Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
And even if you did save frames in those levels compared to the current publication, the fact of the matter here is you're using a level select feature that's available to you at the beginning of the game (i.e. doesn't need to be unlocked) and that breaks a rule that sadly appears haven't been clarified enough.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
The most recent example of the rule being violated, albeit retroactively, is [1990] NES Zanac by adelikat in 03:16.11. When adelikat made his most recent publication ([4866] NES Zanac by adelikat in 34:00.99), it was decided that it would obsolete the previous one instead of being put in its own branch as the previous one broke the in-game code rule. There is precedent in this rule being enforced.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
I was able to manipulate which music played before each stage by pressing random buttons before the stage loaded. The RNG also affected gameplay stuff like the position of the spiked falling poles in the pyramid levels.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
I think the music played in each level is affected by the RNG. Mario 3: Around the World is made by the same developers and has the same music RNG mechanic and I put that to good use in my run.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
In the most ideal case, a clean, unmodified SWF is preferable. In my opinion, I would prefer a modded SWF that ensures 100% determinism without any special instructions.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
PJ wrote:
I am not sure if it works the same way in this game, and I didn't notice anywhere where it would save time, but in Mario 4 you can kind of "store" P-speed by releasing the run button. You lose P-speed if you bump into a wall, but not if you stop running, as long as you are still moving forward.
It works the same way in Mario 3. It was done a number of times to reach platforms where jumping at full speed would make Mario run into a wall and lose P-speed.
PJ wrote:
This is probably not the place to ask, but I'm curious about what emulator issues are preventing Mario 4 from being TASed. When I was practicing, I noticed that the worse emulators (Gens, etc) had a final boss fight whereas Bizhawk just reset when you completed the final platforming level. I assumed it was an emulator bug, but playing on an "authentic" physical copy, it behaves the same way as Bizhawk. The game resets when you complete the final stage and you don't fight the final boss. Probably a tricky ruling to make...
I was thinking it's not emulator issues but actually the consequence of playing the game on a more accurate emulator. Even if Mario 4 was allowed to be TASed on Gens, I don't think I would do it as it lacks something like TAStudio and the traditional use of multiple save states and frame advance isn't something I'm used to.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
HappyLee wrote:
what if someone else makes a even faster version "Hyperfast Mario Bros." one day? Technically I think if someone improves the code further, it can make Mario move faster and further reduce lags and waiting time. Will that obsolete this run?
Probably, probably not. Depends on how accessible the hack is. Also, I don't think it's possible to go any faster than this without it crashing in five seconds.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
FCEUX's TAS Editor doesn't appear to have columns where you can insert coins or do resets. However, you can do it through the traditional way of frame advance and save states outside of TAS Editor, save the movie as an FM2, then import that FM2 into TAS Editor if you want to.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
There are actually powerups in Worlds A - D. The only problem is that a separate run has to be done with those worlds since you have to beat the game 8 times in a row to unlock them, and I'm not sure if it's worth it since there aren't that many powerups in Worlds A - D. If you are interested in doing this, consider using this verification movie as a starting point. It will generate the necessary Save RAM needed to unlock Worlds A - D.
https://tasvideos.org/UserFiles/Info/638008645247617084
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
To summarize your efforts, you recreated the game AI in Rust, created a scriptable working Atari 2600 emulator in Rust, ported a modern Othello playing engine as a script for the emulator, and pitted the modern engine against the primitive game AI thousands of times to get the best result. And yet you still managed to save some frames. I do applaud your efforts for a game like this.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
It looks like the submitted movie doesn't actually reach the credits like the currently published 1 player run. This one does, though I'm not sure if movie will have to be replaced.
https://tasvideos.org/UserFiles/Info/638048456560262145
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
[1926] NES Contra "pacifist" by Soig & zyr2288 in 09:29.08
After noticing that previous pacifist runs had only one player, I was thinking that we should let this one be part of its own branch. It's probably different enough from the main 2 player pacifist runs.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
aiqiyou wrote:
feos wrote:
The difference is you can't go left here until 1 frame later, like there's an invisible block or a collision with the block at the bottom of the screen.
In the NROM version you can go to the left so you can turn right faster, gaining speed and falling to the lower platform earlier. Also in the NROM version you resurrect after death faster. The rest is level loading time difference.
Fine. Our TAS can be well synchronized on the new Rom. in the next step, we only need to carefully check the details.
Ah, I thought this TAS would be resynced on the NROM version, but that doesn't appear to be the case.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
Impressive. Now the current TAS will no longer be nothing but bodyslams and time outs. It's now full of actual pro wrestling, making it more entertaining. I'm gonna miss the YouTube comments on how actual pro wrestling viewers and audiences would riot if they saw nothing but timeouts though.
Editor, Experienced Forum User, Publisher, Player
(52)
Joined: 10/15/2021
Posts: 383
warmCabin wrote:
This is the TASEncoding package everyone uses these days, correct? Would you want me to upload the X resolutions output by it or do you just need a logo + disclaimer on the 4K one for YouTube?
Though I would download from the latest master. I'll ask feos to make a tagged release out of that and call it 3.0.
In the NROM version you can go to the left so you can turn right faster, gaining speed and falling to the lower platform earlier. Also in the NROM version you resurrect after death faster. The rest is level loading time difference.
