|
Post by dshadoff on Mar 18, 2021 23:35:36 GMT
Unless dshadoff comes up with a way to load test programs from a PC, I'd recommend that you wait until you can find a PC-FXGA DOSV card instead of a standard PC-FX. I'll take that as a challenge. I did manage to identify the key address and data lines on the card-edge for the memory board, but I don't think I documented it here. Perhaps the best "first steps" approach is to come up with a board which extends that bus out of the machine for easy logic analyzer probing.
|
|
|
Post by elmer on Mar 21, 2021 19:31:31 GMT
Okay, I was wrong. This wasn't the z280. This was the ez80. So not quite a z280. It has z80 and z180 modes. Ah, yes, the eZ80 is pretty darned cool, especially when running at 50MHz, but you're talking about a chip from 2001 (or 2004 for the 50MHz version). Zilog made *so* many mistakes after the Z80, and having totally-incompatible extended instruction sets in the Z280, Z380 and eZ80 certainly didn't help them. OTOH, I do rather like their Z16F/ZNEO, which seems like a nice modern (but incompatible ) update of their Z8000 architecture from 1979. Anyway, back in the 1980s RAM speed just didn't increase at anywhere near the rate in clock-speed increase for CPUs, and so there really wasn't a way to make fast 8-bit CPUs, and increasing bus and internal processing widths was the only way to make big jumps in performance. That problem was IMHO what really messed up the 65C816, with its keeping the 6502's roughly 1/2 CLK per memory access rather than the HuC6280's redesign of the bus cycle, allowing it to run at 7.1MHz with affordable RAM. I'll take that as a challenge. Excellent! Since the PC-FXGA DOSV is so darned rare and expensive, I suspect that getting a board designed for the standard PC-FX is the only way that we'll allow modern developers to play with the PC-FX.
|
|
|
Post by dshadoff on Mar 22, 2021 1:35:06 GMT
The last thing that I did on the PC-FX was to help out with the Team Innocent translation. Did that ever get finished ? I seem to recall Filler and EsperKnight working on it, then you stepping in to help out... but I can't find any details on a release. Filler seems to list it as a completed script though.
|
|
|
Post by elmer on Mar 22, 2021 17:17:42 GMT
Did that ever get finished ? I seem to recall Filler and EsperKnight working on it, then you stepping in to help out... but I can't find any details on a release. Filler seems to list it as a completed script though. I did the font hack back in 2015, and then I helped Filler find some missing strings in 2019, which may still need to be translated. AFAIK, the whole text is now there, it inserts back into the game ISO, and everything is ready for testing/fixing. Unfortunately, that's not going to be a particularly pleasant task, because EsperKnight's text extraction does not record the size limits of each text string block, and so translation overruns and game crashes are inevitable. From a programming POV, the game is about as easy to translate as you can get. The game seems to be a 1st-generation C project, and the main executable looks to have been compiled from a bunch of individual C source files, with the compiler putting the uncompressed SJIS text strings at the end of each object file. There is a whole load of free memory available at the end of the program to store any text overruns, but because of the way that the extraction/insertion has been done, that memory is currently unused. My current, and possibly inaccuate, understanding is that Filler doesn't really enjoy the dull grind of game-testing, and so the project is a bit rudderless, and it would probably benefit from someone who cares about the game, and who preferably has some technical skills, coming in to finish it off.
|
|