of its success to the barrel-tossing monkey. The red and blue colors on the CoCo version were the"unofficial" hues programmers managed to coax out of the machine's 256x192 black and white hi-res mode (PMODE4). The competition: Good as the CoCo version was, it still couldn't match the best of the home versions, including the Commodore 64's (above). But it still equaled or bettered what was found on other machines (some of which were much more expensive): Sinclair ZX Spectrum Apple II Atari 800 Vic-20 (eeech!)

The CoCo was Tandy's entry into the low-end home computer wars of the early 1980s, doing battle with the likes of the Commodore 64, Vic-20, TI-994a, Atari 400 and 800, and a whole bunch of lesser-known machines whose market lifespan lasted anywhere from a few years to a few weeks. The CoCo - as it was affectionately known - was a bit high-priced and short of power compared with the top of the class. It enjoyed a pretty long run nonetheless, thanks to Radio Shack's stubbornly persistent marketing of it and an overly loyal fan base wanting to convince the world their machine was just as good as anything else (sort of what Mac users go through these days).

Don't get me wrong: I used and loved CoCos from junior high school through my college days (and I'm now a hopeless Mac devotee). I'm simply objective enough to realize the best of the competition was better. The Ataris and Commodore were faster, cheaper, had better text and graphics capabilities, better sound and were much more widely supported by major software houses. The CoCo's advantage? You could buy stuff and get it serviced at your local Radio Shack at a time when service and knowledgeable tech people were often in short supply at Sears and other mass merchandise stores selling home computers. This was probably the machine's strongest selling point. It had a pretty nice version of Microsoft BASIC for users into that sort of thing. I'm not really sure what else. Some hardcore fans said the competition had too many quirks, but the CoCo had plenty of its own.

It started life as a $399 4K machine in 1980, went through its heyday as a 16-64K machine in the $200-$300 range (and didn't my foster parents have a fit watching the price of the machine drop as its performance increased), and ended its life as a third-generation machine with up to 512K selling for less than $200. I should note this entire page deals only with the Color Computer 1 and 2; the CoCo 3 was a vastly improved machine with all kinds of new features including up to 512K of memory and much better graphics, but 1) I never owned one and therefore know little about it and 2) by then it was too little, too late - an entirely new generation of machines including PCs and Macs had arrived. It remains the machine of choice among CoCo hobbyists, but never made the impact of the original machines.

When it came to gaming all of these computers were also competing against the video game consoles available at the time, starting with the Atari 2600 in the late 70s and ending with the Nintendo in the mid 1980s. The Atari, of course, is the darling of any true retro gamer since it really got the market rolling and saw some of the most amazing programming achievements ever for an extremely limited bit of hardware. It was still selling more than a decade later, something nothing except Nintendo's GameBoy can claim as far as I know. And Nintendo's NES, of course, pretty much kicked in the era of modern gaming after the infamous industry crash of the early 1980s.
The CoCo is a real enigma when trying to figure out how it rates against these machines. Even the Atari managed to top it with a number of its arcade titles - its Ms. Pac-Man and Jr. Pac-Man titles play better than any CoCo port, even if they don't look as good - and the sheer number of games would leave me in a real dilemma if I could only bring one to a desert island. On the other hand, there were plenty of games I would much rather play on my beloved CoCo than on the later "advanced" Colecovision and NES machines (Donkey Kong comes immediately to mind). And none of the game consoles could match the computers, of course, for in-depth adventure games, simulations and the like. Not to mention term papers, letters home asking the parents for more money and programming your own stuff.

One other quick hardware note: When it came to joysticks, the CoCo's were definitely near the bottom of the heap, although certainly not the worst ever (that dishonor probably belongs to the Atari 5200 for all time). Most systems used arcade-like self-centering joysticks that recognized movement in eight directions; the CoCo's worked sort of like today's mice and trackballs (not necessarily a bad thing) in returning a set of coordinates - great for trackball and paddle-type movement), but they didn't center and were horribly fragile. Bottom line was they were terrible for most arcade games, where simple, quick and intense moves were needed.

The games available for the CoCo generally fall a notch below the competition. But at least they showed a marked improvement during its lifespan. Even so, the CoCo had serious limitations that no programmer could overcome regardless of talent. And it started with the first things any gamer would notice when comparing titles and platforms - graphics and sound:

• An Atari computer could display 256 high-resolution colors; the CoCo could display four - and that was only through a clever programming "cheat" since the official limit was two. The Atari and Commodore both had screen resolutions of more than 300 horizontal pixels; the CoCo's maximum resolution was 256 wide by 192 high. So everything looked less sharp and colorful before a gamer ever got past the magazine ads or store demonstrations.

• Other computers and video games could play music in the background and play multiple sounds at once; the Color Computer could do neither. Its sound capabilities were limited to a single sound at once that brought everything else to a halt - although again some programmers could sort of overcome these limits with enough effort. But the results were rarely impressive. Again, a gamer used to background music in Donkey Kong, Frogger or whatever, was quickly going to notice at the store if the competition had it and the CoCo was running largely silent.

Other problems were plenty troublesome. It was slow compared to a lot of the competition. I would guess it ran at one-third to one-half the speed of the class leaders. Software was limited: The corporate gurus at Tandy apparently figured they should supply our games and reap the profits, so they didn't exactly make it easy for third parties to license and release popular software titles (other companies, notably Texas Instruments then and Apple in past years, had similar attitudes and suffered the same disastrous results as buyers simply bought machines that would play the titles they wanted). And those analogue joysticks - yikes!

OK, I can hear the flame-mailers now: "Geeze, if this machine sucks so bad why did you have one all those years?"

The above is a listing of the relative speeds of some computers from the 1970s and 80s. Note that any modern machine would likely complete the benchmark in mere hundredths or thousandths of a second.

Calm down - the news wasn't all bad: Just trying to make sure people know the CoCo's warts before we delve into its virtues and how the thing actually works.

First off, all this is comparing the CoCo to the best two or three inexpensive computers at the time. Keep in mind this was a period when there were literally dozens of machines out there, all with competing operating systems and, in general, none of them would run stuff by the other guys. On that list the CoCo ranks high: It compares favorably with Sinclair's ZX Spectrum, a wildly popular computer in the U.K. that still has one of the most active user groups of any machine, and Texas Instruments' TI-994a, a very popular U.S. machine that was a lot like the CoCo in that it didn't have the muscle to be at the head of the pack (and it had a much shorter lifespan, thanks to some of the worst marketing of any company). And anyone who bought one of the dozens of obscure "here today, gone tomorrow" machines like the Mattel Aquarius, Coleco Adam, Spectravision, Memorex MTX and so one would probably make the switch in a second, even if some of them had better raw hardware.

Next, the CoCo and its accessories were reliable and user friendly for the era. That counts for a lot when machines like Coleco's heavily-promoted Adam experienced something like a 30 to 40 percent rate of failure. Commodore - which generally put cost-cutting above all other considerations - sold users disc drives that were slow, arrived well after promised shipment dates and frequently malfunctioned.

Finally, once you're committed to a machine (especially with limited funds), there's almost always going to be plenty of redeeming software for it if it has any shelf life at all. One of my greatest thrills was seeing what people can do with limited hardware and software capabilities, even today. Even though I'm learning more advanced online programming languages, for example, I'm still trying to figure out how to write an honest-to-goodness Web game using only HTML for no other reason than to exercise the right-brain muscles (or are those the left ones - I never can keep that straight).

I also owned the humblest of humble machines as a teen, Sinclair's ZX-81, and I loved it even though it ran at maybe a tenth of the CoCo's speed, was black and white only, had no sound, a miniscule flat plastic keyboard and a memory system that crashed the computer if you bumped it even gently. Why did I love it? Among the many reasons: One of the programs written for it was a computer chess game that used 1 kilobyte of memory, which is probably about the amount that this paragraph requires. Picture it - take the letters in this graph and, assuming two to five of them are needed for each command, imagine writing a something where the computer actually understands the rules of chess. Oh yeah - I forgot; displaying the chess board consumes maybe one-fourth of the memory, so subtract that before you get started. I consider it the greatest accomplishment in the history of computer programming and it was impossible not to appreciate it every time I played it, even if it didn't look fancy or play terribly skilled game (it actually matched my limited skill level very nicely).

Perhaps more understandable to a wider range of gamers, take the Atari 2600. Did anyone who played primitive clunkers on it like Combat and Air-Sea Battle honestly believe they'd be able to play titles like Pitfall II, Stargate and Jr. Pac-Man on it more than a decade later? They were astonishing accomplishments and even if they couldn't match Super Mario Bros and the rest of the stuff starting to show up on Nintendo, it was still a whole lot of fun.

With the Color Computer the discoveries began almost immediately and some of the tricks that became industry standards are truly impressive.

First off was the four-color high-resolution mode. The CoCo officially had several graphics modes, with resolutions ranging from an eight-color 64x32 display to 256x192 in two colors (the higher resolutions used a lot more memory, so earlier games in particular used the chunkier lower-resolution modes). And the hi-res black-and-white screen didn't look all that sharp - bits of color would creep up everywhere (Radio Shack officially called it a "metallic effect"). Actually, it was a pretty simple thing - by turning on every other pixel you could make the dots appear as either red or blue; thus the CoCo ended up essentially with four colors at the highest display level and this became the standard for most latter-day games.

Some programmers figured out how to simulate four sounds at once instead of the standard one, although background sound or music was largely beyond the abilities of most. And for those writing games in BASIC (not many at the commercial level, but there were a few) there was a truly interesting discovery: By typing in a single undocumented command (POKE 65495,0) you could essentially double the speed the machine ran at. The only trade-off was you couldn't do things like save to tape or disk while it was in effect.

There's more, but this should give you an idea of why programmers persisted in trying to find ways to stretch their machines as much as possible. I'm actually much more appreciative of their bedroom and garage office efforts than what I see on today's machines that offer virtually unlimited graphics, sound and memory, not to mention full-scale industrial design teams who put titles together like they're movies (understandable, since they now generate that kind of money). That, of course, is the essence of the retro gaming fan and the reason I suspect those of you reading this rant ever bothered to make it all the way to its conclusion :)