Home Page Forums MAME MAME The SNES WIP topic

Yes, on byuu's board it is said 1A and 1 are the same, but why does higan use 1b rom for 1a games (actually, game) instead of 1, then?

1B is a bugfix of 1A, and byuu apparently cares less about correctness then he used to. Or it's a typo in higan. Take your pick.

Or http://en.wikipedia.org/wiki/List_of_Super_NES_enhancement_chips is wrong about DSP-1A in Ace o Nerae! 3D Tennis.

and who has ever trusted wikipedia? we have in the softlist acenerae as a DSP1A game, but I doubt ReadOnly marked it based on that page...

anyway, since he might have made a mistake, do you know which DSP was there? 1? 1A? 1B?
I think Ace o Nerae has still to be opened up to check which variant was inside, so maybe byuu has decided for the 1B for the time being (Super 3D Baseball, Lock On, Super Air Diver and Suzuka 8 Hours have been verified using the 1A on the other hand and I think higan uses the correct variant like we do for these)

DSP1 was first
DSP1a is a die-shrink of dsp1, same exact code as dsp1
DSP1b has bugfixed code (but still has plenty of bugs)

LN

Asked him, he said it was a guess So it really needs to be opened.

I didn't know there are different version of those.

One of the things I was not happy with, in the slot code used by snesnew, was that some games which should have been partially working with the new code, were instead completely messed up. For instance, trying to start a game in bsxsore after inserting your name resulted in complete garbage



despite most basic accesses being emulated as per available knowledge (kudos to byuu for his invaluable research), and my implementation being a bit rough but pretty much correct even after several reviews...

Long story short, it turned out that the problem was not in the new code, but in a missing mirror in cpu regs (impossible to detect in snes/snespal drivers due to the way old memory handlers were already mirroring each other)
Adding the missing mirror helped to fix most of the problems in snesnew (missing gfx effects, missing gfx layers), and finally allowed bsxsore to show the expected city





This definitely gives better foundations to future attempts in emulating the bsx memory packs!

I still need to sort out a few issues in snesnew before starting merging slots in the old code, but it's a nice improvement nevertheless

and while at it, I have added support for DSP1 games larger than 1MB... this makes Super 3D Baseball aka Super Bases Loaded II aka Hanguk Pro Yagu finally working

ok. now snesnew is officially supported and tests are welcome, as long as you don't report broken bsx memory pack emulation (which is known to be defective, as soon as a memory pack has to be loaded to the flash memory).

OTOH, I'd be very interested to know if there is anything which works in snes and does not in snesnew, or any regression compared to 0.148 (or 0.148u1).

Games broken both in snes and snesnew are slightly less important to report because our driver is still quite behind higan (e.g. most Human titles which do not like current cycle steal code in the CPU, or pilotwings, or tg3k glitches), but given that we have lost the wiki page containing the snes bug list feel free to report these as well