[Bug Report] DAS preservation pivot bug (with video examples)

[PetitPrince]

EDIT: topic renamed from "[Bug Report] SOCD behavior when using DAS (with video examples)" (thanks Kitaru)

Hello,

As mentioned by other when holding a direction button and pressing at the same time the other direction, DAS is not reset and we can fling the piece directly to the other side.

This happens both in Standard and TGM mode.

I believe this is a bug because this behavior is not symmetrical: for me it only happens when doing right then left, and not the other way around.

I've recorded those short videos to show you how to reproduce this behavior, please have a look:

https://youtu.be/DPCt9bVBUj4

https://youtu.be/OU6fH5_R7QQ

https://youtu.be/0ZHywm3FabE

[Ahruna]

For me, it's happening when moving in either direction. I'm using a keyboard. Here's a recording of it happening:

https://www.youtube.com/watch?v=bsSwoJQj9gY

[Kitaru]

This bug isn't exactly related to keyboard SOCD, although it makes it even easier to activate. A frame-perfect pivot on joystick also activates this DAS preservation bug.

[PetitPrince]

I confirm that this behavior was addressed in the patch but is not entirely gone. It's way less likely to activate it, but still doable if you do a frame perfect change of direction.

For reference, here's a video where I can trigger it several time:

https://www.youtube.com/watch?v=YxpyvmiGFPQ

I assume this will be discussed and eventually addressed in future updates; I just wanted to have a video reference somewhere.

[zaphod]

For coding a fix, make it so activating hard drop, soft drop, move, left, or move right actually resets the das counter, instead of just resetting on neutral between left and right. Decrease the das counter by one frame if needed to account for this.

[Kitaru]

Although it seems this item is under consideration for a configuration option ("charge cancellation continuation?"), I wanted to go through one more time to fully and clearly document this behavior and the potential issues it presents.

Action Performed: The player quickly reverses the lever from left to right, or from right to left.
Expected/Desired Behavior: The tetromino moves exactly one space in the new direction.
Actual Behavior: The tetromino responds in one of two ways:

• The tetromino does not move in the new direction at all. This occurs if movement charge for the old direction had not yet accumulated.
  
• The tetromino moves in the new direction many spaces very quickly. This occurs if movement in the old direction was fully charged.

Video clip of expected/desired behavior, using TGM2 as an example:
https://www.youtube.com/watch?v=Fw6JISah_OA

Video clips of actual behavior in TGM4, without charge accumulated (zero movement):
https://www.youtube.com/watch?v=Z22F0U7VZdw
https://www.twitch.tv/kirby703/clip/FragileTemperedKoupreyDoritosChip-MISCMU7x-18Cucmu

Video clip of actual behavior in TGM4, with charge accumulated (rapid movement):
https://www.youtube.com/watch?v=tDEBTqMcOGQ

---

Additionally, there is a subtle behavior in TGM1-3 that would be beneficial to preserve, even with a "charge cancellation continuation" feature disabled. Although reversing the joystick in a single frame does not carry over charge with a tetromino actively falling in the playfield, it does allow charge to carry over or continue to accumulate if an upcoming tetromino has not yet appeard in the playfield.

This subtlety is demonstrated in the following slow-motion video clip of TGM2:
https://www.youtube.com/watch?v=TtB_-PAkZJo

Thank you to Arika for their time and consideration of these reports.