how this can effect the performance ? , this is many constraints !
RARE W, they are open to supporting R6 in an R15 feature? I wonder if they went back on deprecating it
People are gonna use it anyway, so it would make sense to add support.
I noticed that emote animations don’t move the character away from the center. Is that intended?
Speaking of animations, when will we have root motion? I’m an animator, and this new update seems like it could be used to implement root motion since it’s not moving the players in the emotes that are supposed to move.
is Ragdoll humanoidstate will be un deprecated ?
holy peak
This a huge update, with perfect timing too. Was just about to start a project that’s going to use physically simulated anims like this.
There’s little to no impact on performance.
- While the AnimationConstraints are in kinematic mode the performance is actually slightly better than Motor6Ds. The additional instances will not add any runtime cost in kinematic mode.
- When physically simulated (
IsKinematic
is false), the cost is the same as any other traditionally made ragdoll character. Generally, this is fine on all platforms as each client simulates only their own character. - The additional instances only increase the average avatar’s memory usage by 2%.
- Our tests showed this has no impact on join times or out of memory crashes.
YEEEEESSS YES YES YES LETS GO I’ve been waiting for this since RDC I am so excited. This is going to drive all of my characters ASAP, so much emergent gameplay comes from physics toys.
What’s the timeline for being able to turn this on in live? I want this ASAP.
Emergent gameplay designer chads keep winning
@Homeomorph
Tangentially related:
Since the avatar constraints use the motor6d’s positions, that means some avatar packages with incorrect / broken motor6d’s will have physics problems when animating. With the package I use on my avatar, various limbs will straight up detach from the body with some types of animations!
I filed a bug report about this 2 years ago, but I haven’t been able to get it actioned on. Now that joints are going to become physically simulated instead of CFramed, would it be possible to get this looked at?
Question about niche humanoid behavior. In our platformer game, we directly modify the humanoid rootpart’s root motor6d C0
rotation every frame to rotate based on the slope the avatar is standing on. This essentially rotates the avatar on sloped surfaces, making them stay “flat” relative to what they are standing on. One thing I’ve observed with this behavior, is that doing this does not affect physics in any way. The avatar visually rotates, but does not physically rotate. Jumping etc still works as expected as if the avatar was upright.
Will switching to these new animation constraints change that behavior at all? If so, how can we get that effect back where we visually rotate the character, without causing all kinds of physics sideeffects? The slope tilt is mean to just be a visual polish, not something that impacts the avatar’s ability to move.
This is great! Opens up many possibilities. Like this floppy guy:
Self View completely breaks with this feature enabled lol
This is amazing!
As smooth physics interpolation was confirmed in the test game, will it include some kind of physics network prediction?
This unironically the best physics based animation system I’ve seen in any game engine.
I’m already liking this update!
this is very cool ! , W update Roblox !
Please do, it is a quality of life update for R6 that would greatly improve experiences that use it.
Is the replicating towards the client? because sadly bones dont replicate from server to client i hope this one does
REALLY???
Amazing update, will be using this once it’s out of beta!!