A Moogle  🏳️‍🌈 💻 🎮 🎶 🇸🇪

@plgDavid @TheMogMiner I need to look more carefully and thinking I may need to disassemble some of the A'can games to understand better but… are we sure that there's a sound coprocessor at all? (are games loading arbitrary blobs into some RAM or something?)

@plgDavid @TheMogMiner Just saying, going by the reported addresses they look like they may be actual ports, and if these are being done with 16-bit accesses then it's unlikely there's an 8-bit coprocessor in it (due to bus size), and it seems to be wavetable-based…

Additionally, the 68k uploads perfectly valid 6502 code to the shared 64kbyte space. In my local code, the 6502 is now polling inputs and handing them off to the 68k. Further, it's clearly processing what appears to be some kind of note data, and poking SPU registers.

I'm in the process of manually converting the player code to C-like pseudocode to get better insight into exactly how the player program expects the SPU to function.

In general, just saying, many of your tweets at @plgDavid and I make it sound like you're falling into the same trap that so many others have, which is assuming the MAME team are a bunch of idiots who somehow YOLO'd their way into emulating thousands of systems.

Being able to observe valid 6502 machine code being uploaded to a shared port is so incredibly obvious that it's honestly a little insulting that you would just assume that we're so stupid as to think that there's a 6502 there if there isn't actually one.

The problem is that there's so much speculation around and the source code isn't saying much what's *confirmed*. And I mean, on the Mega Drive side we used to think the SVP was a SH-1 and were proven otherwise only the moment it got emulated, so I tend to be sceptical about this.