byuu's message board

Ah ok, let's start to rewrite all Snes games to suit other plataforms.



So your "big ideia" is to change all the 65816 opcodes in a Rom to x86 equivalents, so the Rom can be used like an .exe and run directly?

Are you referring to AnusPush?

No, I'm n....teeheehee XD

This is really quite simple. All Sora needs is to enroll in a master's degree program in Computer Science from a well-regarded accredited university with a strong emphasis on the theory of computing, number theory, and algorithms.

Then Sora will see first hand just how fun it is to mathematically describe Turing machines and algorithms and can move past the Wikipedia level education that's been demonstrated by Sora's posts.

Sora's got these great new ideas though, so it shouldn't be long before we see a paper on how they've solved P=NP.

You don't need to have a degree in computer science to understand the basic principles of P=NP or a Turing machine. Just some patience to get through all the set theory symbols a Turing Machine throws at you . I'd dare say you don't even need a computer science background to appreciate the limits of computation (I've also seen a decent P=NP explanation for the laymen before).

I guess it's safe to say that our friend from Kenya got bored. This thread was thread was ultra-epic.

Lost net

Once SNES opcodes are translated to suit opcodes of another hardware (with shortcuts where applicable), within an emulator framework the ROM can be automatically patched to make it an EXE with opcode/offset changes.
My point is from a cycle accurate algorithm a shortcutted/step reduced algorithm which will retain accuracy and break the cycle accurate cycle/hardware state data can be done to have a functional emulator using the least steps to maintain accuracy as well as be specific to another hardware's opcodes.

That can then be patched automatically into any ROM with some extra data to adjust offset information and also have additional emulator code for controls, display etc embedded (as well as trim padding data).


For shorcutting/step reducing the original cycle accurate algorithm, one can keep in mind that the expected result is matching numbers for output/temp and is done using math.


From there, individual tweaks is generally something else

PEOPLE HAVE DONE THIS BEFORE
PEOPLE HAVE BEEN DOING THIS FOR DECADES
PEOPLE HAVE BEATEN IT TO DEATH OVER THEIR ENTIRE CAREER

I don't rightly have any response to this other than the oft-repeated "you have no idea what you're talking about"

Here we go again...

Seriously though. SoraK05, what do you have to show us that is not just more text? A practical demonstration would be very welcome.

Skipping architectural and clock cycle accuracy impossibilities, it'd be a bad idea to vanilla compile port everything anyhow. Many ROMs busy wait, so they'd tie up a core on a modern system, while an emulator would throttle said clocking to that of the arch the ROM was made for.