If the game runs and you can play it relatively glitch free, you're happy. For many many people, likely yourself included, accuracy doesn't matter. That is what 'accurate' emulation means, you would get the same response through the emulator as you would on the actual machine. In this case, it means a game might not be running the exact same as it did on the original hardware. A game running is not always a sign it is being run accurately. The main point is 'all details' and 'exact'. Well first you should look at the definition of the word, if you aren't clear on it. The argument being made is that lr-snes9x is the more 'accurate' emulator, but what does that entail? Removed CPUShutdown speedhack, DisableHDMA and Remove POSIX dup and access calls, and rename qword toįixed crash relating to double-closed descriptor. (windows port, unix port and gtk legacy config) (Jonas Quinn) (gocha)įixed BlockInvalidVRAMAccess config file option. (zones, OV2)įixed movie snapshot unfreeze inconsistency. (OV2)įixed the problem that echo buffer breaks IPL ROM.

New APU timing hack, fixes various games that exhibit (BearOso, zones, OV2)įixed C4 for Mega Man X2's 'weapon get' screen. Improved screen interlace and sprite interlace supports. Improved overall timings, now Snes9x can handle events in Unix: Added Xv support and fixed several bugs.

(OV2, libretro team)įixed MMC bank register bit 7, restored 64mbit ExLoRom Speed hacks, but also regresses a few speed-hack games.

This has the effect of increased accuracy, fewer (LuigiBlood)Ĭhanged the S-SMP core module to one written by byuu. (qwertymodo)Īdded support for BS-X Satellaview memory packs and data (LuigiBlood) Here's the list of what we're up to so far:Īdded support for the MSU-1 coprocessor. One can simply view the current snes9x changelog ( ) to see what has been added/improved.

Snes9x 2010