Draft Vehicles Frozen in Place

Client

Using canary? Any

Server

Artifact version: Any

Incident

Summary: Draft vehicles become randomly frozen in place with their attached horses running in place typically. This is most commonly seen in towns during a time of day with increased draft vehicle spawns. One or more draft vehicles will become frozen and then hold up all other traffic. They are unable to be unfrozen via any traditional method. However, I just found a method to control the behavior in an unexpected way.

Expected behavior: Draft vehicles shouldn’t be frozen when relevant or in the focus area
Actual behavior: Draft vehicles become frozen when relevant or in the focus area - actively seen and approachable

Steps to reproduce:
Can be done with one player:

  1. Go to a town like Valentine or Blackwater midday and wait for a draft vehicle to become frozen. It usually doesn’t take very long. You can just move about the area or remain stationary and eventually you’ll notice one frozen or traffic backing up.
  2. Once you are looking at a draft vehicle that is not moving, but clearly should be, run SetFocusPosAndVel() (normally would pass parameters, but this should just default to the world origin, well out of the area)
  3. The LOD should drop and the broken draft vehicle that was incorrectly frozen now begins to move as if it were relevant again.
  4. Run ClearFocus() and the LOD will return and the broken draft vehicle will freeze again.

Therefore it certainly seems that the issue is focus/relevancy related and maybe this ability to toggle it will help track down the source of the bug. Figure it can’t hurt to report :smile:

4 Likes

Thanks for yet another great report. I’ve seen it before and tried to fix this issue. Going to note down some stuff here.

The dependence of focus area for this bug seems to be an important addition to the information we already had. Also worth to mention that this issue can happen even when there’s only a single player (i.e. local one).

According to my earlier tests, this bug first appeared after this commit:

So far I didn’t succeed in finding what’s going on, after a few attempts. The code in the commit seems to be correct, according to what R* is doing in their code. Going to try again with a fresh head soon.

2 Likes

Following up here to see if there has been any success triaging this!