A bit disappointing that the article doesn't go more into the technical details of the hardware and how the two designs differed, and how the Atari design influenced the 16-bit Commodore machines (the Amiga).
IMHO the 8-bit Ataris were a lot more innovative than the Commodore 8-bitters (including the C64), despite being a couple years earlier to market.
For instance the Atari display chip ANTIC was an actual co-processor, running a sequence of simple instruction via DMA to produce the display (very similar to the Amiga's copper).
The C64 (and other 8-bit machines) could only reprogram the display chip mid-frame by directly writing hardware registers with the CPU at the right time (often controlled by a raster interrupt).
The 8-bit Ataris were the actual predecessor to the Commodore Amiga (not surprising, because the same people who designed the Atari later designed the Amiga).
Indeed. ANTIC did beam racing based on display lists and could switch modes and palettes mid-screen (tricks the Amigas inherited). The sound was modest, but nothing in the 8-bit space (and on the PC clone side until EGA) comes close in the visual side. And the SIO could daisy chain intelligent devices running their own tiny little OSs just like we do with USB-C today, which is really cool for 1979.
The 64 had the just-right choices for its 16 colors, which provided some amazing visuals when combined, and had the most beloved sound chip of its time.
From retrogeek to retrogeek, if you are in doubt, get both. And an Apple //e to finish out the 6502 family. All three are extremely interesting on their own.
> IMHO the 8-bit Ataris were a lot more innovative than the Commodore 8-bitters (including the C64), despite being a couple years earlier to market. ... The 8-bit Ataris were the actual predecessor to the Commodore Amiga (not surprising, because the same people who designed the Atari later designed the Amiga).
Indeed. While the closest thing to a true successor to the Commodore 64 and 128 (ignoring the unreleased "Commodore 65") in the "16-bit" era was the Atari ST, which like the C64 was optimized for low cost.
I had a Commodore 64 because the Atari 2600 promised a keyboard option to turn it into a computer. It never came to market. So I switched to Commodore who had a video game console with a keyboard and BASIC.
The Atari 800 was nice I heard, but there was a war between Commodore and Atari users on the BBSES.
The Atari 800 had more colors than the Commodore 64, but the SID chip on the Commodore 64 had better sounds, which attracted musicians to it.
Oddly enough management of Atari and Commodore switched and the Atari ST was the successor of the Commodore 64, but I had the Commodore Amiga 1000 that was designed by Jay Miner and other Atari engineers.
The story of how Jay Miner and the rest of very talented engineer crew actually ended up under Commodore is a whole essay and a half - needless to say, that wasn't what they had in mind when they started out at all! There's a fantastic series of articles on the subject:
https://arstechnica.com/series/history-of-the-amiga/
(skipping past the first 140 seconds, because I think the emulator is not doing it justice, for either audio or video, until about 150 seconds in -- the game needed a real NTSC TV)
Here in Germany I knew basically no one who had an 8-bit Atari, it only became popular enough with the ST. And even then it led a bit of a pariah existence behind the Amiga, just like the Midgard RPG was behind the Dark Eye RPG (to keep with the 80s/90s nostalgia mood).
In Norway the Amiga vs ST wars felt quite real. :) The Former was by far the best, except for Midi of course - and as such it raged on. (I had the Vic20, C64, then Amigas).
For me it being in my mid 40s now it was the original religious tech war. Now we have Android Vs iPhone etc.
It was in the UK too - the worst part of which was the games that developed for ST, as the most limited platform, then ported to Amiga. So many Amiga versions were no better than the ST because of that, as it was obviously cheaper than supporting both platforms properly.
The two most impressive things about the ST were the late addition of midi, and that it was designed and launched in some absurdly short time - six months or less?? - which explains many of the compromises, and it getting Gem.
Having had Amiga snatched from under him, Jack Tramiel wanted a fast response to spoil their party. Tramiel who'd only left Commodore the previous year.
There's so many what-ifs to play with this timeline.
Well Atari and Commodore enthusiasts were already used to crippled ports through their experiences with Apple II ports with that awful aliased color palette. :)
To show how real it was the Populous the Promised Lands expansion had a skin which was Atari vs Commodore, on a computer paper landscape with digits for the sea as I recall.
Not just regional, but also tribal. I was firmly in C64 camp and we always make fun of ZX owners or Sharpists and vice versa. It is silly looking back two dacedas, but hey we were all “kids” back then.
It is interesting how biggest flamewars are always between groups of people that are most in common and exclude rest of general public.
I spent a week at a computer camp when I was 10 or so, and was exposed to the Atari 800 (and 400). To 10-year old me, they seemed far superior to my neighbors' C64 machines. My only clear memory is being blown away playing some kind of Star Wars based first-person game. For the day, the graphics seemed incredible.
The Atari is referred to as "being built like a tank" due to the shielding, but when did this stop being an issue? Did the FCC relax its requirements, or did RF interference simply become less of a problem now that people aren't as reliant on antennas anymore?
It never stopped to be an issue, only today there are better ways to solve it, mainly multilayer boards with internal ground and power planes and generally better understanding of EMC.
Apple basically found a way to game the system and ended up with a design that had fewer compromises and lower cost (on this front). I've heard several interviews on ANTIC The Atari 8-bit Podcast with various designers of the Atari computers who were very frustrated at being compelled to design their computers like a tank while Apple largely, in their view, just ignored the FCC and got away with it. Wikipedia has a more generous take on the situation, but even Wikipedia concedes that Apple was exploiting a loophole in the regulations.
Later, the FCC did relax its requirements, but by that point, Apple had established itself as the dominant player.
Anecdotally in Canada the hackers who actually discovered how to code for the machines tended to be on Atari and C64 owners tended to be more consumer oriented. I started out thinking I'd get the ZX81 kit with my own money which got vetoed by my parents saying that was a toy and pitched in for an Atari 400 (with no storage). I'd alternate between writing a game playing it for a day or two, reset to the Asteroids cartridge, repeat.
...but during the same period Radio Shack attempted to compete with the C64 and Atari 8-bit machines with their TRS-80 Color Computer line; specifically the Color Computer 2 (the Color Computer 3 was a more worthy competitor, but since it had 128K and came into being at the tail end of the 1980s when things were transitioning to the PC and the 8-bit machines were "dying out" - it doesn't really figure here).
The thing about the Color Computer 2 (this applied as well to a great extent with the 3), is that it was mostly a reference design by Motorola, and subsequently virtually all of the main component ICs were Motorola parts.
In this case, the CPU was the 8-bit 6809 (clocked at roughly 1 MHz - though there were ways to double this speed), and the "display chip" was the MC6847:
As you can tell from the wikipedia article, it's capabilities were nowhere near that of the machines it was competing with. There was a way to get some weird modes called "semigraphics" via messing with another chip in the machine called the "SAM" chip, but they were rarely used.
In short - for everything the Color Computer 2 did, it relied almost entirely upon the speed and capabilities of the CPU itself! Graphics blitting, sound generation (there was no "sound chip" - merely a 6-bit DAC - which was shared with other parts of the system as well), keyboard/joystick input - all went through the CPU.
It has been argued that the Motorola 6809 is a far superior CPU (in design and capabilities) than the CPUs used in the Atari or Commodore 8-bit machines (or the Apple IIe); in fact, many arcade and gambling machines of the era also used the 6809 (generally in multiples) over other offerings.
Where it fell down, though, was it's price. It was easily several times more expensive than other offerings from other manufacturers, which the other machines used (mainly variants of the MOS 6502).
I've often wondered what/why Tandy/Radio Shack took this route? Now originally, the Color Computer (well, the 6809 and 6847 combo) grew out of a "videotext" style terminal that was originally marketed toward farmers, with the idea of receiving up-to-date "almanac" and planning information via phone lines and the terminal. This was during the late 1970s/early 1980s. I can't recall what this device was called, but I recall it had a blue case, but looked identical to the Color Computer otherwise; it can be found in old Radio Shack catalogs now online, if you care to look.
Did Tandy/Radio Shack have some kind of exclusive license deal with Motorola, to only use Motorola or Tandy-designed parts for the machines? I don't think the SAM was a Motorola component (?); and later with the Color Computer 3 it had it's own custom chip (the GIME) too - but everything else was Motorola...
The price point of the CPU prevented them from putting multiple CPUs in the machine (I often wonder how things might have been different had there been a second 6809 CPU in the Color Computer to handle sound and/or graphics); but what prevented them from putting in a cheaper CPU (like the 6502?) or a sound chip (there were a few cheap ones out there made by TI and others)?
Instead - for additional sound capabilities, you had to purchase cartridge expansion - either third-party or from Radio Shack (the Speech/Sound Pak, or the Orchestra-90 Pak), as well as an expansion device (the Multi-Pak interface), especially if you already had a floppy disk system and/or modem/rs-232 pak. That upped the price considerably, and as a result, few did it, and so those sound options were not well supported in games or other applications.
Instead, the built-in sound was exploited as far as things could take it at the time (and more recently, even further). Later, to improve the graphics of the system with the release of the Color Computer 3, Tandy added the GIME chip, which basically took over the duties of the SAM chip and 6847, but only partially (semigraphics modes were not available - all other modes were, though); but it was a seemingly in-house designed chip, and they also had to have a third-party update the BASIC (that was originally provided by Microsoft) to support the new resolutions and colors enabled by that chip.
The company that did this was Microware, who had long supplied a multi-user multi-tasking OS for the 6809 called "OS-9"; they later produced a version for the Color Computer 3 called "OS-9 Level II" - not that this should not be confused with OS9 for the Mac (totally different products).
Microware essentially hacked the BASIC. The original code was from Microsoft, but they had moved on and didn't want to do the updates to the code, so what Microware did was kinda ingenious - they had their ROM of "patches", and on bootup, they would copy the ROM code for the MS BASIC over into RAM, then overlay hooks to their ROM routines in the RAM copy, then run the modified RAM copy of the BASIC interpreter; this was called "all-RAM mode", and it was something that was long possible to do with the CoCo 2 (and was used in a similar manner by third-parties to extend commands in BASIC for their own products and extensions). Microware's only changes were being able to do it at boot time. Anyhow, they extended it and it worked (mostly) well enough.
But even at that late date (circa-1986) - Tandy had to go with a custom chip, and not anything third-party outside of Motorola (from what I recall, Motorola didn't have any display chips of a better spec for the 8-bit line; they had mainly moved on to their 16 bit 68K lineup). So what was preventing them? I honestly don't know...
Interestingly, some time in the early 2000s it came to light that Microware had probably the only (maybe outside of Tandy itself) "prototype" of the Color Computer 3 - it was a giant wire-wrapped board; it's most curious feature being that it lacks a GIME chip! This is especially important as there is no documentation/datasheet or internal information about how the GIME chip worked, outside of what was provided to programmers. In short, the GIME chip is an "unobtainable" component (there are two versions - one from 1986, the other from 1987, with minor "bug fixes"). It is hoped that this rare board might provide clues as to how the GIME chip was designed, but right now the board is a curiosity (I can't recall who it was given to, but it was given to that person by employees of Microware, which was still in business at the time, and may still be (?), who found it in a back closet or something like that).
IMHO the 8-bit Ataris were a lot more innovative than the Commodore 8-bitters (including the C64), despite being a couple years earlier to market.
For instance the Atari display chip ANTIC was an actual co-processor, running a sequence of simple instruction via DMA to produce the display (very similar to the Amiga's copper).
The C64 (and other 8-bit machines) could only reprogram the display chip mid-frame by directly writing hardware registers with the CPU at the right time (often controlled by a raster interrupt).
The 8-bit Ataris were the actual predecessor to the Commodore Amiga (not surprising, because the same people who designed the Atari later designed the Amiga).