News for hackers. Stuff that matters.
Re:The Super Mario Maker community faces its final boss by TillE on Saturday March 23, 2024 @06:38PM (#39803532)
That was always by far the most likely explanation, it was the general opinion of top-level SMM players. Something like 10 frame-perfect inputs with a lot of precise movement, it's just not plausible.

He's not really clear about how exactly he did it, but it's pretty easy to just take a controller cable, wire it into a microcontroller, and program inputs. A lot of people were saying "there was no TAS in 2017", meaning I suppose there were no specific tools for emulator at the time, but that's just irrelevant when you can take a day or two to trial-and-error it.

Re:The Super Mario Maker community faces its final boss by kevinventullo on Saturday March 23, 2024 @07:13PM (#39803699)
I mean tasvideos.org goes back to 2003: https://tasvideos.org/SiteHistory#2003
Re:The Super Mario Maker community faces its final boss by LukeShu on Saturday March 23, 2024 @08:12PM (#39804000)
I don't think anyone is saying that there was no TAS in general in 2017, but that there was no TAS for the Wii U in 2017.
Re:The Super Mario Maker community faces its final boss by bakugo on Saturday March 23, 2024 @07:15PM (#39803712)
I'm fairly sure the Wii U was already hacked by 2017, so he could've just played the inputs back on the console itself. Either that or just hack the game into thinking he cleared it when he didn't.
Re:The Super Mario Maker community faces its final boss by pjerem on Sunday March 24, 2024 @04:14AM (#39805775)
No need to hack anything, you just have to make a Bluetooth controller with the same characteristics than Nintendo’s controllers. They are pretty standard.
Re:The Super Mario Maker community faces its final boss by Retr0id on Saturday March 23, 2024 @08:41PM (#39804151)
Even with a microcontroller, doing frame-perfect inputs is tricky because you don't have anything to sync off of - even tapping into the vsync signal from the video output is an imperfect reference because there's no guarantee that the game engine is actually synced with it (although I have no knowledge of SMM, maybe it does work better than I expect)
Re:The Super Mario Maker community faces its final boss by a_t48 on Saturday March 23, 2024 @09:59PM (#39804528)
You can start off with a start input to unpause the game, for instance.
Re:The Super Mario Maker community faces its final boss by Retr0id on Saturday March 23, 2024 @10:13PM (#39804606)
That doesn't solve the problem, in the general case (it might work acceptably well in practice, though)
Re:The Super Mario Maker community faces its final boss by ikiris on Sunday March 24, 2024 @02:59AM (#39805537)
if your controller has such loose timing that you can't match a 10ms window reliably, you're doing something very wrong. '
Re:The Super Mario Maker community faces its final boss by lmm on Sunday March 24, 2024 @03:34AM (#39805641)
Match a 10ms window how far away from your last sync point? Every clock has a drift rate.
Re:The Super Mario Maker community faces its final boss by themoonisachees on Sunday March 24, 2024 @06:32AM (#39806255)
I get where you're going, but also consider the level is 17 seconds long. It's not unreasonable to think he just ran it a few times until the clocks decided to cooperate.
Re:The Super Mario Maker community faces its final boss by braiamp on Sunday March 24, 2024 @12:11AM (#39805117)
Except that many top players said it was very unlikely due the time it was uploaded and the availability of TAS methods at the time. The creator took a day to beat it, because not only couldn't sync inputs with the game engine current state, but also there's real rng in the trick.
Re:The Super Mario Maker community faces its final boss by ufferop on Sunday March 24, 2024 @01:51AM (#39805372)
The funny thing is if you watch the clear vid it actually looks really straightforward.

The original purpose of the video was to coyly announce the TAS tool the maker created, but the video was mostly ignored because it looks like a regular clear vid.

That’s how it so effectively accidentally flew under the radar. The level lacked a TAS label, and TASs are usually obvious and flashy.

But then you play it and realize those jumps are insanely precise.

Yes, frame perfect, but also mostly pixel perfect x,y precision.

There is even RNG manipulation necessary to ensure a bomb is spawned from a note block in a specific direction.

Streamers had already beaten it in chunks so it was theoretically possible for a human to do, but once people started offering monetary bounties the uploader felt it was time to let everyone in on the 7 year old joke.

Re:The Super Mario Maker community faces its final boss by ufferop on Sunday March 24, 2024 @02:21AM (#39805438)
> He's not really clear about how exactly he did it, but it's pretty easy to just take a controller cable, wire it into a microcontroller, and program inputs.

The streamer featured in TFA, ThaBeast721, said that he and some other high profile streamers playing the level were contacted by the maker, whom many in the community knew prior and so there was chat history to authenticate the claim, and shown a video of the device breaking the level.

Thab described the device as a “raspberry pi or arduino type device hooked up to an actual controller somehow”

When the clear video failed to go viral the maker dropped the gag, and forgot about it until recently, because the device was a pain to use anyway.

The maker said it took 3 months to upload the level. Something about the device causing the console to crash when programming the inputs.

https://m.twitch.tv/thabeast721