"Sleeper" physics issues

Hi, so I was told by a Roblox staff member that an issue I was having is “not sleep related”

Here’s a video of the bug:
https://gyazo.com/02214d45953e9e0c85a5b894f790214e

This rarely happens, and supposedly should’ve been fixed with Roblox’ latest update…but the person I spoke with told me to come to the Help and Feedback section; for whatever reason…

I can provide more information if necessary, this has been happening in my game for as long as I can remember. One detail to note is that my whole game is rendered client-sided only, for best physics possible.

5 Likes

Maybe it’s an issue with mass? Try making the pins massless

Edit: or try changing the CustomPhysicalProperties of the bowling balls and increase the density property (which determines mass).

1 Like

This actually doesn’t have an effect unless it’s part of an assembly with a non-massless root part. Instead, use CustomPhysicalProperties.

1 Like

Have you checked if it is welded to the surface. Roblox has a auto joint feature now. Thats my only guess. These threads can help you.

You can use these functions to check if they are welded/anchored:

BasePart:IsGrounded() 

Or

BasePart:CanSetNetworkOwnership() 
1 Like

…that is disabled by default. It used to be enabled but it’s not anymore. You have to manually enable it.

2 Likes

Hi, I’d like to investigate this further. I will need you to reply (or update the post) following the format outlined here: How to post a Bug Report.
I don’t have enough information to know what might be going on or how to debug it.

1 Like

Hi, I’ve actually already attempted to file a bug report, but because the repro file I attached “did not replicate the issue” my post has seemingly been on standby; waiting for another reply from a post approval member.

I’d love for you guys to look into this issue though. Please let me know if you want me to do something.

I’m not sure if you’re able to access this post, but here’s the link to my bug report:
https://devforum.roblox.com/t/engine-bugs-physics-sleep-still-an-issue/482765

Yea I can’t access it that post. You can just reply with the info from it here.

You will need to make sure that your repro file can reproduce the issue, or there isn’t much that can be done.

A quick hack I’ve found to help is to manually set the part’s Velocity to a small value beforehand. When I did this, I was able to “awaken” the part.

5 Likes

Hey there!

Not entirely sure this is the same bug but it seems similar

Previously (when the bug did not occur) I was instantiating parts on the server and having them not anchored from the get-go. However, this caused my game to have some bad performance issues as I was creating and removing a lot of parts constantly.

I got an idea to have the parts be anchored on the server and have the clients remove the anchor once they replicate, that way the server does not have to deal with all of the physics.

Which lead me to have this issue, once the parts spawn, they remain in the air even though their “Anchored” property is set to false, I’ve also noticed that once my character comes closer to the part physics wakes back up.

My guess is that because of how the Network Ownership works, the parts’ physics owner is the server until the player gets close, which then it switches… Now that’s great and all, but it does have some edge cases like this.

I would give physics ownership to the player but due to the nature of the game, any given player can be around that said part…

I’ve asked some friends if they can replicate this issue a few days ago, and @1TheNoobestNoob got a gif of it!

3 Likes

Unfortunately, since Roblox is the weirdest platform I’ve ever worked with before, it hasn’t done it in the repro but it does do it rarely in my main game, even though literally everything is as close as possible. Ball physical properties, pin models are exactly the same, model is created client-sided entirely, same ball size, literally everything.

No idea why this is such rocket science to investigate. There’s dozens of other reports that signify that the physics solver is still performing poorly.

I’ve approved the topic and included as much information from the discussion in Post Approval as seemed useful.

1 Like