So with the launch of OpenFPGA on the Analogue Pocket, along with the development Analogue Pockets making the rounds. I'm curious to know what the chances of Pico-8 appearing on the Analogue Pocket?
I haven't got the foggiest about programming cores for FPGA, but from what I'm aware is that Pico-8 is mainly a piece of software that sits on top of an existing system, this makes me wonder what kind of core would need to be made to make Pico-8 work on an FPGA chip, be it either the Analogue Pocket or even the MiSTer. I would love to have all my GameBoy games and Pico-8 games on my Analogue pocket when I receive mine later this year.
Thinking of which, could there be a means to run Pico-8 on a GBA core and run it like that on an Analogue Pocket and MiSTer?
This has been suggested before, but it's orders of magnitude more complex to do than most people initially think. @dddaaannn actually had a good comment about this recently in another other thread.
In this case, I'd be less concerned about the potential processing power of the FPGA than the FPGA-implemented architecture it would take for PICO-8 to run. PICO-8 isn't a real CPU architecture, it's a runtime environment on a host OS. The fastest way to get it running on a MiSTer or Pocket would be to install a conventional x64 or ARM architecture core, run Linux on top of that, and run PICO-8 on top of that. That might get it on the device (and if that's the goal then there you go, at least in theory :) ) but it doesn't exactly take advantage of the FPGA-ness of it.
I would doubt that a custom architecture specifically for a custom PICO-8 runtime would be all that compelling. The PICO-8 runtime includes a full Lua interpreter, which compiles internally to Lua bytecode that probably can't be meaningfully built into a custom CPU, which means there's a bytecode interpreter on what's probably a conventional CPU. I won't stop anyone who wants to take a crack at a custom but conventional CPU, a C compiler, a Lua interpreter, and a PICO-8 runtime on top of it. But PICO-8 would probably still live entirely in software. There might be some fun in recreating the I/O peripherals (sound?) as if they were hardware chips. Give PICO-8 a couple of SIDs! ;)
Hmmm... So Pico-8 would need an operating system too. I was hoping that maybe it could just be an ARM-based piece of software that could be loaded like a GBA game. Clearly, it isn't as simple as that, but as mentioned before, if there was a means to get Pico-8 running on a GBA, that would simplify the chances of getting Pico-8 on the Pocket.
Well damn, that's the first time I discovered that the GBA over has 256KB of RAM, damn...
So yeah, you really need something more modern to make that work. Shame.
If we're talking platforms, @SepticLemon, APK EXPORT would be perfect. That's 11,000 hardware right there including all cellphones and tablets.
Failing that, yeah, those early portable gaming systems don't offer too much RAM. When I was programming GBA I was limited to 64k of RAM space myself.
I worked with it, but it was rough.
FPGA is playground processing. No need to cast a chip. You program what it should do. I think Pico8 has enough hard limits to be able to translate it into an FPGA system. I however am not the technical genius to do it, but if Zep was ever able to find a collaborator with FPGA know how, I think it could happen.
Basically the real use of an FPGA is to prototype new chip/hardware designs. Emulation and all that is just a fun side effect for us old timers. Thus, a true Pico8 Chip could be devised using FPGA. Whether it can handle the current virtual version of it with the state of FPGA as it is now is...well, that's to be discovered. However, if the work was put into it, then the fantasy console would become a reality console. How neat would that be?
[Please log in to post a comment]
