Commander X16 thread - A brand new 8-Bit computer!

I've been somewhat following this and it's very cool but I don't get it. It's strange because I'm the kind of person who should be the perfect target for something like this, someone who really likes these machines and this period of computing history.

I know it's not supposed to be a raspberry-like device, it has a different scope and purpose as a new 8-bit machine but, not being backward compatible with any of the original 8-bit machines, IMO its usefulness is really limited.

For someone who still has some of the original machines and/or an FPGA version and uses them, what's the point? It can't replace them, being incompatible, so you can't justify it as a replacement for if/when one of those old machines finally dies or as a way for using less of the original ones, increasing their lifetime, like the FPGA.

But for someone who wants a new device to tinker and to add to their collection, it's very nice (although expensive).

Maybe I'll try one of the future, less expensive, iterations.

VectorVoyager said:

Anyways, I have the same hope with you. I hope gen 2 and gen 3 will be cheaper so more people can appreciate David and others' work.

Click to expand...

Sure! One thing I didn't mentioned earlier was how I admire this project and that I hope it is successful, even though I'm still not convinced it is for me. It deserves to be successful.

VectorVoyager said:

Your points are valid. I didn't buy one neither and would instead buy an FPGA or just emulate the system I wanna use on an SBC like Raspberry Pi but David opposed FPGA/emulation from the very start of the project because he wanted to have a completely real 8-bit system made with completely new parts, which proved to be quite the challenge! Ironic thing that they chose the Yamaha chip for X16 of which supply issues made the product even more expensive and late to production!

About the backwards compatibility I can't agree with you. X16 is its own thing. Making it a Commodore-compatible machine would undermine its "superior" capabilities as many people would use it as a generic C64 and not try its additional features. I think what X16 needs is old games getting rewritten + enhanced for the system which is done slowly by the small-mid community it has. Therefore X16 actually needs more people interested in it, which is hard as the product is expensive as you have stated. Maybe David should do contests etc to attract devs to write software with the emulator that would enlarge the software category.

Anyways, I have the same hope with you. I hope gen 2 and gen 3 will be cheaper so more people can appreciate David and others' work.

Click to expand...

The opposition to an FPGA is weird since he is using one for the sound/video Vera chip.

Neat idea, but the price is much too high for me personally to consider buying one.

I have a huge respect for the engineering and sheer amount of effort it must have taken to get something like this into production, but it's definitely not for me. I get my (limited) retro-computing kicks from emulation of the platforms I used back in the day, be that Atari ST, Super Nintendo, Atari Lynx, etc...

I've followed this project on and off as it came together, and I highly recommend the 8-Bit-Guy channel if you're at all a fan of retro/8-bit computing. But the X16 falls into that great big black hole of possible hobbies that I just don't have time for. Like steelghost said, emulation has gotten so good over the years that I feel like I can dive into the games, software, etc. I enjoyed on my Commodore machines as an 80s kid really easily already.

The one thing that I think about is getting back into 65xx machine code hacking, which way-back-when was the last time my artist's brain was able to understand the logic of programming, but again... tackling that on a whole new platform is not super likely with all the other things that compete for my time. I hope the overall project succeeds, however, as the idea of a brand new 8-bit platform in 2024 is just the best.

I’ve followed the genesis of the X16 through Youtube videos from the start, and also took the emulator for a spin, actually writing some BASIC code.

I think the raison d’etre for the Commander X16 is twofold: first, it’s a hardware project, using 8-bit era components that are still available new today. That worked for the CPU, RAM and so on. Less so for the sound chip, where the choice ended up with "new old stock" but that turned out less available than expected, and it failed with the video part. So they went with an FPGA (cheating!). Perhaps they could have found some VGA chips that are still being made today? Or more modern video chips and run them in VGA compatibility mode? So if you want to build a somewhat useful computer with an actual 6502 today, this is it.

Second, there’s the software side. I grew up learning how to program the C64 and later C128. So the dumb C64 BASIC, the much better C128 BASIC, Forth, for which I bought an interpreter on cassette tape, and of course 6502 assembly.

And I think that is a really good way to learn the basics of computer architecture. So it’s great that there is now a system that will let you do that, but which is not decades old and ready to fail, with no reasonable options to connect to a modern display. Although IMO, this is also a weak point of the X16 as it’s all VGA or worse, no HDMI. And something that drove me batty using the emulator, there is no C64-like border in VGA mode. The characters are pressed directly against the edge of the display. This looks terrible. But at least VGA is something that looks good using a 20 buck adapter to HDMI, unlike the TV outs from 1980s 8-bit computers.

The advantage of the X16 is that its CPU can run at 8 MHz. This means that it’s actually somewhat possible to write games in BASIC. An 1 or 2 MHz 6502 or a 3.5 MHz Z80 is just so insanely slow that it’s just too frustrating. Also, the old 8-bit computers had weird keyboards, slow / no longer available storage and so on.

Also note that although it is not hardware compatible with any Commodore stuff, AFAIK they got a license for the C64 firmware and the BASIC and KERNAL are actually very close to fully compatible with the C64.

I would have made other choices, like prioritize HDMI and using the 16-bit version of the 6502, the 65816. Or perhaps the 68000. That is such a clean and forward-looking design. But all in all I’m happier to live in a world where the X16 exists compared to a world where the X16 doesn’t exist.

BTW, I would be happy to do the Ars review of the X16.

I don't consider using the FPGA for graphics/video cheating since it's roughly similar to spinning up a custom chip back in the day, except FPGA makes the low volume affordable.

There are other systems like the Foenix that might interest people in terms of capabilities/features that live in the same ballpark.

kefkafloyd said:

I don't consider using the FPGA for graphics/video cheating since it's roughly similar to spinning up a custom chip back in the day, except FPGA makes the low volume affordable.

Click to expand...

Yeah, but then why use "real" old school parts voor the CPU and VIAs?

Obviously just the VERA FPGA doesn’t turn the X16 into a MiSTer (an FPGA-based hardware emulator of many old computers), but it does detract from the charm of "realness" a bit.

I’m also conflicted as to whether just using a Raspberry Pi, and either take advantage of modern languages etc or emulate the old stuff in software isn’t a better solution. One thing I do like about the real or emulated 8- or maybe 16-bit stuff is that it’s all so simple without unfathomable amounts of software sitting between you and the hardware.

iljitsch said:

Yeah, but then why use "real" old school parts voor the CPU and VIAs?

Click to expand...

Because David has more money/clout than sense. The Foenix 256K is, IMO, a better/cleaner implementation of a similar idea while still using a 65C816 for its CPU. Remember, he wanted it to be mostly through-hole at first, then he tried doing his own wave soldering until finally conceding that for an actual manufacturing run he needed to work with a real manufacturer. The project's been plagued with problems due to short-sighted decisions but at this point it's in a decent enough state thanks to learning from said mistakes.

There's also other system design considerations where these affordable FPGAs do have limitations; it's likely easier to design a system where a cheaper FPGA can be used solely for graphics/sound acceleration while leaving CPU and other functions on other chips.

iljitsch said:

Obviously just the VERA FPGA doesn’t turn the X16 into a MiSTer (an FPGA-based hardware emulator of many old computers), but it does detract from the charm of "realness" a bit.

Click to expand...

There's not really another off-the-shelf solution for graphics processing for this class of machine. It's not just about basic video output; it's about image processing that we'd expect from chips like VIC-II and Paula. The whole point of FPGA is the ability to spin up custom circuits without spending a boatload of money on real ASICs. If they made enough of the systems to make an ASIC worthwhile, there's nothing stopping them from replacing the FPGA with an ASIC. And if the FPGA was replaced with an actual ASIC, there'd be no difference in the functionality or overall design, except for the fact that an FPGA can be updated at any time.

Solutions like TinyVICKY and VERA are functionally equivalent to using Denise and Agnus, and implementing them in FPGA versus an ASIC shouldn't be held against them in terms of "realness."

kefkafloyd said:

There are other systems like the Foenix that might interest people in terms of capabilities/features that live in the same ballpark.

Click to expand...

Didn't knew them until your post. I'm trying to understand their (expensive) products but their site is a bit confusing... For example, what do those FDD drives have that make them worth 300 to 500$ ?! I mean, I can buy an adapter that turns any old PC FDD into a Commodore compatible drive for 40$. Is it just because of the cool custom case?

Yeah, the Foenix is an even weirder amalgamation of old and new stuff, and it’s even more expensive to boot.

How hard would it be to have true classic CPU such as the 65816, RAM and ROM, and then interface with a Raspberry Pi through the serial SPI protocol (it seems the W65C22 VIA can do that?) or parallel using a W65C21 PIA?

So basically this thing.

Then use the the Pi for all I/O, including display. This does mean that all display data must go through a generic I/O chip. That will probably be slow, but it would be possible to optimize some things to minimize the amount of data that needs to be transferred.

And you could even load image and audio data from the Pi’s storage to the Pi’s RAM where the Pi could build the display or play audio as directed by the 8/16-bit CPU without that CPU having to touch the data itself. So it may still be possible to write games!

iljitsch said:

Yeah, but then why use "real" old school parts voor the CPU and VIAs?

Obviously just the VERA FPGA doesn’t turn the X16 into a MiSTer (an FPGA-based hardware emulator of many old computers), but it does detract from the charm of "realness" a bit.

I’m also conflicted as to whether just using a Raspberry Pi, and either take advantage of modern languages etc or emulate the old stuff in software isn’t a better solution. One thing I do like about the real or emulated 8- or maybe 16-bit stuff is that it’s all so simple without unfathomable amounts of software sitting between you and the hardware.

Click to expand...

I think I rather have a Raspberry Pi or use one of the new RISC-V SBCs out there instead of the Commander X16. This is really a Niche Product.

I got the Raspberry Pi 400 kit that comes with a book that’s a lot like the books that taught programming on computers like the C64 or ZX Spectrum back in the 1980s. If I were a kid today I’m sure this would be right up my alley. But I’m not a kid and haven’t started with the book just yet...

If Apple feels like a really high end hotel room and Windows like the blandest possible 9-to-5 office building, the Pi feels like camping in the woods: sometimes a bit primitive and you have to do a lot of stuff yourself, but endless opportunities to explore. I’d say those 8-bitters are a little play house on a playground: not much there, but within the limitations there is tons of fun to be had.

iljitsch said:

I got the Raspberry Pi 400 kit that comes with a book that’s a lot like the books that taught programming on computers like the C64 or ZX Spectrum back in the 1980s. If I were a kid today I’m sure this would be right up my alley. But I’m not a kid and haven’t started with the book just yet...

If Apple feels like a really high end hotel room and Windows like the blandest possible 9-to-5 office building, the Pi feels like camping in the woods: sometimes a bit primitive and you have to do a lot of stuff yourself, but endless opportunities to explore. I’d say those 8-bitters are a little play house on a playground: not much there, but within the limitations there is tons of fun to be had.

Click to expand...

hello, what book came with your raspberry pi 400 kit?

VectorVoyager said:

Yeah, he was eventually forced to use FPGA at VEGA for video and sound as the alternatives were very poor quality and/or out of production. Using a Y2151 for an additional sound channel already created tons of problems. If he were to not have a sound/video chip and make a single chip 65C02 board then the hardware would be lesser than 80s microcomputers. Its near to impossible to make 100% 8-bit computer with good specs for 80s with today's supply after all.

Click to expand...

You can easily get "late 70s" 8-bit CP/M recreations, mostly Z80 https://rc2014.co.uk/ also 6509. Also easy for CP/M to simulate/emulate DOS, Digital Research basically did this back in the day. and anyways the problem is it should be a 16-bit machine! Why the 16 in the name? IMO kinda ridiculous to use authentic chips AND try to get it to play games like a recent stock "vintage game console" with HDMI, or based around RPi or similar board. The problem with Pi et al., is that it can't recreate timings of classic machines, a problem Mister solved with FPGA. I think there's room for a MiSTer competitor, to bad this wasn't considered from the start. Reminds me of many stories of genius engineers totally screwing up marketing (ha! a recreation of the 80s in the bad way) or they could have gone the RC2014 way, which BTW can be upgraded to play MSX machine games. Basically, not only a pass for me, I don't see a reason for Arstechnica to cover this, or maybe as a post-mortem.