Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
the tas looks... fine i guess but there is obviously a lot of places where you should have done tetrises, so this tas is clearly not super optimal
voting meh
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
If we use the built in "start from SRAM" option on 0.9.11, it just crashes the emulator and we don't know of any fix possible, so that was the only way to use sram without the emulator crashing.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
Gonquai wrote:
Last one to include frame boosting is generally a good cutoff (though that comes with its own flaws, since smba does technically have it, though it is a hybrid of sorts).
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
My argument was that if there was only a "fastest completion" on Multi, getting the true ending is suboptimal since it gets an additionnal fight. Of course, if we have both a "fastest completion" and "best ending", that problem is gone.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
The amount of modes available from the start is also stunning. Technically we allow each mode to be published separately, but I don't particularly like the idea of having 20 branches for Meteos, per mode and per ending (whatever the actual number would be).
Simple mode is just some CPU battles with the ability to change some rules and planets.
It is clearly uneligible for a branch.
Deluge is a stamina mode, where the game ends when you fail. A "max out" (9 999 999) is theorically possible on every planet, but I feel like only one branch would be acceptable : "fastest max out". However we would need a verification movie for the unlocked planet if it is not a starter planet.
Star Trip has 3 different modes, Straight, Branch and Multi.
Straight is a series of 1v1 battles versus random planets with the exception of the player's planet. RNG manipulation would be important in this mode since every planet could be fought. It can be its own branch.
Branch is like Straight but planets are pre-determined with their position, so some route can be faster than another one. It can also have it's own branch for fastest completion, regardless of the ending.
Multi is a bit special in my opinion. Reaching the The Ascent ending (the true ending) unlocks Layazero so it is a special ending, the other endings not unlocking it. Furthermire, to reach True Meteo, you have to fight an additionnal fight compared to the other two endings, so if we have a single "fastest completion" branch of the Multi route, this movie is suboptimal for reaching the true ending.
So for Multi we can have 2 branches : "Any Route" and "True Ending".
And for Time War, there are 4 different modes, a 2-minute score attack, a 5-minute score attack, a 100 meteo sprint and a 1000 meteo sprint. Since they are all on different planets, maybe a TAS that clears all 4 with the highest score/fastest IGT could be accepted, but I'm not completely sure.
Extended input that's not present in the published movie but only in the encode is only allowed if this input is trivial for a publisher to execute. Since in this case it's not trivial, and also it increases the overall entertainment, having it in the primary movie makes the most sense to me. Arcade Tetris also has this.
Yeah that's fair, since it's not just an additional input to clear text boxes but a complete superplay. It's ok to have the movie file replaced by the one with the superplay.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
it's got to take at least three frames to move a meteo (tap old location, tap new location, release)
if you need to move a meteo for more than one space, you need to hold the new location for more time, one frame by space
also sorry for the inactivity, i have got a lot of work to do recently, i do have a credits tas done, but i really don't think it is interesting enough and i don't know what exactly to do to maximise entertainment.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
As for the credits sequence I think honestly it'd be nice to actually have extended input for publication just for screwing around and adding something more instead of just watching the credits roll.
Do you want me to TAS the whole thing and post the file here afterward? Like, yeah I understand it could be a bit interesting to see actual TAS superplay since the whole run is just sending garbage to every planet as fast as possible and not ultra long combo (plus the Metamo Ark's playfield is super high, making a lot of possibilities). If it helps with the submission I can do it.
Would I then need to cancel this submission and make another one after or will you be able to just change the file here?
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
Just the high score (compared to everything else) on the last song shows that this game almost don't forgive mistakes on Hard Rock! difficulty. Normally TASes are made with perfection in mind, but this TAS proves that in the right circumstances, failure is entertaining in a TAS. For a rhythm game TAS, this is a very original approach and I like it.
Yes vote
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
I'm pretty sure there is no credit sequence if you beat all modes. Even if there were, I don't know if you would need to beat all bombliss, shadow and stage racer levels.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
I can see you run between jumps, I'm sorry but I'm pretty sure the best way to conserve speed is getting it as high as possible (it ossillates) and frame perfect jumps.
You can butt slide to get a bunch of speed if you can't get P-Speed, which you don't do in level 1.
There are questionable slowdowns.
Probably not an acceptable hack.
No vote.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
I found the monkey choice (I just hate baby, gongon best monkey) and the goal of real time over per level optimisation to be a questionable one
baby loads a lot faster than other monkeys, even on real hardware, and optimisation for real time is mostly because only the camera speeds up, while every other SMB game where the stage also speeds up, making some IL strats impossible/slower. but all that was already in the submission text so you should know that already.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
upon further testing, the bad movement might be due to the game being 30 fps, and the author might not be completely comfortable with this, and the low rerecord count might mean there is suboptimal play overall
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
this movie doesn't sync for me, what settings do you have?
(actually nevermind, just realised i got a PAL version of the game, which doesn't have the first message at the beginning and might have different rng, while this movie is on the US version)
also gave it a small attempt, moving pieces in this game is fairly slow actually, so while it seems unoptimal, it's mostly because of the game. might give it a better attempt later
oh yeah, also this movie starts from sram
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
I reaaaally don't like you going to visit Sans, it wastes time and is not fun to look at. Since it's at the end of the run, you could probably take that out of the TAS.
Other than that, the run looks really nice. Yes vote
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
I started at level 0 by accident
I mean....
couldn't you have taken like a minute or two to fix that? shouldn't have kept that, save states are basically there to fix mistakes.
Also the fact you couldn't have a clean stack on level 0 is reject worthy.
No
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
IL WRs vs this TAS (for comparison) https://www.cyberscore.me.uk/game/1029
1- Snow & Go
WR : 34.58
This TAS : 35.53 (+0.95)
2- Jigsaw Jumble
Inaccurate since this is full screen and optimally you'd want less space for pieces to fall faster, but this is time trial so we can't do that.
3- Peak Precision
WR : 6.11
This TAS : 5.95 (-0.16) (nice 1 frame improvement vs previous TAS)
4- Speeding Bullet
WR : 26.54
This TAS : 26.54 (-0.00)
5- Chain Event
WR : 20.80
This TAS : 20.81 (+0.01)
6- Ballistic Beach
WR : 10.53
This TAS : 11.15 (+0.62)
7- Upward Mobility
WR : 21.18
This TAS : 21.18 (-0.00)
8- Pizza Me, Mario
WR : 32.56
This TAS : 35.98 (+3.42)
9- Flinger Painting
WR : 12.41
This TAS : 13.28 (+0.87)
10- Pit or Platter
WR : 22.41
This TAS : 16.91 (-5.50)
Full run comparison
WR : 2:56.95
This TAS : 2:47.84 (-9.11)
Optimisation looks overall good, but you lose time on multiple mini-games over their WR.
Also would going in a circle during Pit or Platter save time, if you can manipulate it to go like that (actually nevermind, looks like it's a fixed pattern)?
Voting meh.
Experienced Forum User, Published Author, Player
(62)
Joined: 12/4/2017
Posts: 66
Voting yes for the nice usage of the l+r glitch, I feel like a more complete run would be much better for this game though, since that glitch opens more routes and speeds up water levels by a ton.
But we'll see when one is done. This run is nice and different enough compared to the NES version.
