12 minutes ago, Scott Robison said:
I'm not @rje and I haven't talked to him about it. I didn't *laugh* at your specs, but they seem overly optimistic to me for the technology that was available in the day for "consumer hardware" or effectively a game console.
Don't get me wrong, I loved my C=64 & C=128DCR back in the day. Wish I still had them. That's what appeals to me about the CX16. But I don't see any 8-bit being a realistic path forward for much past the C=128. I think they could have created upgraded C=128s that had more RAM and perhaps had a few more sales (I mean, 5.7 million C=128s were sold, that's not bad, it's only bad in comparison to how many C=64s were sold over the years in combination with the profit margin of the more expensive hardware). But the world was moving away from all in one computers other than for game consoles. By all in one I don't mean the modern definition, I mean "you get this one video chip and this one sound chip". The world was demanding expandable / upgradable hardware. That's why I think they had to focus more on 16 bit / 32 bit platforms. They just didn't know how to market anything that wasn't a toy, more or less.
The 256 would have two cartridge slots a-la the MSX architecture, plus the user port and the RS-232 port, not to mention the standard datasette port and the daisy-chainable floppy port. The 640 in either horizontal upright or monitor tower versions will have at leas four and possibly as many as eight expansion slots, each with at least four times the bandwidth of ISA as implemented at the time. Either having MOS Technology develop an FPU solution for an expansion card or cartridge, or commissioning a third party to do so would have hardly been beyond Commodore's means. And there is a logical CPU upgrade path in the form of the upcoming 65832 core architecture.
And as for computers with good chipsets already baked into the system being on their way out, consider that the Sharp X68000 sold over 16 million over a lifetime of seven years (1987-1994) and three major chipset revisions. If Sharp had marketed a version in the West without the expense of the Kanji font ROM and the Video RAM to support it, it could have sold at least half as many more, allowing it to beat out the Commodore 64 and NEC PC93/98 for best selling computer series before the Windows 95 era.
The 640 would have all the ease of use of the Macintosh, but in living color, with better sound than the beeper, an actual language baked in ROM, and be just as upgradable, for less than half the price, with at least double the RAM, and with all the PET, CBM 400-800, VIC-20, Max Machine, and 64 software stack available right away (with caveats involving PET's 8' floppies and {for the 640} VIC-20 cartridges, but still). The then new Mac, on the other hand, would have had the Microsoft Office suite and that's about it, and to program it, they would have hade to write a compiler or interpreter for it from scratch (Smalltalk and Hypercard only came out in '86 and '87, respectively) with a data analyzer.
Apple would have been (even more) screwed, and even the PC cloners would have gotten a run for their money. The 16/32 bit era can wait for 1987 to fully blossom, by which time, I would have had the next machine in the series ready...