Commodore introduced its final 8 bit computer, the Commodore 128, on January 5, 1985 at the Winter Consumer Electronics Show (CES). It was an unusual and and often misunderstood machine. I have never seen another product subject to so much armchair quarterback. On its birthday, let’s take a look at what this computer was and why it was built the way it was.
The last of the 8-bits
It was clear in January 1985 that the era of 8 bit computers was coming to an end. It wasn’t a matter of if, it was a matter of when. The IBM PC was gaining popularity. And three companies, Commodore included, had advanced new machines based on the Motorola 68000 CPU either on the market or coming very soon.
The Commodore 128 existed solely to solve 2 problems. The Commodore Amiga was not ready, and probably wasn’t going to be ready until fall. And it was going to be expensive. Commodore needed something more expensive to sell alongside its $149 C64. And having something priced in between the $149 C64 and $1395 Amiga seemed prudent.
Commodore’s engineers had been tinkering with the idea of some kind of upgrade for the C-64. Commodore management rushed it to market after they found out about it.
Commodore 128: A hodgepodge of spare parts
The Commodore 128 seemed like a hodgepodge of leftover parts, mostly because it was. It was essentially a C-64 with a Z-80 CPU, an 80-column video chip, and some extra RAM bolted on. The 80-column video chip was a leftover from the cancelled Commodore 900 project. The Z-80 provided CP/M compatibility but also solved a compatibility problem with the C-64. Certain poorly behaved C-64 cartridges wouldn’t start on the C-128. But since the Z-80 starts from a different memory address than the 6502, they solved the problem by letting the Z-80 initialize the machine and bring it into a state where those poorly behaved cartridges worked, then hand control over to the 8502 CPU, a 6502 derivative that handled C-64 and 128 mode.
The Z-80 powered the machine’s third mode, which ran CP/M.
The native C-128 mode addressed the major shortcomings of the C-64. It had a more powerful implementation of the Basic language, to make it easier to program. It had a higher-resolution RGB display that could display 80 columns of text, making it more useful for word processing and spreadsheets. And it had a faster serial bus to allow a faster disk drive than the notorious 1541 drive sold with the C-64. The 1570, 1571, and 1581 disk drives were all faster when used with the 128 than with the C-64.
It was an upgrade from the C-64 that retailed for $349 and could be discounted to $249.
Was the Commodore 128 a success or a failure?
Since the Commodore 128 didn’t sell nearly as well as the C-64, computer fans have argued about it ever since. How many units it ultimately sold is controversial, but it was far less than the C-64 no matter how you slice it. Initially it sold well, but sales dropped off once the Amiga 1000 hit the market in November 1985 and never recovered.
But indirectly, the 128 didn’t have to sell to be a success. By late 1984, demand for the C-64 was tailing off because of a lack of an upgrade path. The existence of the 128 helped slow the decrease in C-64 sales, since a natural upgrade path existed. The C-64 would not have sold 2 million units in 1985 without the C-128 protecting it in the lineup.
Re-imagining the Commodore 128
Whether Commodore should have made the 128 differently is a subject of another blog post. But let’s address two of the common criticisms, that the 128 should have been more like the Apple IIgs.
The Apple IIgs didn’t exactly set the world on fire sales-wise either. In both cases, the presence of a middling machine got people in the door, and sometimes they chose a cheaper 8-bit machine or a pricier 16-bit machine instead. Simply having a third choice made them more willing to consider the product line.
It would have been nice if Commodore had used a 65816 CPU like the IIgs did, but there was a problem with that. The Commodore 128 shipped in 1985, after the 65816 CPU was announced, but well before it was available in quantity. Commodore couldn’t afford to wait for the 65816.
Similarly, a video chip that was fully backward compatible with the VIC-II while offering higher resolution and colors would have been nice. But Commodore didn’t have time to develop that and get the machine to market quickly.
The Commodore 128 was the best machine Commodore could show working in January 1985 and deliver in quantity soon after. It existed in two forms: the wedge-shaped “flat” C-128 that didn’t include a disk drive, and the C-128D, which included a 1571 disk drive in a PC-style enclosure with a detached keyboard.
An end user perspective on the Commodore 128
I actually had and used a C-128 in the mid/late 1980s, so I can provide an end user perspective on it. Overall I liked it. I liked being able to write programs with graphics and sound. The 80-column video chip it had was always controversial but it had interesting potential that was fun to explore.
Lack of native-mode software was frustrating but also inevitable. The 64 was the lowest common denominator, so relabeling all software to “Commodore 64/128” was the safer bet than developing native C-128 software. In theory, Commodore could have primed the pump by producing some first party software like they had in the early days of the C64. But by 1985, they regarded internal software development as overhead. They produced a handful of productivity titles but that was it. The major publishers of C-64 productivity titles ported their titles to the C-128 with enhanced capabilities, and Infocom made a few native C-128 games.
Was the Commodore 128 fully C-64 compatible?
I also frequently hear criticisms of its C-64 compatibility, which I don’t understand. In C64 mode, it was very nearly 100% compatible. I only ever found one program that didn’t work, and it was a pirated title, so it’s not like it cost me any money. To this day, I don’t know if it was the fault of the publisher or if the cracking group introduced the incompatibility. The thing about lack of 100% compatibility is not every revision of the C64 was 100% compatible either. And the Apple IIe and IIc fell slightly short of 100% compatibility with each other too. So the C-128 gets penalized for a flaw that its peers also had.
The only explanation I have is that maybe people who say it wasn’t very compatible are trying to run C-64 software from native C-128 mode. Some titles were C-128-aware and would load in 128 mode to boot faster, then switch to C-64 mode. But the memory map between the two modes was very different, so a program had to be coded specifically to do that. If you switched to C-64 mode and then tried to load a program, it generally worked. Like I said, I only ever found one title that didn’t.
David Farquhar is a computer security professional, entrepreneur, and author. He has written professionally about computers since 1991, so he was writing about retro computers when they were still new. He has been working in IT professionally since 1994 and has specialized in vulnerability management since 2013. He holds Security+ and CISSP certifications. Today he blogs five times a week, mostly about retro computers and retro gaming covering the time period from 1975 to 2000.
It’s interesting to read Brian Bagnall on the C128. While pursuing development of the machine, Commodore in its usual wisdom simultaneously decided to close down its in-house software division, which could have jumpstarted the market for games that took advantage of the C128’s upgrades over the C64. So the machine never really took off as much as it could have, although of course it sold reasonably well regardless.
I agree, one reason the 64 had a chance was because Commodore produced a couple dozen games and a dozen or so productivity titles for it so it had an instant software library. For the 128, Commodore commissioned a handful of titles but nothing like what they did for the 64. Its native mode was practically an orphan machine within a year of its release.