Live data from Hacker News

Commander X16

commanderx16.com

51–60 of 116 posts

Re: Commander X16

#51
post #43

Earlier quoted context omitted.

The MiST does all of that: https://amigastore.eu/en/358-mist-midi-fpga-computer-with-mi... 221,95 Euro (Approx 271 USD)

Close enough. Do I need to buy the "Full Amiga Configuration" or can I build that myself since I had an Amiga? EDIT: some other sellers: https://www.amedia-computer.com/en/accueil/275-mist-midi-13-... https://www.dragonbox.de/en/consoles/other-consoles/mist-fpg...

You can build that yourself, provided you have images of the Kickstart ROM and Workbench disks. You can boot the thing off of a hard disk image, the same kind used by modern *UAE emulators, or you can boot it off a floppy disk image.

Re: Commander X16

#52

All these machines end up involving one or more FPGAs -- to varying degrees -- simply because of the requirements and logistics. There's varying types of purism involved ("FPGA only where we would have used a custom chip" or "FPGA but not for the CPU" etc) but once you start down that path the whole thing starts to seem silly as with many of these machines the entire functionality could be done on a single relatively…

In a lot of cases, they aren't left with much of a choice. Many chips aren't in production any longer, either due to them being custom chips or due to them serving a need that the market no longer has. Stuffing everything into an FPGA isn't really a solution in this case since it effectively makes it a hands-off machine.

What do I mean by hands off? In terms of software, it doesn't make much of a difference. You can still develop for the machine and it will appear to be the genuine article. In terms of hardware, it is a different ballgame. At best, slots and ports may be exposed so stuff can be added to the machine. It will be a bit like using a modern desktop, albeit with the opportunity to develop your own expansion cards since the expansion bus can be more accessible than PCI-e. On the other hand, most modifications to the hardware itself must be made through the FPGA. The FPGA route will be more flexible, but it requires an entirely different skill set. Relatively few people have the skill or interest to deal with FPGAs, particularly if they are seeking something more comforting than modern hardware.

How is replicating something like the video chip with an FPGA somehow acceptable? Presumably it is because it has to interface with the rest of the system in a conventional way and since people were unable to modify the video chip in vintage computers anyhow. Since it is interfacing with the system in a conventional way, you can also mess around with the video circuitry surrounding the chip.

Re: Commander X16

#53
post #7
post #4

Earlier quoted context omitted.

Defining success is tricky for these. For some of them I think the people creating them are perfectly happy with just getting something for themselves and a handful of enthusiasts. I agree with you, though. If I absolutely wanted physical hardware, I'd get a real one, or an FPGA reimplementation on a general board like MiSTER [1] that'd also let me run other cores, or something like the Ultimate 64 [2] (fit's in a re…

An FPGA implementation is missing the big educational aspect though, namely that this product probably intends for it to be feasible (just like it was with the 8 bit computers back then) to fully understand the hardware, prod it, modify it, and interface to it. Usually up to IC boundary, but the key there is that the ICs are distinct and (hopefully) not too complex. Even a fully "open" modern platform is usually just…

I personally didn't find Verilog all that hard to learn, and it's frankly simpler to understand and modify the "hardware" in that form than using a multimeter and logic analyzer and futzing with solder. I built myself a simple retro-style machine around a RISC-V core and my own custom designed video controller in a few weeks, without any prior Verilog experience. Although I am a software engineer, I don't feel like those skills actually helped that much.

Re: Commander X16

#54
The retro designs that I really want to see made real would be the ones done by "Arne"[1].

Most of the designs would be impossible, being pretty steeped in trademarks, but they've explored some interesting ideas, like alternate keyboard layouts (ABC, 4 rows etc.) and mashups of all kinds of "retro" machines.

We're also not seeing that many 16-bit faux-retro designs (i.e. not straight-forward Amiga/Atari clones). Either because they tend to be more complex, or because this follows the same trends as fashion and music, where we e.g. had 80s synthwave followed by 90s vaporwave and now more 00s sounds and looks. Since the time for that should've come by now, I think it's more the complexity angle. "Just" hooking up a 6502/z80 seems an order of magnitude easier than coming up with whatever the equivalent of custom chips would be. Never mind that because of the scale economy, you'd have to do the latter with FPGA, which is a step down on the ladder of retro authenticity.

[1]: https://androidarts.com

Re: Commander X16

#55
post #2

I'm a big retrocomputing enthusiast (in the C64 and Apple // realms) and the people doing this project are well-known in the community, but I just have trouble believing this will be successful. I don't want a computer that's kinda like a C64. I have a C64, and anyone who really likes that platform can either buy one on ebay or emulate it. A new platform that is a shadow of a platform from 40 years ago is such a nich…

It is like model railways. You are selling to an older demographic with little interest from younger people.

Older people do have money though. In 1983 it was not possible for them to upgrade to a C64 from their ZX81 or VIC-20 because they lacked the money and Christmas only came once a year.

Nowadays though they can buy these toys without having to leave their chair.

Re: Commander X16

#56
post #46
post #13

Speaking of "modern retro computing" there is already a player in the town: C256 Foenix [1]. Based on 65c816, affordable, powerful and - most important - already available and well tested. It definitively deserve for user interest. For curious - there is an emulator available [2] (even one-and-half, because mine [3] is rather limited and created mostly for Forth [6, 7] development). 320x240 (double-pixel), 640x480, 8…

I thought the Foenix was canceled. We had a discussion about it the last time the Commander X16 came up: https://news.ycombinator.com/item?id=23801266

Yes, I thought so too, and I also thought it sounded like a shame because the tiny scraps of information I'd heard sounded good.

But there's no info about it on that website. Is there anywhere that I can read about design, specs etc?

Re: Commander X16

#57

When David Murray (8-bit-guy) first announced his 'dream computer' I was quite excited. His early descriptions were akin to something like a C128 but with more colours, faster CPU and modern connections, all in a single case, like the C128(not D) or C64. Plug into a screen or TV and just 'go'. Somewhere along the way serious scope-creep happened, and now we've got something that looks and feels just like a mini-ITX D…

Where did it lose the retro vibe for you? I see a board with very few simple chips where you could reasonably understand the whole system top to bottom. The hardware is still very limited.

Re: Commander X16

#58
post #7

Earlier quoted context omitted.

An FPGA implementation is missing the big educational aspect though, namely that this product probably intends for it to be feasible (just like it was with the 8 bit computers back then) to fully understand the hardware, prod it, modify it, and interface to it. Usually up to IC boundary, but the key there is that the ICs are distinct and (hopefully) not too complex. Even a fully "open" modern platform is usually just…

I personally didn't find Verilog all that hard to learn, and it's frankly simpler to understand and modify the "hardware" in that form than using a multimeter and logic analyzer and futzing with solder. I built myself a simple retro-style machine around a RISC-V core and my own custom designed video controller in a few weeks, without any prior Verilog experience. Although I am a software engineer, I don't feel like t…

Kudos to you if you can go from zero to your own custom video controller implemented in an fpga in a few weeks.

But for normal humans that just want to get involved with hardware and dabble somewhat with software/firmware in the context of vintage computing, taking a detour into fpga-world would be a deep unpleasant rabbit hole unless one was specifically interested in fpga's and there was no other viable option (which in the case of custom ASIC's is admittedly quite possible).

Compared to the nightmarish complex toolchains used for fpga's, digging up some datasheets and working with a DMM, scope, and logic analyzer is a breeze and actually quite pleasurable.

Re: Commander X16

#59
Ah, I still recall the April 1st joke about this: https://www.youtube.com/watch?v=PWBUMgAl7wY which was one of the better ones by far this year.

The channel "Retro recipe" on YT been mostly driving this and has few video's over time with updates and news nuggets upon the progress and now - finally here. Worth checking out the channel if your into that kinda stuff and one of the better ones for that retro nostalgia.

Re: Commander X16

#60
post #13

Speaking of "modern retro computing" there is already a player in the town: C256 Foenix [1]. Based on 65c816, affordable, powerful and - most important - already available and well tested. It definitively deserve for user interest. For curious - there is an emulator available [2] (even one-and-half, because mine [3] is rather limited and created mostly for Forth [6, 7] development). 320x240 (double-pixel), 640x480, 8…

This sounds like a much better spec. Using the 65c816 keeps the 8-bit vibe but doesn't force you to play the bank switching game to access more than 64K. DVI instead of VGA itself is also enough to give it a much better chance of success.
Post reply on HN