tufty wrote:The C128 was released in early 1985, and failed because a: it was shit
But that's the thing. It didn't fail.
It sold millions of units. By some sources, about 4.5 million units sold - which places it comparable in sales to the Apple II (5 million sold) and outselling most other computers of the era.
I'm agreeing that it was shit, for what it could and should have been, but it was not a commercial failure in the slightest. What it was, was a horrible missed opportunity.
Indeed, what I'm asking for did already exist. The BBC Micro had a 2MHz 6502 CPU and 4MHz RAM speed, so that CPU and video can alternatively access the same RAM. Acorn did this in 1981 through third-party manufacturers. Commodore - who owned MOS Technology to fabricate their own chips - could have done this in 1985 easily.
Honestly, it's not remotely an unreasonable request. It's even the obvious engineering thing to do. Many 6502-based systems followed this "CPU and video take turns to access the memory bus" with RAM clocked double the CPU and video chip speed. The C64 did this with a 1MHz CPU, a 1MHz VIC on a 2MHz bus. So if you're upping the CPU to 2MHz, then you up the video chip to 2MHz and put it on a 4MHz memory bus (like the BBC Micro already did).
And note that if you were not upping the VIC's actual performance, so it's the same VIC-II, then you could even have had it still operate at 1MHz, but just able to handle a 2MHz clock signal coming in. The scheme could have gone from "CPU - VIC - CPU - VIC" to "CPU - VIC - CPU - nothing - CPU - VIC - CPU - nothing". So the VIC's hitting the bus exactly as often as it used to - 1MHz - and it just surrenders its second time slice. So, really, the only change you need to make is that it can accept a 2MHZ clock, but just ignores every other clock to still operate at 1MHz.
I mean, if you are making changes to the video chip then, yes, I'd encourage you did more than that - add new registers and capabilities, actually making it perform at 2MHz so it could grab more data and be less in contention with the CPU ("bad lines" and so forth).
But the most minimal thing here was just make the VIC-II still work when you switched the 6510 to 2MHz. That's it.
So you didn't have to do the supremely weird thing of sticking two video chips into the same machine. Just make the VIC happy to work whether the CPU is in 1 or 2 MHz mode. One video chip, which just works at either CPU speed.
This was the obvious, reasonable, feasible thing to do. So please understand how weird Commodore's decision here was to not simply fix the existing chip to work for both CPU speeds. But instead to develop a completely new fucking chip - the VDC - for running in 2MHz mode, while the old VIC was still in the box at the same time to run in 1MHz "C64" mode.
This was more expensive, more complicated and more stupid than what I'm asking them to do instead. Which was just to fix the chip you've already got to still work when you've upped the CPU to 2MHz. Because then you only have to pay for and design for one chip, not two. Then you can put the VDC development effort into just making a better "VIC-III" instead.
But they chose the "weird business decision" path rather than the good and obvious path. Which, you know, was unfortunately a thing Commodore went on to do a lot.
Because the business suits wanted "80 column CP/M" - for some reason - and they moved Heaven and Earth to make that shit happen, adding in an extra CPU for it. Yes, the Z80 was also the boot CPU, but only because, once you've got it in the machine, make it do something useful. Because, otherwise, in your "GO 64" mode, the Z80 was sitting in the machine doing fuck all. Sitting in there, hoping that someone might want CP/M at some point, which nobody did want to do.
An extra video chip, an extra CPU. So that we have, effectively, three completely different computer systems stuffed into one box at the same time. This was more complicated, more expensive, more confusing than what I'm suggesting.
In fact, Commodore themselves did eventually take this path later on with the C65. The real world "VIC-III" that could do all the old C64 modes, but added more graphics capability that, yes, it could do an 80 column mode.
I get it. Because I can be the weird radical one - my signature warns about my "wild words" on occasion - but, honestly, in this case, I'm being the reasonable one. The "voice of reason" to do the cheaper, easier and obvious thing, to Commodore being the weird "radicals" doing some mad shit that made no sense.
But, yes, the C128 - despite all my criticisms - was not a failure. It sold very well, thank you very much. Indeed, 4.5 million is just shy of the 5 million that the Apple II or ZX Spectrum sold throughout their lifetimes. This "sequel" of a machine might have been "Breakdance 2: Electric Boogaloo" in terms of being shit and bizarrely designed (again, the blame is not on Bil Herd for making the best out of some really bad business decisions, but on those bizarre business decisions in the first place).
But, hey, it made big box office. Like, people did go out in large numbers to buy "the new C64 with twice the RAM". Even if it was a disappointment for not being usefully more than that, when it could so easily have been.
Commercially, it wasn't a failure at all. And you might say that with the Atari ST and Amiga looming over it all, the 8-bits were dead. But Commodore did still ship 1 million C64s in 1989. You and I might have moved on by then, but it was still selling extremely well, even that late in the game. 8-bits became the "budget computers" for a while, even as the 16-bits took over.
And while I can't totally say that Commodore realised their mistake, as they started developing the C65 - far too late because, yes, by that point, the 16-bits had taken over - which actually does what I'm saying the C128 should have done, because many years had passed, so the C65 tech is more advanced than what you could have feasibly done for the C128 in 1985.
If they'd had the right "C65" attitude - make a "better C64" and double down on its gaming strengths, rather than wander off aimlessly to CP/M land - then the commercial success of the C128 would have been matched by an actually better machine, while folks waited and saved up for the Amiga. And, even in its "death throes", Commodore were still shifting a million "dead" 8-bit machines, even as late as 1989 - it would still have made commercial sense (and, indeed, the "C65" plans were exactly to pursue that but timed far too late to be useful, as Commodore was heading towards bankruptcy from their bad decisions by then).
And that's not only my take. Admittedly, the 8-bit guy was thinking a built-in 3.5" floppy would have been the thing to pursue - and he's got some good points - but I think, design-wise, my request is the simplest and most reasonable of them all. You've already got a VIC-II in the box. Just make it still work, when the CPU is clocked up to 2MHz. Just that. Don't have one of your components suddenly stop working when you switch the CPU mode. It's hardly unreasonable.
"Words ought to be a little wild - for they are assaults of thought on the unthinking."
- John Maynard Keynes
