forum

[resolved] [confirmed] [CtB hypers] Ryuuta going out of screen

posted
Total Posts
15
Topic Starter
Drafura
Problem Details:

Ryuuta going out of screen due to the "slipping" effect on extremly fast hyperdashes. Tried many times to catch, it's just impossible imo. Related infos : https://osu.ppy.sh/forum/p/2324454 ?

Video or screenshot showing the problem:
Replay

Map [blood AR10] 00:17:287 (1,2) - the hyper



osu! version: 20140812.4 (latest)
Deif
Easily replicable. The problem appears by minimal distances between notes in the timeline: where the player has to press left to catch the next kick-slider and the HDash gets you to the complete opposite direction + dissapears.

This map is kinda controversial anyway, since the problematic slider is totally unsnapped:


Anyway, tried the same map but the AR9 diff and the same happened to me:

MillhioreF
This has been happening for a while, but the sheer speed and note density required for this means that it doesn't occur in even the most intense ranked map. Not sure if it's worth a confirm just yet.
Loctav

MillhioreF wrote:

This has been happening for a while, but the sheer speed and note density required for this means that it doesn't occur in even the most intense ranked map. Not sure if it's worth a confirm just yet.
I disagree. It might be a low-priority bug, but it's a confirmed unwanted behavior. the catcher shouldn't behave like this even in the most edgy cases (especially not disappearing entirely)
We might move this as low-priority since it's not happening in any currently ranked or pending map apart of some edgy exceptions, but it's worth confirming imo.

Marking as such.
Shivarion
ryuuta go picnic LOL
Kitokofox
This has been around ever since the hyperdashes used the "arc" movement. When they go straight, it's actually very very easy to get through a stream of hypers, like it should be.

This seems to happen when the game attempts to estimate a fast movement with a following slow movement for the arc, but because it's too short, it ends up over-estimating the distance because it ends up eliminating the second movement since it's too short.. In some cases this can happen in a frame, sending ryuuta/yuzu off the screen as a result. You can mediate this a little by walking or stopping through some hypers but that's limited and very difficult.

This didn't happen with the hyperdash behaviour that went in a one set speed to the next fruit, letting go upon reaching the next one. Since there is one speed and one distance to cover, even if the movement happens in the next frame, you'll still land under the next fruit.

http://puu.sh/byFM9/0ed938472a.osz - This is an example that I had laying around about ticks being hypers. It's not necessarily passable but at least it can help understand the stability of hypers and how they behave.
Kurokami
I already discovered this a year ago. Back then Marcin said this is not a bug.
MillhioreF
I've fixed this on my local dev build, but I need to test it some more before it goes live in case something else broke badly.
peppy
I'm not sure what this "arc" movement you talk about is. Hyperdash velocity is still constant.
peppy
Please test in cutting edge. It should at least behave a bit more sane now.
Kurokami
Still goes out but it comes back immediately to the same position. I think this is fixed now.
Deif
The catcher doesn't go out of the screen anymore (which was the purpose of the thread), but it still takes you to the opposite direction where you're pressing. No idea if that can be fixed, but the main bug was solved.
Kurokami
For me it goes back to the same place. o.o
Topic Starter
Drafura

Deif wrote:

The catcher doesn't go out of the screen anymore (which was the purpose of the thread), but it still takes you to the opposite direction where you're pressing. No idea if that can be fixed, but the main bug was solved.
I have the same behavior here.
MillhioreF
Fixed on latest cuttingedge.
Please sign in to reply.

New reply