BC UR Flap

[earnest light]

Moving all of the progress for this run into a thread, hopefully this will encourage people to finish this run. Right now we are trying to complete the 1 shroom ultra with Daisy instead of Yoshi. With the current setup (done by shxd), we are only a little EV short of making ultra. According to Monster, we need to have at least 176 EV on the frame we hit the tree, we have 2 ghosts, one with 175 EV and a bad angle, the other with 169 EV and a decent angle. It is essential to avoid any geysers to ensure we can delay the inputs from an SSMT to get good cycles. Important to note that Monster only has 161 EV before hitting the tree on his successful attempt and we have around 163 EV, so different QM on the tree clip might be all we need to get enough EV.

Successful ultra with Daisy on lap 1
#tas-files message
https://www.youtube.com/watch?v=Ut-aONFKjX8

Ghost with 169 EV and good angle
#tas-files message

Ghost with 175 EV but bad angle
#tas-files message

Ultra landing strat (Yoshi)
#tas-focused message

One shroom ending cut (Yoshi)
#tas-focused message

[dull ridge]

i will try to work on both 175 and 169 ghost this weekend i think

[dull ridge]

@forest walrus deperately need you

ive tried for the last couple hours

sorry for ping but if you just want to take a look at it be free to

[dull ridge]

excluding the days of the other times ive worked on this over the last few months

[earnest light]

At what point do we dm Justin

[dull ridge]

idk

[forest walrus]

I'll give it a try, but don't expect much

[forest walrus]

is it better to not touch any of geysers so that we can adjust the cycle for flap?

[earnest light]

yes

[forest walrus]

I see 👍

[dull ridge]

yeah i think thats the main reason i cant do it

[forest walrus]

closer attempt over this https://discordapp.com/channels/214574342137315328/600410593937195069/1284978788471410749 but ghost doesn't sync if you change ssmt release timing

[earnest light]

hopefully if we get it we can reference the inputs for 3lap

[dull ridge]

you dont need to not hit geysers in the 3lap

[dull ridge]

how many EV did you get?

yeah the thing Is that without hitting any geysers its hella hard to get the required EV, if you go on geysers you can get a little boost on the slope of it

and perhaps also a better angle

[dull ridge]

also how does It not sync? like its an ssmt, It should sync everytime

oh did you like release It so soon that you still had the momentum from turning

[small scroll]

values still fluctuate during an ssmt. in general flaps don't sync if you change the frame when you release ssmt

[forest walrus]

^

I released the ssmt at the same time as you did

sometimes ghost desyncs at sg entry, sometimes at sg ejection

[dull ridge]

oh i see

[earnest light]

i wonder if its worth finding a cycle where we go over a geyser while its low to make the ultra while still having a good cycle at the end

it would be hard but maybe possible

[dull ridge]

we would need to like time everything and if It doesnt match we're cooked

[earnest light]

thats why i dont know if its worth it

[earnest light]

this ghost gets enough EV, we just need to have a good angle now

gonna have a little look at it now

[dull ridge]

alr

🤞

[earnest light]

we might have to change the landing strat based on the landing

[dull ridge]

you think we arent gonna land on the boost panel?

[earnest light]

this is where we currently are

[dull ridge]

oh wow that close

this is monster's effect chat

it shouldnt be too hard to land on the boost panel

[earnest light]

slightly closer attempt to Monsters ghost

also this gets more height but less distance

[dull ridge]

finally someone that sends csv

⠀

[earnest light]

thats only because i didnt want to overwrite the ghost :P

[tranquil venture]

you need to be a lot more high up and also inside this wall

[earnest light]

i know

this is the csv

[tranquil venture]

why does that look much worse than this attempt

[earnest light]

good point, here is the csv for that attempt

[dull ridge]

i got 181ev

probably wont make a difference but

[earnest light]

monsters ghost gets 182 xyz EV and 176 xz EV, i probably shouldve made that clearer

[dull ridge]

oh, i only saw the 176 then

[forest walrus]

This may not be essential, but adjusting the speed to around 120 on the frame you hit the block will make it easier to keep EV a little while still getting the height

[dull ridge]

lirwm (or you idk) gets 83xyz and 78xz

[earnest light]

hopefully if you can fix that we can finally move on with this flap

[dull ridge]

i damn hope so

[tranquil venture]

i'm a little confused

why do we need to avoid the geyser for cycle changes

the yoshi flap already gets a good cycle

just start the flap at the same time as the yoshi one, but using daisy instead

[dull ridge]

well both the sg and flap might not take the exact time as the yoshi one 🤷

[tranquil venture]

even if daisy is faster by like 0.4 somehow, the geysers are still relatively the same

they move slowly

and we can always just prevent that by starting the lap ~0.3 later than the yoshi one

[earnest light]

i think we got pretty lucky with the yoshi cycles

the daisy run will start the lap about 4.3 seconds before the yoshi run

[tranquil venture]

aim to hit this tree with daisy (don't worry about avoiding geysers) at around 1:06.0

boom, there's your flap setup

[earnest light]

ill have a look at delaying the flap tomorrow

[dull ridge]

but should we really do that? like we are almost close now

[earnest light]

also idk how useful it would be but if we are able to copy inputs for the 3lap run, avoiding the geysers would make landing the ultra again easier but thats a bit optimistic

[tranquil venture]

I don't get it, what's the alternative if we don't do that?

Do a flap with Daisy that gets a bad geyser cycle and it ends up slower than the WIP Yoshi one?

And we can't delay inputs with TTK anyway because of how QM-heavy the ultra is

[tranquil venture]

we won't be able to reuse inputs for 3lap because the 3lap setup approaches from a different angle, due to recovery from a strat we'll use to skip the last turn

[earnest light]

ill see what i can do tomorrow

monster said the run said the run sometimes desyncs at the sg ejection, if we can get that far with the good cycle timing hopefully it shouldnt be too hard to get a good clip

[dull ridge]

i cant get anything by the way

[dull ridge]

with this

[earnest light]

synced the clip later

might be too late though

like .6 later than the yoshi flap

[dull ridge]

of what ghost

like the clip of what ghost

yours right?

[earnest light]

i think it was the most recent csv i sent

[dull ridge]

so now we can hit geysers

according to ebay chan

⠀

[earnest light]

we can definitely hit the first geyser

we might miss the last geyser

this is roughly where we will be at the geysers

the cycle in the yoshi flap is pretty tight because it barely gets the first geyser trick as it gets bigger and barely gets the final trick as the geyser gets smaller

this should be .1 sooner but misses the tree clip

This is the best one so far

clip syncs and hits the tree at 1:06.090

in case it isnt obvious im going through every frame to get as close as possible to the yoshi cycle

same as previous, except a frame sooner

i will send a rkg of the best one

clips tree at 1:06.006

nice

[dull ridge]

will this even gonna end up useful, if we go on the geysers 1 the time would be different by a bit once you hit the tree anyway 2 if the end doesnt have good geysers (hence daisy or the flap ends up being faster) we are cooked

[earnest light]

i think even if the geysers are bad there are still frames where releasing the ssmt at a different time the run will still (mostly) sync

[dull ridge]

not if you hit the geysers on the setup, unless theres some cycle where It syncs both for the setup and the end

but It would be a pain to get the exact time the geysers move on the setup if It ends up bad

idk man im not convinced about this

[earnest light]

in my run i dont hit any geysers

[dull ridge]

ik, but the point of this Is to make the setup hit the geysers so its easier right?

or am i misunderstandinf

[earnest light]

we dont need to hit the geysers because we already have enough EV rn

if i dont get any progress to making the ultra with the current setup then we can do the other setup

[dull ridge]

i genuinely cant make this btw

we might need to call for backup again (hes the only one that can actually make it 😭)

[earnest light]

i had a quick look at this and couldnt get anything

would really appreciate it if someone new had a look at this

[earnest light]

#1344592240164470806 message

making sure this is easy to find

[earnest light]

What We Want
between .98850~.98855 race completion,
~176 yaw
What We Have (#1319937435152416768 message)
.98864 race completion
173 yaw

[earnest light]

i tried for a little and got this attempt with 176.9 yaw and .98855 race completion but it has not enough EV and bad qm

@ Monster with more experience landing the ultra and the new conditions would it be possible for you to try this once more? I presume it would be easier for you but i dont want you to feel forced into it if you dont want to do it

[earnest light]

ive been able to get clips with the correct yaw and race completion but i lose too much EV (i get 173-174 EV)
because we start from a ssmt its probably worthwhile to release it earlier if we have to lose time to get good cycles later

[forest walrus]

how much does Daisy's speed bonus save over yoshi?

would she even beat the yoshi run if we get a slow ultra

https://youtu.be/Ut-aONFKjX8 https://youtu.be/H9EHd2ANh2o

seems daisy's slow ultra loses around 11 frames before the landing over yoshi

[modern trellis]

where does daisy gain time

I feel like most of the race is spent either at 120 with EV or at lower than top speed

[small scroll]

i don't really understand anything about this ultra. is the "slow ultra" because daisy has a worse offroad stat than yoshi? in that case, why was yoshi not considered for 3lap?

[forest walrus]

I don't completely know

It is sensitive and affected by minor things

[modern trellis]

is it something like yoshi gets more EV in the offroad? or does it gain in some other way

[earnest light]

Ask Ebay G

[dull ridge]

from my experience, building EV with daisy in offroad was harder

but that doesnt matter in the lap

[tranquil venture]

i didn't realise how slow the daisy ultra was compared to the yoshi one w/ reverse superhopping

maybe we should just stick with yoshi then...

[indigo creek]

If anyone needs it, this was the latest yoshi flap

[gentle kelp]

there is a better ghost than mine it's this one:

#tas-files message .601 by thomas

[indigo creek]

Oh oops I didn’t see that because I have Thomas blocked

[earnest light]

13.596

ending superhops can probably be improved more

[forest walrus]

the ending is cool

[pine rain]

#tas-files message
At that time, right after Thomas released the 13.601 data, I had the data that I improved, so I will send it here.

13.582

After that, there was a run using Daisy, so I thought there was no point in updating it, so I didn't publish it until now.

[forest walrus]

got an improvement on the shortcut with 2nd shroom

trying the ending shortcut with 1f early QM

[forest walrus]

13.557

[tranquil venture]

love the reo drift

well, maybe it's just a slipdrift i guess

but still

good work

[forest walrus]

landing position could be changed more to the right which would make the line you take mid air shorter

[forest walrus]

thanks

[pine rain]

After all, is Yoshi faster?

[tranquil venture]

Daisy would be faster for the driving after the ultra, but we get a super fast ultra with Yoshi, and cannot get the same ultra with Daisy

So, we chose to continue with Yoshi

It could be that Yoshi just achieves better ultras, for reasons we don't fully understand yet. Rather than it being an issue with difficulty/skill, it's likely to do with his character stats

[small scroll]

we know that yoshi's traction has no effect on ev generation, daisy and yoshi lean the same way. but it's unclear if it has an effect on ev dissipation

[forest walrus]

13.453, couldn't get better ending cut or landing

[indigo creek]

managed to get a -0.010 but i lost the savestate

ok better

13.440, better last 2 tricks + ending SH

[pine rain]

13.422

[narrow peak]

Is this useful for 3lap? I don‘t think it is sadly

[tranquil venture]

you need a mushroom for it

[narrow peak]

alright

[limpid ferry]

more IV leads to more EV gain in a SG and you start the SG in the offroad with a character that has an offroad bonus

[small scroll]

wait what how

[limpid ferry]

idk how to explain in technical terms and this is mostly just empirical but i think it has something to do with your vehicle having a more “stable” state in a SG with more IV. I know as well that you gain a LOT more EV when you shroom into a slope in a SG

But im 100% sure it has something to do with the difference in IV

[small scroll]

this game is bullshit

so basically the ev gain per frame is not a constant depending on idb/odb? i thought it was

[tranquil venture]

https://youtu.be/ixRo4WABTfY?feature=shared&t=39

from the description:
"....A few months ago however, I discovered that using three mushrooms, one before the launch for increased external velocity and...."

[limpid ferry]

It might just be something to do with the relation between slopes and IV which there are several of during the SG on BC but idk for sure

[limpid ferry]

Also was this testing done with SG specifically?

[tranquil venture]

you build more EV on downhill slopes because of gravity

which is why it's possible to supergrind downhill on a kart

but that still doesn't explain why shrooming increases EV gain

[forest walrus]

interesting. I think speed bonus gain an advantage with Superhopping too

when I tried using some characters (like Mario and Daisy) and TAS Coded the same ghost that does superhops, speed and handling bonus made decrease in EV on the ground a bit smaller

but not sure about shroomimg, in my experiance shroom boost doesn't help much in increasing EV on superhops

[tranquil venture]

in my experience shrooming can make superhopping worse

[small scroll]

superslide. i don't know it's crazy that it's all different

[small scroll]

this also seems crazy i wouldn't have guessed at all

[limpid ferry]

shroom vs. no shroom (same inputs)

might just have to do with the slopes in particular

anyone know of another SG with a flat surface?

i thought of LC but i remembered that the first straight wasn't completely flat

[tranquil venture]

the end of TF

moonview highway ending (supergrinding from before the boost panels, not at the top of the hill)

[small scroll]

am i stupid what changed here exactly

[tranquil venture]

nothing

that's the point

EV is identical, so shrooms mustn't matter on flat ground

[small scroll]

oh i thought i was supposed to look at the ghost

okay i getit

[main valve]

It's the case for the lean effect. But when you do a SG, you gain EV 2 different ways : The lean effect, and something else.
I tried to look at it with MBP on LC's supergrind, and a very rough estimation is that the lean effect only plays for like 35% of the EV gained. But I have no idea how this "something else" works

[ebon plinth]

I bet the 2nd thing comes from being ejected by the floor every frame

[tranquil venture]

maybe move to #tas-focused

[tranquil venture]

Most recent BC UR Flap ghost ^

[pine rain]

399
I started changing input midway through the big jump.

[indigo creek]

yay .3!

good job :D

[forest walrus]

🔥

[limpid ferry]

.395

[pine rain]

it seems like different data

[small scroll]

lmfaoo i sent the mg 3lap on accident sorry

this should be right

[limpid ferry]

^.394

[woeful basin]

-206 μs over this

[indigo creek]

another step towards 393

how far off is it now

[woeful basin]

before it was a full ms off, now it's 0.794 ms off

[woeful basin]

first time using csv edit

other than AGCing or respawn syncing

[forest walrus]

13.382

[earnest light]

Nice!

[woeful basin]

Incredible

[forest walrus]

thanks

[forest walrus]

13.353

[earnest light]

[dull ridge]

woah

[small scroll]

how

[indigo creek]

earlier improvement obviously

now we have to optimise the end again 🙃

[pine rain]

13.352

[tranquil venture]

I'm wondering if I can upload this flap once it's done

Last time I uploaded a collab with more than 3 people was over 5 years ago, and the ultra landing strat is probably one of the last cool glitch strats I'll find in a while

[dull ridge]

did this backtrack? because i have inputs from the RFH until the first hop after landing i think

#tas-general message

but idk

[forest walrus]

I edited the vertical inputs in the huge air during ending cut from the previous ghost and did the rest part myself

[earnest light]

Do I have any inputs on the run?

I improved the landing cut but then shxd and alego got it no stop

[forest walrus]

Ive never touched the rfh inputs before the second shroom so I think shadow has inputs

[dull ridge]

#tas-general message you did this but then i did this #tas-general message

i remember you working on the landing after though? i mean you should know

you mostly worked on daisy's run

[earnest light]

iirc my inputs started at the rfh for the landing cut and then I saved time at the very end but monster redid the ending

[dull ridge]

so basically i have inputs from where this tooks off until the landing of it

i remember trying to improve it by tweaking the shroom / rfh ejection timing but i got nothing

[dull ridge]

thats unfortunate then

[earnest light]

I’m talking about the rfh for this cut btw

Idk what ghost was used for the no stop ending cut

[dull ridge]

oh

you did? idr

well what was used then

[earnest light]

I improved it originally then a while later I got the mt for the ending cut

[dull ridge]

yeah thats what i sent, but i actually now dont remember what i used for the non-bonking landing i did

[earnest light]

This was my first improvement

[dull ridge]

can you check where these ghost desync? i cant check rn

to see what inputs i took from

[tranquil venture]

gonna have a look at this now that i finished exams

[indigo creek]

WOOOOOOOOOOOOOOOOOOOOOOOOo

[dull ridge]

wtf

how

[earnest light]

Are the geysers still good?

[indigo creek]

bro is the least washed taser in history

[silver kelp]

Its the nunchucks prob cx

[tranquil venture]

BC UR Flap - Big improvement

[narrow peak]

Congrats!

[forest walrus]

nice!

[indigo creek]

always has been

[tranquil venture]

i'm not sure how to do the ending sc properly

is it necessary to hold a left drift for a while before the RFH?

[forest walrus]

I think so

[small scroll]

i was putting in some attempts yesterday. i think part of the problem with your approach is that you land so much further left, that there's a much shorter path to the ramp and you get 2 less left inputs during the sg. i cleared the wall with a much better angle but the xz ev was less than monster no matter what

[gentle kelp]

managed to almost resync the trick 6f ahead of the old ghost, but it's a bit short in height to clear the wall

[tranquil venture]

you don't have to land further left if you don't want to

I think you can just hold softer inputs during the counterdrift after popping up through the road

[small scroll]

i know. it's possible to just turn more right before the wallclip

this was my best attempt, ghost is alego. it's either 4 or 8 xz ev short don't remember

[tranquil venture]

how does that compare to my ghost

cause doesn't mine make it over the wall and land back as well

[small scroll]

your ghost is going too far to the right after the cut

[tranquil venture]

doesn't your ghost land way too far right as well

I swear they land in super similar spots

anyway it doesn't really matter

[small scroll]

yeah i just loaded yours it's less different than i remember

one more reason why this wallclip landing is not viable then

[tranquil venture]

it's absolutely possible to save time from this improvement, it's not not viable

even if it means we don't get the full 7 frames of timesave from it

[small scroll]

it lands on the top floor 3 f earlier. of course it's viable. i'm saying the shorter path to the ramp isn't

[tranquil venture]

gotcha

[indigo creek]

Losing 7 frames for qm 👀

[small scroll]

i made it over the wall 2 f ahead but the trajectory is still slightly off so i can't finish the lap

sorry if this isn't useful. i tried but i really struggle with movement like this

[mint mesa]

if I tested right, the first mogul only seems to be trickable starting from f4985

[small scroll]

i don't think that's right

i just made the line after landing work, need a bit to see if it's actually trickable 2 f ahead

[mint mesa]

i have this rn

[small scroll]

the flap leaves the first mogul on this frame. how does it work, does it store the surface and only allow the trick from the frame you said?

[mint mesa]

nvm it does seem to be 4984. If I slow down to trick on this mogul it starts the animation earliest on 4985

[small scroll]

hm that's awesome frame 4984 is the 11th frame of airtime if i try to follow a similar line

i definitely don't know the game well enough to work around that lmao

yeah i have no idea if this is salvageable, have fun guys

i forgot to turn off my music sorry it's loud

[mint mesa]

this is not correct, I just tricked late.

it looks like it depends on when you leave the surface

f4975 vs f4976, both try to trick instantly

so f4976 is the earliest we can leave the mogul and get a trick

[small scroll]

i wasn't able to leave the mogul on frame 4976 but i'm also not that skilled with odb

either stick or 4975

[mint mesa]

so u got 4973 as well?

[small scroll]

i think i miscounted a lot when i typed that lmfao

[mint mesa]

i got 4974 only changing this line slightly

[small scroll]

yeah i was able to get earlier than 4975 too, not later though

[mint mesa]

i mean if it was possible, earlier should be better no?

[small scroll]

no clue

[mint mesa]

i wonder if its possible to delay lap1 by a few frames and resync it. the cycle kinda sucks anyway cuz you have to go forward a lot instead of full left. this mogul is so small atp youre not even getting airtime here

the problem is that one mogul is still expanding during the setup

[small scroll]

it would be a lot less effort to lose one frame from what i did i reckon

[earnest light]

Resyncing wouldn’t be easy because monster doesn’t stop for the ultra and how incredibly qm dependent doing the ultra is

[tranquil venture]

man it kinda bugs me knowing that there's multiple frames of missing timesave here now

[tranquil venture]

wait so the bkt leaves at 4977 and we can potentially get 4976 to trick

[tranquil venture]

one thing to note is that an ending strat like this is possible without tricking off the first geyser

[indigo creek]

the lap 1 sg for ultra doesn't use optimal ev inputs anyway, so it's not the only thing missing

just another thing for the next tas to implement

[tranquil venture]

the ultra is so unpredictable that we don't even know if that would save time

this yoshi ultra is faster than all the daisy ultras that have been made

[indigo creek]

that's a fair point

[gentle kelp]

landed the cut 6f ahead, but didn't got a good bounce of the wall after, but i think it can be fixed

[tranquil venture]

alegoat

[small scroll]

the angle can be fixed by starting the sg a bit more to the right and adding some turning inputs during neutral frames in the sg

but are you sure that this can carried without the trick

[tranquil venture]

I'm working on fixing it now

[gentle kelp]

i was thinking that if the trick was not possible , we could do this

[small scroll]

oh i didn't see that, but that looks like getting that setup from the cut will be harder

[tranquil venture]

tiny bit further left

got a landing that might be workable?

less yaw but only slows down to 30 or something when it touches the wall

[tranquil venture]

have to go for a bit

someone else can try

the ramp ejection is very sensitive, editing a single input during the SG seems to almost ruin it

[earnest light]

are we trying to get a good bonk off the wall

(one that gives airtime after)

[tranquil venture]

basically just replicate this as best as possible

so yeah the 1f airtime is part of it

[dull ridge]

might get back to this

[earnest light]

got the airtime after but the landing was slower and didnt go far left enough

[earnest light]

i believe i land slightly earlier than ejays ghost and i only get one frame of airtime after the but i think thats because i have a not great angle going into (or its )

i think ejays ghost is more useful rn because he has a faster bounce upon landing

[dull ridge]

i was able to get a slight improvement after landing but i think angle isn't ideal

[dull ridge]

yeah i can't make this faster

[mint mesa]

vs Ejay. bonks on the corner to drift pretty early

[tranquil venture]

paul's started collab tasing, it's over

pack it up

[mint mesa]

not just gonna sit on my non TAS BKT having ass and watch

[tranquil venture]

that ghost bonks the wall about 7.3 frames ahead of this video: #1319937435152416768 message

the flap time (linked video) before we started tricking off the geyser was a 13.557 (slightly unoptimised)
the flap time before i found the improvement to the beginning was 13.352

so it seems like not tricking loses 0.2, but 7.3f = 0.120s < 0.2, so it's not looking good for saving time with this, unless someone finds a no-geyser-trick line that's faster by 80ms

[small scroll]

we should delay the release of this tas by 3 more years to fully test daisy and geyser cycles

[mint mesa]

vs the 352. got a delay trick here leaving the geyser at f4976 w no airtime. angle sucks balls tho

[mint mesa]

I only got this to work but I counter steer so late that I lose lots of momentum. didnt have any luck doing it earlier yet.

[mint mesa]

ultra attempt about 8 frames delayed

I only did the optimal ev building in a few places cuz it messes with ur line a lot and I couldn't correct it well, so EV is basically the same

[tranquil venture]

mad respect for trying

[mint mesa]

well I haven't really tried until after the tree clip cuz I'm not familiar with this ultra. should've said setup attempt instead ig

the setup might be ahead too much so u get air from touching the wall on this frame while the ghost doesnt (I think?)

[small scroll]

for ledges like this it's a qm issue from bad positioning. there are infodisplay sections for position and ev difference, try to match the best you can

[gentle kelp]

landed the cut with paul's ghost

the cut can maybe be fully resynced

ghost is previous completed lap with 8f delay

idk how to do the landing strat, and idk if it's possible with this landing position

[forest walrus]

wow amazing

[gentle kelp]

Ty !

[mint mesa]

incredible work man

[gentle kelp]

thanks ! and it's thanks to your work that i'v managed to get it !

[tranquil venture]

neither do I (and I'm the one who did it )

[small scroll]

took a quick look at alego's input vs monster's original. the qm matches up, but the reason why the landing falls short is that less ev is preserved after the wall collisions.
after the tree collision, you end up with 1.2 xz ev less than the initial ultra

and after the pillar collision, you drop to 2 ev less than monster. for this reason you end up slowing down earlier by the end of the ultra.

but before the ultra, position and ev were synced well, so this should definitely be possible to resync. i think instead of replicating the landing with this attempt it's worth trying to optimize this ultra to carry more ev

i would like to take a look but i'm busy with exams sorry

also, it would be interesting to compare this to the current bc 3lap daisy ultra. maybe we can figure out why the attempt with daisy is so slow

[mint mesa]

this much air might be required to not hit the boost panel

[small scroll]

wait what is the player

[mint mesa]

the 3lap lap 1

i used agc player

[small scroll]

oh yeah i forgot the spiral strat was completely different lmao

the main thing i was curious about was the ev building. obviously daisy can still reach 120 u/f and get all the qm, but she has much less ev so she slows down early. are you able to compare with agc if the rate of building ev is similar between her and yoshi?

i don't think it's confirmed if yoshi's offroad rotation actually makes the ultra faster at all, for example by giving a bit more ev

[mint mesa]

daisy has 7 more EV before the tree bonk

[small scroll]

does yoshi start with more ev bc of the reserve sh setup? or gain ev more rapidly throughout the sg

it's well possible that the flap may be able to spend time before the lap begins to get a faster ultra

[mint mesa]

nope daisy has more ev before the first geyser, building rate seems basically the same

[small scroll]

more ev? does she take a faster path through the grass to end up with less later?

[mint mesa]

then after that she has to correct right a bit so before the shroom she has only 15ev more. but yeah the sg entry allows daisy to hold left a lot more

but she can turn more left after that because she has less yaw than yoshi

well more yaw technically, -136 daisy to -148 yoshi

[small scroll]

i see. ty for checking

it's good confirmation that at least the character is right bc that wasn't 100% clear. sounds like the 3lap ultra can be improved then. although it's likely too late now

[mint mesa]

a big difference is also the angle into the tree bonk, yoshi gets sent a lot further back than daisy

[small scroll]

that's interesting, yeah there are a ton of moving parts to this ultra and you prob figured out a bunch of them with your attempt, there def should be enough freedom of movement to fix that wall hit

[mint mesa]

almost same speed as Monster's original glitch but lacks height and -Z distance

working setup a little more delayed

there's a lot of stuff you can do with Alego's setup (he changed some inputs during the sg and gets a higher tree bounce) cuz his ghost is more in front of the original glitch. We'll get it eventually

[gentle kelp]

got almost the same speed, i think this can get closer to the ghost position at the landing to be able to get the turnaround

[small scroll]

looks good! rather than change ev i would try to mess with pitch angle shortle before/after colliding with pillar since the difference in speed happens from there. it looks like your trajectory curves a bit more upward

[gentle kelp]

ty ! and yeah it might be possible to match even more with the ghost, but this is the best i managed to get for now

[small scroll]

i feel like it may be possible to replicate the previous input with this already

[gentle kelp]

if it's possible to get a good landing with this bounce, it can be way faster than the one before i think

[mint mesa]

successful turnaround

i used this track edit w/ inversed material culling to see whats going on during the frame advances above

[tranquil venture]

how are you getting this good so fast

[soft hare]

holy

[earnest light]

New Taser of the Year 👀

[small scroll]

i called it before anyone else just saying

[echo birch]

dammit i was gonna say was lirwm said

i didnt wanna clog the chat

[gentle kelp]

wow nice job paul !!

[tranquil venture]

let's gooo

how easy was it to sync

[small scroll]

🔥

[gentle kelp]

the hardest part was to finish to resync the turnaround, the rest was pretty easy

finished resyncing the turnaround and turn skip, the ending can still be improved i think

[tranquil venture]

wasn't the og improvement like 7 frames ahead?

[gentle kelp]

it was 6f ahead

[tranquil venture]

ok so we should be able to save like another 10ms

[small scroll]

full resync should give 252

[tranquil venture]

by messing around with the geysers

[tranquil venture]

the geysers aren't exactly the same

it probably doesn't matter much but still

[dull ridge]

oh shit!!!

[pine rain]

I think it is still possible to optimize within the castle.

[tranquil venture]

A few of us got that improvement previously, but the ramp trick QM is very strict

Also the further left you are, the higher the trick needs to be to make it over the wall

[small scroll]

also, the trajectory in the air needs to be considered. if you trick from the ramp more left, you will need more right rotation to clear the wall, but too much right rotation gives a bad landing for the end of the lap

[pine rain]

I was able to shift the landing a little to the left after the jump trick. If you keep running like this, you may be able to ahead.

[tranquil venture]

Might as well save the rkg

[tranquil venture]

Nice!

[main valve]

for future reference, when you want to resync a flap setup like that : it's much much easier if the original flap setup lose a significant amount of time by doing nothing (in like a stand still for a while), so you have time to make tiny position adjustement and get basically much more freedom to resync later if needed

Not using vertical inputs in the setup is also very helpful, if you need to tweak those vertical inputs later to get tiny adjustements

[mint mesa]

insane stuff alego

[mint mesa]

yeah I've thought about this as well. It's not possible here though because we cross over an expanding geyser during the setup. but for other tracks that require a cycle during the flap that doesnt affect the setup it's def good to do this

just waiting at the start might be the play

[small scroll]

most flap setups begin with an ssmt for this reason, or at the very least have waiting periods

[earnest light]

Just wait after landing the ultra

[small scroll]

for whatever reason monster happened to do this one in one motion

[main valve]

I also mean, when you tried to do the ultra setup again with like a 8f delay : you could have done a SSMT and waste time anyway, and whenever you get close to monster's attempt, you can start modifying the position at the SSMT, to fully resync from there

to make the process from almost resync to fully resync easier

[tranquil venture]

Because he wasn't making a flap setup

He was just trying to prove that one shroom was possible

[small scroll]

oh lmao i didn't know the setup was never remade

[tranquil venture]

Doing the ultra sucks so much

Nobody wanted to remake it

[mint mesa]

kind of lucky then that the geyser cycles were relatively good

if the lap1 was like 30f ahead it'd be cooked

[tranquil venture]

The fire bar too

[mint mesa]

38f i would not have been able to adjust

[small scroll]

yeah i assumed the cycles were sort of calculated

[mint mesa]

Monster is just built different

[tranquil venture]

the original flap was a 15 something

I hadn't found the new landing and wallclip strat when the ghost was made

Which saved 2 seconds

[pine rain]

13.250

[mint mesa]

.249

[small scroll]

244

[mint mesa]

243

[gray flint]

242

[echo birch]

why did the cut have to be remade? was pauls desync(?) at the geysers a cause or was a mini-optimization found

[small scroll]

ejay's wallclip improvement saved several frames, enough that the first geyser was not yet trickable which lost way more time than the improvement

[echo birch]

right

241

[mint mesa]

aight so who has inputs and where

[hidden crag]

everyone getting credit at the end 🔥

[echo birch]

i started my input as yoshi hits the ground fromt the last geyser trick

[mint mesa]

mine should be 3789-4392, 4632-4800, 4949-end

Alego should be 4023-end

[echo birch]

right i should count frames

5080-end i beleive

[mint mesa]

Aoiro 4536-end I think

[indigo creek]

This tas is just covered with new tasers and I’m all for it

[dull ridge]

241 - 0.03ms timesave

according to exact finish

[dull ridge]

👍

[indigo creek]

keep in mind that was at the start of the night, when timesaves hadn't started

sub-millisecond cuts are probably fine now that we're at -0.001 territory

[echo birch]

ic

[dull ridge]

i also don't know if my inputs from like 1.5 years ago were backtracked, btw

[indigo creek]

i think he was talking about the improvements that just came in

since no videos were shared

(i think videos seem to be a standard for figuring out inputs now)

[dull ridge]

makes sense

[mint mesa]

no I meant in general

[indigo creek]

oh

[dull ridge]

oh

[mint mesa]

oh

[small scroll]

oh

[dull ridge]

⠀

[echo birch]

you ruiend the oh chain

[indigo creek]

good luck with that because this project is older than this thread

[small scroll]

this was posted before aoiro got the 250 so the ending was known to be worse than the previous version (now it's better)

[indigo creek]

and has had multiple restarts for various reasons

[mint mesa]

the recent ones you can just check w pycore infodisplay once it's no longer tied

[dull ridge]

this were my inputs btw and i don't know if they were backtracked

[mint mesa]

do you have an rkg to that?

[indigo creek]

yeah it's in tas-files on the same day

#tas-files message

[small scroll]

i have no clue where i have input. @gentle kelp when you carried ejay's improvement to 6 f before paul redid the entry, did you use my 2 f improvement input at all?

if so my imput starts at the wallclip from the floor below, otherwise it's somewhere in the landing idr

[indigo creek]

if you just search "bc ur flap" in #tas-files youll get an idea of the full history of the project before it got here

that applies to any run made before december 2024 btw

[small scroll]

(also it's just easier to simply list authors and not waste 3 hours determining who did what)

[indigo creek]

unless it was also made before july 2021, in which case, good luck

[echo birch]

that really is your thing corvid

[small scroll]

yes.

i think it's important for a project like this with a bunch of people who saved 1 ms, it sounds insane but it's worth the same as the people who did more work to me

[dull ridge]

now that i think of it a cool little script that takes 2 rkgs or csvs and returns the frames that aren't equal would be cool (but useless maybe?)
it's not like it's hard anyways

[indigo creek]

it isn't helping anybody complaining about name display every time it's brought up

[small scroll]

so i will keep saying it every time, and kierio will fight me over it every time

yup

[dull ridge]

lol

[echo birch]

lmao

[mint mesa]

[indigo creek]

im not even fighting him over it this time, im just saying, as a moderator, that i think the complaining should stop

[dull ridge]

ban him

[main valve]

there is a command line for this iirc, like the diff of 2 files in general (works for csv)

[dull ridge]

oh nice

i was thinking of either player + ghost CSVs, or file dialog of 2 rkgs

[indigo creek]

reinventing git again...

[mint mesa]

you can also just compare them in VSCode

[dull ridge]

we should put our CSVs on a big ass repo to see the diffs.

[indigo creek]

thats actually how celeste tas community works ^

[dull ridge]

i have the git extension but i can see diffs only if i commit or merge conflict
i assume there's a standalone extension for doing that

[indigo creek]

every tas has its own github repository, and usually its own discord server

[echo birch]

own discord server is crazy

[indigo creek]

any time a project is started, the invite is shared in their equivalent tas projects channel

[echo birch]

then again i dont know anything about celese tases

[small scroll]

am i allowed to complain about the github idea being brought up again or is that a moderator-exclusive privilege

[indigo creek]

they regularly pull in like 30 active tasers per day

so organisation is like a massive top priority for them

[dull ridge]

do you mean that git-like process that Blounard brought up

(again, git not github)

[mint mesa]

its built in actually, if you open the folder with the .csv files, right click one then click "Select for Compare"

[small scroll]

yeah. it's just kind of... unneeded

[dull ridge]

wait rly

that's cool i didn't know that

[indigo creek]

we're straying from BC discussion, should probably move to #tas-general

[hidden crag]

how many?

[small scroll]

paul, token, skip

[hidden crag]

i might try be a 4th

[small scroll]

you wouldn't be new bc mg flap

[hidden crag]

shucks

[dull ridge]

that doesn't stop you from trying

[indigo creek]

by new i mean recent not no bkts

[hidden crag]

hell yeah

[indigo creek]

yeah see if you can turn shadow's 241 into a 240

13.240 is a cool time

[mint mesa]

241 is also cool cuz we saved 6.66 frames over the 352

[echo birch]

https://tenor.com/view/mischievous-m4ng0z-gif-devil-emoji-devious-gif-gif-17606276402347397816

[gray flint]

Just get sub 13 real quick shouldn’t be too hard just be faster

[echo birch]

yeah

[dull ridge]

i agree

[echo birch]

its only 250 ms rta people do cuts like that all the time why is it so hard for us

[dull ridge]

just do the ultra with daisy rq

[small scroll]

it's not too late to restart with daisy...

[mint mesa]

I found some massive time save recently

step 1 is enable this button right here

[dull ridge]

i would enable the one above actually

[small scroll]

have you tried enabling 300cc too

shad shut the fuck up

[dull ridge]

bro stop copying me

[small scroll]

zitto

[dull ridge]

first with daisy and now with ts

[small scroll]

taci

[dull ridge]

polentone

[echo birch]

pizza

[mint mesa]

spaghetti

[echo birch]

https://tenor.com/view/italian-river-speed-boat-perfecto-gif-17332909

now im off topic

[dull ridge]

https://cdn.discordapp.com/attachments/848304619733647363/1291061055174213693/caption.gif

[mint mesa]

that worked really well tbh

[dull ridge]

ok back to bc ur flap

[gray flint]

🦗

[hidden crag]

idk anything about this track, how useful is something like this?

i'm slightly ahead and slightly more to the left

[mint mesa]

ejay 4339-4620
shadow 4496-4620
lirwm 4496-4620
corvid 4733-end
monster full run?

[indigo creek]

Monster, LirWm, shxd, Ejay B, ego, César, clem, Alego11, Thomas, Aoiro, kierio04, Justin, Tactical, Paul, Corvid, Skip, token
these are all the people who have helped with some version of this project at some point

up to someone to figure out who gets cut due to backtracks

[woeful basin]

^

I got cut due to a backtrack though

[spiral token]

you say ''with some version of this project at some point'' but does that count or is it too outdated? https://discordapp.com/channels/214574342137315328/600410593937195069/1221954050786853024

i'd assume outdated since the whole flap was redone

[indigo creek]

missed a few oops

[mint mesa]

shad up to like 4720 I'd say since he was the first to the get the fast ending cut with the sg

#tas-files message

[dull ridge]

oh yeah i remember i also got the drift landing

[hidden crag]

i'm posting this here on the off chance it's useful

[dull ridge]

seems like you are higher there, but maybe it is

[hidden crag]

iirc after i landed i just let tas code do the rest

i hate outside drift with a burning passion and couldn't be bothered to do more

[dull ridge]

that's fair lol

[dull ridge]

^

[hidden crag]

shucks

[dull ridge]

but it would need testing

nothing for sure

[mint mesa]

Monster start-end
Paul 3789-4392, 4632-4732, 4949-end
Alego 4023-end
Aoiro 4192-end
Ejay 4339-4620
LirWm 4496-4620
Shad 4496-4732
Corvid 4732-end
Skip 5040-end
Token 5080-end

please double check for urself and then give me a

[echo birch]

for me3

[indigo creek]

still 10/17 not bad

[woeful basin]

Is there any room for improvement

Everyone jumped to name display right away fsr

I feel like at least another ms could be saved

[pine rain]

This data started to change the input immediately after hitting the tree.
It's easier to see when comparing it with the ghost in 13.264

[mint mesa]

should be this then. I noticed it but I wasn't sure if it was from my input because I experimented with these frames a lot as well. thanks for clearing it up 👍

[gentle kelp]

no i didn't used yours

[dull ridge]

if we considera my last ghost then i actually also have some frames at the end

but i dont really care

[pine rain]

I was able to move it further to the left.

[small scroll]

are you able to get a good slipdrift from the wall? i was able to land further ahead and more left than the 241, but it always resulted in a worse wall collision

[pine rain]

I achieved great results. I was able to pass through the wall without hitting it.
However, I can't perform the jump trick at the first geyser...

[tranquil venture]

Time to delay the flap again

[pine rain]

oh...

[dull ridge]

holy

[tranquil venture]

Am I correct to say that gets airtime from the geyser, but still can't trick?

[mint mesa]

nice work! i think this should be doable though, the 352 left the geyser on 4977, so 1f later than when it becomes trickable. now with 8f delay, getting ahead 9 frames should still work

[pine rain]

yes. probably he touched the geyser before the jump trick was possible.

[mint mesa]

yeah you have airtime on f4975

[pine rain]

airtime was "1" at 4975f

[mint mesa]

#1319937435152416768 message i got this 1f ahead of the 352 where I touch the rim of the geyser, but idk if its possible with your ghost to go more to the left

[mint mesa]

can u send the ghost? i think lua core info display might be different from pycore (with UI delay removed)

[pine rain]

ok

[mint mesa]

ah nvm its the same. you probably need to be even more to the left during the cut, as to not go over the bigger part of the geyser and get airtime early..

[ebon plinth]

using git to track csv changes would go kinda hard

[dull ridge]

btw the problem with pycore infodisplay was that it was 1f delayed, lua core was 0f always

[mint mesa]

only ever used lua core for a TAF lol

yeah you can get more left quite easily

[dull ridge]

where do yall delay the flap btw? this setup doesn't use SSMT

[mint mesa]

i did this to delay it

[dull ridge]

oh so like just manually trying to recreate the position

[mint mesa]

yup

[dull ridge]

that must be annoying

[mint mesa]

yup

[dull ridge]

when i worked on daisy's setup i intentionally first did SSMT and tried to avoid hitting any geysers in the SG, so that would make resyncing much easier

but one of the problems might be not hitting any geysers, in yoshi's setup iirc going on the geysers make you gain that bit more of EV

[main valve]

if you want to re do the glitch, I highly recommend doing a stand still, and wait as much as you can (like if possible gain time on the beginning of the run to have more time to waste during the stand still)

so when you are close to getting the ultra again, you can try correcting it just with slight position modification during the stand still part (which is easier when it's longer)

[hidden crag]

so just to confirm, we're redoing the setup again

[mint mesa]

i managed to get a bit more to the right than this image

[mint mesa]

no we're gonna try to make aoiro's timesave work first

[hidden crag]

oh i was able to get further left than aoiro

gimme a sec, i gotta load it up

[mint mesa]

and more to the right than me?

[hidden crag]

oh we need to go more right?

what's up w that

[mint mesa]

no we want to be left of aoiro but not too much, we want basically the perfect alignment so that we can hold full left through the drift and scrape the edge of the geyser to get airtime very late and get the trick out

[hidden crag]

why not delay the setup again so we can be as far left as possible without causing issues

[mint mesa]

if we're too far left we have to counter steer a bit which fucks the momentum off of the geyser

just send the ghost

[hidden crag]

yeah so uh

i don't actually have the ghost because i was reading this chat and thought we'd have to do the setup again

so i just didnt bother saving it

• i don't know if it's more to the right than yours

[mint mesa]

ah damn

[hidden crag]

i could just recreate it, took me about 10 minutes

i'll do it later though, doing other stuff atm

[main valve]

also it could be nice to use agc load player before the last ramp to delay the player, then take control back and see which delay is optimal

[small scroll]

i think that is way too much effort when we're using the wrong character

this flap will leave some timesave on the table, and that's okay

[hidden crag]

wait, it's confirmed that daisy is faster?

[tranquil venture]

It's not confirmed

[small scroll]

we had a conversation about this. i don't know what other conclusion to possibly draw

[tranquil venture]

are we sure that even though EV is the same during the SG for both, the EV vectors are too?

I kinda assumed the reverse SHing changed the way the EV is released at the tree bonk

providing more left momentum or something

[mint mesa]

i feel like it just comes down to the approach no?

because of the different approach yaw and roll are quite a bit different, also the sg ejection happens at different qm

first frames of airtime, left is yoshi, right is daisy

ignore the frame count I might have trolled w agc

[small scroll]

the direction of the ev vector appears to be roughly the same (i did check with pycore's speed display which isn't the best quality but it seems like about the same angle)

after the block touch, daisy has -3 ev. after the pillar clip, daisy has -7 ev. and that's the whole reason why she is so slow in the air. but at the tree hit she actually has +1 ev

there is simply no way that yoshi's stats make any difference for that; it must be a qm difference

[mint mesa]

also daisy reverses on the wall a lot more noticably as she has backwards distance to make up for

[small scroll]

pay attention to the "ev diff" values in the bottom left of infodisplay

[earnest light]

where would daisy gain time during the flap

[mint mesa]

daisy just gets way too much air i feel like

[small scroll]

after turnaround

and after wallclip

[mint mesa]

I thought the air is needed to avoid the boost panel but I managed to avoid it using yoshi on my ultra setup

[tranquil venture]

I don't think we should worry about lap 1 of the 3lap

[mint mesa]

alright then who wants to do the simple task of redoing the ultra with daisy using yoshi's alignment/approach

[tranquil venture]

and the flap is good enough already, for how volatile this whole lap is

[small scroll]

but is it possible that optimizing the ultra better can make us beat the l3 firebar cycle?

[hidden crag]

the council has unanimously voted for you to do that

[dull ridge]

can we just optimize the last bit of yoshi, upload it, and then worry about daisy later?

it's already pretty much optimized

[small scroll]

for the record i agree with this. and i also think it's not worth it to redo the whole setup to carry aoiro's improvement (sorry man)

[mint mesa]

again

we can likely carry Aoiros improvement still with this cycle

maybe not all of it though

[mint mesa]

[dull ridge]

this is how close we were with daisy btw

[earnest light]

Why have we gone back to doing the ultra with Daisy 😭

[orchid plank]

I guess we're all dreaming of sub 13

[dull ridge]

we didn't

[mint mesa]

man fuck this geyser

[mint mesa]

yeah I'm giving up on this pace. might need to delay the 3rd shroom to get this to work

have to touch the geyser on this pace by being a bit further back (which the ghost is, but it gets airtime early)

player ^^^

ghost ^^^

I couldnt slow myself down any further because of how precise the qm is. I already managed to go a little slower by changing f4749 but lowering the input more gets me a huge bounce while landing, or when avoiding the bounce through other changes I dont get airtime before the drift, or the drift doesn't go as much left as we need

[mint mesa]

I just did this with this ghost which goes as far left as I was able to from Aoiro's ghost.

3 more frames of delay seemed the best as I'm a frame ahead and have less X coordinate distance, 2 also seemed very good

plus how much Daisy is faster than Yoshi but ig it's hard to estimtate

for now I'd be fine uploading this run once Aoiro's timesave has been implemented as good as possible. imo not worth delaying the ultra again, at least not before we've tested the ultra with Daisy extensively and can't get the fast yoshi ultra with her

[gentle kelp]

so here we are trying to carry aoiro's timesave ? is that correct ?

or are we trying to redelay the ultra

[mint mesa]

In my opinion, yes

I'm gonna try to make the ultra with daisy from the shroom to prove it's possible

agc player up until a few different points without changing the input. I honestly don't see why it wouldn't work

before that, it desyncs because of different stats (mostly handling stat) and thus doesnt get a good tree bonk

since the UR 3lap is at the lap 2 ultra, it might be better to try and go for the fast ultra there first and put this flap on the back burner for now once the yoshi run is finished and uploaded

[mint mesa]

It seems really difficult with this pace. Delaying the shroom always gave me a bad landing cuz the qm is just sooo precise, might not even be worth the effort then. I don't mind uploading the 241 instead and implementing Aoiro's new strat in the future flap but I don't have a strong opinion on it.

[earnest light]

Will anymore work be done on this or are we calling it finished

Also who is uploading it

[woeful basin]

monster I think

[earnest light]

Main problem with that is his backlog of videos
I think Ejay said something about wanting to upload it but idk

[indigo creek]

I mean if you asked monster you could probably upload it yourself

[pallid holly]

@tranquil venture upload or no

keep the momentum going

[tranquil venture]

Probably not anymore

[mint mesa]

I wanna get the conversation about what to do w/ the Yoshi run going again. Since Monster and myself pulled off the fast ultra with Daisy ( #1344592240164470806 message ), this can ofc be applied to this category too.
There's been lots of effort by a lot of people to get this far and it'd be a bit of a waste to not have the progress reflected in this category's BKT history. More importantly, Yoshi is just a really cool guy.
In the end it's gonna come down to whether the character switch, some new strats (Aoiro's new ending sc strat paired with better geyser cycles) and the time between uploads make the runs distinct enough to warrant an (unlisted?) upload of the Yoshi run.
When the Daisy will be finished or even worked on in the first place is another question ofc. I personally want to focus on the 3lap of this glitch and NG categories first (won't have time until October tho).

imo an unlisted upload is the way to go as to not take too much shock value off of a final, optimised Daisy run

[dull ridge]

i think we should upload it not unlisted at all, and then put a disclaimer (in the description probably but no one reads it) that a daisy run is in the works already

imo all the strats and time each player has put in this run is not worth an unlisted video
it's been almost 2 years since the start of this run

[small scroll]

i am not a fan of uploading unlisted. don't let perfect be the enemy of good

[woeful basin]

you're in a boost for a lot of the flap so daisy won't save much, I'd say upload it and wait another 2 years to pull it off with daisy

[tranquil venture]

I agree with Paul and think an unlisted upload is a good option

[dull ridge]

i don't agree. just why

[dull ridge]

as to not take too much shock value off of a final, optimised Daisy run
as if this yoshi run isn't optimized

i mean i was one of the first ones to actually work on making the ultra with daisy and complaining about continuing the work on yoshi, but now that it's fully finished and optimized i don't see why it should get unlisted

[tranquil venture]

because it's not optimised

[modern trellis]

Wouldn’t it be more correct to say it’s not optimal?

[dull ridge]

^ i think it's optimised enough

to get an upload

[tranquil venture]

it's not fully finished because the geyser cycles have once again cost us time

the first flap is going to be the one that everyone sees, so I'd rather have that one be the flap that we know will be faster

[tranquil venture]

This isn't true

We're in trick boosts for a lot of the flap, but rarely in a wheelie simultaneously

Which means Daisy will save a lot

[small scroll]

what kind of bullshit is this? who is "everyone," would it not be better to not upload the yoshi run at all then since everyone in the community will be spoiled by this suboptimal run?

[mint mesa]

that is not what I was implying. I just said the Daisy should also be optimised obviously lol

[woeful basin]

the whole "wait till the run is absolutely perfect before upload" culture is the thing I dislike most about mkwii. I could care less about more shock value and attention going to a slightly unoptimized run and on top of that, the average viewer won't nitpick this run and say "erm, you could save 0.00000001 seconds if you use daisy, so you have to spend another 2 years redoing the run". Also, yoshi deserves the attention tbh

[mint mesa]

It also depends on who's gonna upload it. ik Monster likes his uploads to be pretty much perfect so I doubt he wants to upload the Yoshi run, but he's the main author, so that's why I suggested an unlisted upload in the first place.

Need to hear what he says first

Tho it's also true that Monster gets a lot more views than most people uploading runs, so it's likely many people would see the flap there for the first time even with a public Yoshi run out there already

[woeful basin]

"lirwm upload since he created the thread" 🔥

[pallid holly]

as someone who has no stock in this I think you should upload yoshi because yoshi is cool

[woeful basin]

because DashQC is cool*

[tranquil venture]

by everyone I meant the general public

the first BC flap that's publicly uploaded to YouTube will be the one that gets the most attention

the community won't be spoiled because there won't be a bkt ping for it

also why are you getting angry for?

[tranquil venture]

The average viewer can't tell whether a run improvable by 0.5 is unoptimised or not, it looks the same to them - so under that logic why bother optimising anything before upload? It's because you upload something that you're proud of to your own standards

[woeful basin]

it's not like this is a 2:09.223 on kc ur 3lap. This is a very optimized TAS that uses optimal strats for the character used. The only know timesave (other than daisy and extreme micro-optimizations) is a little from the geyser cycles. Disregarding my personal upload standards these days (which are a lot better then they used to be), I still believe months of work is not worth 0.5 seconds of timesave in a video game, and therefore, the yoshi flap should be uploaded.

[tranquil venture]

ok. I disagree

hang on are you an author of the run?

[woeful basin]

I don't think so, just sharing my opinion

pretty sure I got backtracked

[tranquil venture]

the people who spent this most time on this run (paul, alego, me) seem to be the ones who want it unlisted lol

[woeful basin]

you want it unlisted? I thought you didn't wanted uploaded period

[woeful basin]

somehow didn't see that

[tranquil venture]

[indigo creek]

insane to see you say this

but yeah this is just a repeat of the toadette/baby daisy situation on rsl

just upload the best we can get with the current character (yoshi) and then go all in on the new character (daisy)

[tranquil venture]

not really though because the Yoshi run has more timesave possible

I was under the impression that the toadette run was thought to be optimised

[indigo creek]

yeah it was

i didnt realise yoshi had more left

[tranquil venture]

the geyser cycles are bad again

[indigo creek]

well if the run's gonna be redone with daisy then they can be perfect in that tas

[tranquil venture]

sure

[indigo creek]

yoshi was redone enough times and deserves to be uploaded properly

[tranquil venture]

nah

[indigo creek]

why

there is no point in not uploading it publicly, we have no way of knowing how long it will be until the daisy tas is finished

[tranquil venture]

all the points are in chat already I'm not going to repeat myself

main authors of the run want it unlisted, that's that

[indigo creek]

who cares about which tas gets the most attention, the tas exists and not uploading it because you want an improved version to get more exposure is a bit silly imo

[tranquil venture]

why is that silly?

why is it silly that we want something we're more proud of to get more attention

I think that's perfectly reasonable

[indigo creek]

why have we somehow gone back to not being okay with uploading a tas that can be improved? i thought we finally got over that a couple years ago

[indigo creek]

also i have to agree with tactical here - it is so tiring to have to defend tases being shown to the public over and over again every time the semantics are ever so slightly different to some general understanding in the past

it shouldn't have to be this complicated

[tranquil venture]

it's really not complicated. run improvable + wrong character = rather showcase optimised run + right character

you're not even a key author of this run so idk why you're pushing so hard

there is no "culture" that tactical speaks of, nobody gets flamed for uploading things that aren't perfect anymore

it's only people's personal standards these days

[indigo creek]

because as soon as this decision is made for one run it will start happening in other runs

sure, we have improved compared to the past - people are no longer flamed for not having the same standards

but people still have unnecessarily high standards, and that is just harmful for the community imo

[tranquil venture]

something something logical fallacy

if it happens again then it'll be because of the BC flap authors' personal standards (which have existed for ages), not because of precedent set here

[tranquil venture]

the standards are higher but the tools to achieve those standards are easier

[indigo creek]

the justification for not uploading an unoptimised run is always that it won't take long for it to be optimised and that never turns out to be true

[tranquil venture]

is that the justification for this? I don't think so

[indigo creek]

well do you think it will be long before daisy flap is uploaded?

and is that part of why you want to wait for it to be the first public flap?

[tranquil venture]

I don't know and I'm not really making my decision based off that

I redid everything in the Yoshi flap from the ultra landing trick all the way up to the ramp trick in like 10 minutes

I don't know how long it'll take with Daisy

[indigo creek]

well exactly, it could be weeks or months before someone is interested in optimising everything again with daisy - during that time, people in the public still won't have seen the yoshi flap that could have been uploaded in the meantime

[tranquil venture]

that's the point

[ebon plinth]

just agc it and finish in 3 days

[indigo creek]

.

regardless of whether it is true or not, i think it shouldn't be used as an excuse at all

the problem is still in the fact that people are willing to throw away nearly completed runs because of known improvements

[tranquil venture]

why?

we're literally living in the thriving era of perfect runs and it's never been better

I'm loving it

and again that's just based off of authors' personal standards

[indigo creek]

having high standards is great when all the strats are found and the run goes straight to being uploaded without any hurdles

[tranquil venture]

there's still no community culture

shadow and tactical already complained about NG being optimised to corvid so you can read corvid's responses in this convo #tas-general message

I'm not going to repeat his points

[indigo creek]

im not talking about tasing being too hard or anything

[tranquil venture]

I really don't have time to talk more, and my mind is pretty set on this flap

[indigo creek]

im really disappointed that you guys are deciding to just upload it unlisted

a lot of work went into the yoshi flap

[ebon plinth]

I have no right to an opinion here but I don't even think it should go up unlisted

[indigo creek]

if i cant change your mind then so be it

[tranquil venture]

and that work will be preserved on the bkt history

[indigo creek]

i really hope this kind of mindset doesn't spread to other projects though

[tranquil venture]

would you be advocating for BC3 NG 3lap to be uploaded if it was driven with Rosalina and had to lose 0.2 for bad thwomp cycles?

[indigo creek]

no, rosalina was never faster

yoshi was thought to be optimal

then it was discovered daisy could do the good ultra

[tranquil venture]

Yoshi was never faster either we were just coping

[indigo creek]

after the tas was basically done in fact

[ebon plinth]

we have the good ultra with daisy right? the hard part is already done, it could genuinely be finished in a few days

[tranquil venture]

yes we just need to replicate QM on the landing trick

[indigo creek]

this is true but also i think you're missing my point

imagine if 2 months ago people found out that a different character was faster for MT UR 3lap

before the end of lap 3 was done

would you be advocating for that run to be restarted

[tranquil venture]

yes

[ebon plinth]

I would've agreed with you a year ago but now AGC invalidates your point imo

[tranquil venture]

well I'd probably finish the bad character run and upload it unlisted first

[indigo creek]

it's not just the fact that it may or may not be done quickly

[ebon plinth]

work on an old run can easily be copied over now, or in this case can easily be referenced for matching QM. The work isn't lost

[indigo creek]

it's that the first TAS had all the hard work put in to develop strats and come up with new optimisations by all the different authors, and then you're saying we should just upload that unlisted because the new run with less strat development and less people involved is "more important" to be published first

[tranquil venture]

you keep talking about hard work, but the people who put in the hardest work are the ones who are wanting an unlisted upload lol

I discovered the 2 main strategy changes as well

[indigo creek]

i mean you're forgetting the dozen or so other authors who got backtracked or overwritten in the creation of the final yoshi flap

[tranquil venture]

tough luck sadly

I've had my contributions to the ends of runs be overwritten tens of times

just happens

[indigo creek]

i mean yeah i agree with that, i'm not saying we should be crediting those people or anything

but you're saying that the only people who put in hard work are the final authors and that's just not fair

anyway i've got all my points out, and you are set on this being unlisted, so let's just stop this here and move on with getting that uploaded

[tranquil venture]

I'm happy to upload the flap unlisted to my channel

[small scroll]

rsl was found to use the wrong character and miss a strat before upload. i was motivated to work on it to redo, but the slower run was forced to be uploaded anyway. pissed me off badly, but i understood that's how the community operates.
bc flap was found to use the wrong character and have bad cycles before upload. zero people have been motivated to redo it. so why the fuck is this run not forced to be uploaded too?

[tranquil venture]

so the real answer is that you would personally want an unlisted/no upload, but you just want this uploaded out of consistency so that it makes you feel better about rsl

[indigo creek]

i think that but instead of spite it's just consistency

[tranquil venture]

and very convienient that you miss the screenshot i sent right above kierio's message

https://cdn.discordapp.com/attachments/1319937435152416768/1411895752241844315/image0.jpg?ex=68b651e3&is=68b50063&hm=603e3a9595e405d9f6f0b67c0db7369c2ed4438c8a6cc3bf236570521682265d&

this was about GV2 NG

[indigo creek]

it wasn't convenient you're missing his point

[small scroll]

i'm not going to argue with you if this is SOMEHOW what you gather from my message

have fun

kierio gets it

[indigo creek]

it's hypocritical that people were up in arms about rsl if this is what happens on bc

[tranquil venture]

it's only hypocritical if it's the same people involved

[indigo creek]

i mean it is, you're involved

iirc you were involved with rsl

[tranquil venture]

was i forcing corvid to upload rsl?

as far as i remember it was campbell

[indigo creek]

well no but you were one of the people who were advocating on the pro-upload side

[tranquil venture]

i went back and read chat and it was corvid arguing with citri and campbell

i was neutral

i said this to try and diffuse the argument

[indigo creek]

alright fair enough

i think corvid still has a point that it's not fair he can be pressured into having that tas uploaded if the same isn't going to happen here

it's been the general consensus that uploading rsl was the right thing to do so uploading bc should be equivalent

[tranquil venture]

isn't the rsl video unlisted anyway

[indigo creek]

is it? i dont even know

[tranquil venture]

yeah it is

[indigo creek]

i never wanted to watch it after the encode was ruined

citri also uploaded it so i was never recommended it

[tranquil venture]

blame campbell and citri for that

[indigo creek]

but yeah if rsl was unlisted i'd say there's not much different to rn

citri uploading it was practically unlisted anyway

[tranquil venture]

if anything i don't get why corvid apologised for uploading something unoptimised, but is angry that we don't want to upload something unoptimised now

why aren't we allowed to

[indigo creek]

he's not angry that you dont want to upload something unoptimised now

if he were an author he wouldn't want to upload it either

[tranquil venture]

corvid is an author

[indigo creek]

wait really

[tranquil venture]

yes

[indigo creek]

ok then whatever if he was the main author

point is he wouldnt want to upload it if he was the one uploading

he's angry that this same level of support for not uploading wasn't around when rsl happened

anyone who was around for rsl who was actively arguing was in favour of uploading except corvid

[tranquil venture]

if he's not angry at me then he should be civil and not call my opinions about BC bullshit

[indigo creek]

when did he say that

[indigo creek]

pretty sure thats just him trying to lead you to the same conclusion as what he wanted for rsl

"if you think this is unoptimised and dont want people to see the yoshi run, why not just not upload it at all?"

[tranquil venture]

ok so now what does he want to happen. i understand that he's frustrated because of rsl, but does he want every single similar, future scenario to have the same outcome as rsl now?

[indigo creek]

he just wants equal treatment

[tranquil venture]

because people enjoy the documentation of runs on the BKT site

but corvid has openly said that he thinks documentation is stupid so he would disagree with that as well

[tranquil venture]

through the means of us doing what

apologising? uploading BC public to make him feel better?

[indigo creek]

if what the community wants is unoptimised runs in these scenarios published then every unoptimised run in these scenarios should be published
if what the community wants is for them not to be published, then he feels that rsl should not have gone the way it did, and i would imagine an apology would be appreciated

[tranquil venture]

why do we keep talking about the community

why can't it just be individual preference!!

if this was tactical and shadow's project, and i was just someone who saved time on the last turn, i would be ok with them uploading it public

it's all relative standards

[indigo creek]

unrelated but fwiw people's individual preferences can influence new tasers' preferences, and thus after long enough could become community-wide preferences

hypothetically of course

so it's not like a few people having individual preferences is meaningless

if it could influence other people then care should be taken

[tranquil venture]

sigh

ok, EVEN if that was some side consequence that had a chance of happening, it shouldn't be a reason for people not being allowed to have their own preferences and standards

[indigo creek]

yeah im not saying you cant have preferences and standards but it's important to know that your preferences can affect other people so just saying "individual preference so it doesnt matter" doesnt work

this conversation is going nowhere can i go now

[tranquil venture]

well i appreciate you speaking on behalf of @small scroll but neither of us really know if these are the options he wants

so yes conversation over for now

[small scroll]

idc to backread just make a fucking choice for the future

i hate inconsistency

[tranquil venture]

i get that you want a standardised choice but i'd much prefer we just have individual decisions based on what authors want

[tranquil venture]

maybe before tas projects are in full swing, the main authors can jot down their standards in the thread so that other people are aware of what their intentions are for upload

[indigo creek]

just btw following author intent goes against wanting full documentation of bkt history

the two are not compatible

[small scroll]

stop pinging me

[indigo creek]

thats why we have massive holes in ur categories' histories (monster doesnt release a lot of his tases)

[tranquil venture]

i don't have time to ask why tbh

[indigo creek]

me or corvid

[tranquil venture]

u

corvid's cracked it i think so i don't think he'll be around in the next couple days

[indigo creek]

huh?

[tranquil venture]

he left the thread

[indigo creek]

okay

he just doesnt wanna be notified

fair enough

[indigo creek]

• some authors have certain standards for what is considered "publishable", which means you can end up with optimised tases that are never published provided the authors are dissatisfied with it enough. depending on the author and depending on how the authors feel on the day, a better tas could be less publishable than a worse one.
• full documentation of bkt history relies on some sort of consistency for what counts as a separate entry in the progression of a category, and going by "what the authors want" is inherently inconsistent. we have been going by author intent on the site, i.e. we don't forcefully add tases that authors don't want on the site if they aren't happy with the tas, which has resulted in several strat changes which have been properly TASed to go undocumented on the site

just in defense of going by author intent: it has allowed us to avoid having to make executive decisions on whether an unlisted or unpublished TAS deserved to be put on the site, since if it wasn't finished and sent off to be verified, it automatically won't end up on the site, making the decision fall on the author

[soft hare]

bruh

i'm all up for transparency and unhoarding but i'm not sure why this is a problem now of all times

like idk every run is a different thing with different authors and choices

personally i'm all for not rushing stuff and uploading minor runs unlisted so bkt history is preserved but i don't get why a single strict rule has to be enforced when there are valid arguments for both options

since there'll always be debates just let each main author/uploader have it their own way

[mint mesa]

No I want it uploaded personally

[mint mesa]

this is why I think having it unlisted would be the most "diplomatic" way to handle it

Assuming Monster is still the main author, he has inputs in the whole run so I did.

[dull ridge]

by main author do we mean who did the most important stuff (e.g. the ultra) or who has the most inputs

[soft hare]

most inputs doesn't necessarily translate into most effort put so i'd say the former

[dull ridge]

yeah same

[real vine]

I think whatever the strat, Monster deserves to be the one to upload it

[earnest light]

Yeah but he has a long backlog of videos he hasn’t uploaded

[mint mesa]

@forest walrus Can we hear your thoughts on this please?
to recap:
Since you and me pulled off the fast version of this ultra with Daisy, this category is gonna be using Daisy too and I think you as the main author will want to upload the Daisy run. It's probably safe to assume that that upload is still a very long time away, which is why we've been discussing if we want to upload the Yoshi run in the meantime (public or unlisted?).
Note that the Yoshi run is not fully optimised, Aoiro's faster strat on the ending shortcut would need us to delay the ultra again by 3 more frames for the geyser cycles to be optimal again. Especially with the imminent character switch I don't think anyone wants to do that tho, so the .241 is likely the final time with Yoshi.

[forest walrus]

I think yoshi should be uploaded public if nobady wants to touch it again.
don't care who is uploading yoshi or daisy runs

[mint mesa]

sounds good to me!

it's my first BKT contribution so I'd like to upload this myself if that's fine with everyone

@tranquil venture @gentle kelp final verdict?

[tranquil venture]

for sure, you put so many hours into the supergrind

can you enable this gecko code?:

04257E64 30000000

no blur background code, ntsc-u

[gentle kelp]

yeah sure

[dull ridge]

what are the authors again

things have been backtracked so much idek

[woeful basin]

I could record if you want