SwedenHey, I updated the rules a little (added some more allowed emulators from the SMB1 leaderboard) and added patches to the resources section. I haven't managed to get a hold of the patch for the version used by the top 2 times but I believe the 'X1' patch works mostly the same just looks a little weird, or you can use the 'with warpzones' patch which I think also applies to a normal smb1 rom.
You can also use the newer anniversary version if you'd like. Not sure if we're going to separate it from the other patches.
sorry for the necro, but just to clarify that feature doesn't actually "shave frames" off anything, the run time at the end would be exactly the same length, and it also in no way alters any mechanics. :)
the potential advantage runahead could have is if you need to react on the fly to something very quickly, then it could possibly give you a frame or two of increased reaction time. the feature shouldn't be allowed, but i'm skeptical that it would be a good way to cheat in any game.
Well you have to jump for a specific number of frames off the elevator when it's at the correct height.. There are a few spots and number of jump frames that work, but, not very many. You can see an example here: .. see the inputs on the right, and the exact jump spot he used.
Anyway I think this is probably a bit far off topic from whether there should be a two player category. :)
These inaccuracies could be fixed in emulators and flash carts, I'd say it's probably a better approach to file issues with these projects and try to convince them rather than try to convince every game to separate emulator runs.. I know the Mesen dev has been working on it.
Again, great work Nihilist!
RAM is more variable for sure, most consoles seemed to have one 'main' starting state that they hit more often, maybe up to 50% consistency depending on the hardware. I think all seeds should be be accessible on all consoles, there are only about 1000 possible question seeds and the way the rng advances is pretty simple.
The difficulty mapping out seeds on hardware is that flash carts also initialize ram and ppu, so they are effectively the same as emulator when it comes to this. To get the actual seeds for an original cart you'd need to come up with some clever game genie codes or something similar to find your starting state. Seems doable, but more work than emulator at the moment since no one has made the codes.
I personally don't really mind either way, seems a little silly to separate runs more considering how few there are in total.. but if they were separated I think the board with more players should be the main category. At present that would be the board for non-original hardware, which would have 11 vs i believe 2 (me and chambers) for original hardware.
Thanks for coming to a decision Nihilist! I think that's a reasonable solution, even if I think the categories are a little silly since they don't really change the game in any notable way.. Unless someone comes up with new strats for 20k 1P2C.
Chambers does have a fair point regarding emulators in that it is considerably easier to map out and target Family Feud seeds on emulator because of how the RNG is implemented, compared to NES + original cart.
But since this is an issue with a number of NES games, it feels like there'd need to be a general consensus view in the wider community to take any action.
To clarify in case anyone is unaware, Family Feud's RNG used for the initial question seed is affected by initial ram state (it uses uninitialized memory) and i believe also the PPU/CPU clock alignment. This is always the same in current emulators, but fluctuates on NES.
I personally don't mind either way.
It kind of invalidates the current runs based on just what was assumed to be the rules for the category, but there aren't that many runs on the board so I don't know if that matters.
Edit: I did just try two family mode and I will definitely give SpaceColonizer that is more fun to play that way, the reduced downtime makes it a bit more engaging.
Nice! Didn't want to give it away, since I saw it in the code.. Not as fun as figuring it out on your own.
Jump from a specific spot on the elevator, for an exact number of frames (don't remember how many, but you can check the video) and you'll fall right through the floor.
Watch the FPGs in this and you can see the inputs on the right. Just land on the correct spot before the flag while not holding B and jump very close to the block.
Hm what specifically is disallowed, pressing the left and right buttons simultaneously on keyboard? Just to clarify for dummies like me. :)
If I'm not mistaken, that cart was only released in PAL regions, so it's likely that you are running a PAL version of SMB.. The 8-2 issue you are describing happens in that version, because the movement speed of mario is sped up to compensate for the reduced framerate of the console. Any other version, such as the Nintendo Switch release, would be using an NTSC mario rom which makes the 8-2 stairs a lot easier.
FDS PRG data is loaded in at $6000-$DFFF instead of $8000-$FFFF, even if the actual code were as identical as it could be, that would be enough to change minus world since everything is moved into different locations.
You won't be able to diff the files since FDS is a whole different format, but the "US/Japan" rom you're looking at is likely the cartridge version. The FDS version would have an .fds extension, and not have "US" in the name.
If you want to actually compare them, you could start both versions of the game in an emulator with a decent disassembly view, copy out the disassembly and diff those, then account for the differences caused by the different load location vs actual code changes.
Try selecting all the frames (ctrl+a), right clicking and pressing 'ungreenzone', or go into File > Save Compact, save the file and open the new file. If it's because you've edited memory or the rom file in some way that should clear it up.
The coin areas are reused per level, the exit pipes are loaded based on the level you're in. The only place where you would be able to save time that I'm aware of is in 1-1, where the unused coin areas send you to the end of 1-2. In all the other levels I believe they send you to the start of the current level.
But yeah, wouldn't expect anything like that to be found.
It does remove lag which should definitely not be allowed, which seems to put Mesen in a similar position to RetroArch which afaik is disallowed primarily because it has this setting..
edit: Mesen has been banned!
