Thank you for reporting this. This has been in the back of my mind for a while now, but I never got around to it.

I've spent a silly amount of time trying to make sense of these joint gizmos, ever since I first implemented them, which I realize in hindsight was largely a result of me getting tripped up by the way that single-body joints work in Godot Physics, which I for some reason felt the need to accommodate for (over two-body joints).

With this behavior having changed in #725 I should rightfully have changed the gizmo accordingly, but forgot to.

Anyway, this should be fixed when #862 is merged. It now seems to line up nicely with Jolt's own debug visualization, as well as Nvidia Omniverse, which happens to use a very similar visualization in its USD Composer editor.

extremely cool and swift execution. nicely done.

just compiled the master - is this the intended display of the gizmos?

i.e. intuitively I would expect the gizmo to show the range of limits in the direction of the effect, but now it looks like its showing it on the opposite side.

might be an issue with negative values - I just checked and it looks good with limits above 0°

edit: nope, still weird.

Would you mind sharing the scene/project? It's hard to tell from the screenshots how you've set it up. It seemed to make sense in the scene I tried it with.

Oh, sorry, I had completely glossed over the original repro. I'd had this in my backlog for so long I didn't even bother to check it, so thank you for double-checking the fix.

But yeah, it definitely looks weird/unintuitive in that scene, and now I'm having trouble recalling the setup I had where it made sense to me.

Anyway, I'll need to take another look at this, clearly. I'm very curious what the gizmos in Nvidia Omniverse's USD Composer looks like with this setup.

I'm pretty sure all we need is to add or substract half PI from the gizmo angle when calculating the points for it. However while trying this I don't see any effect on the gizmo drawing at all. I must be missing something crucial about the whole plugin build. I had a path error and therefore some problems with testing my code change. back on it.

sorry, bit pressed for time so I'll be sloppy.
basically you need to revert your change from #862, it's doing the wrong thing
instead line 83 in the jolt joint gizmo .cpp should be something like const float angle = p_limit_lower + angle_step * (float)p_index - Mathf_PI / 2.0;
I also now noticed that the limits must be in range -180° to 180°, which propably follows naturally. It's just the UI doesn't limit this and maybe should.

see suggested fix in #872

Yes, sorry, I saw it. I've been somewhat strapped for time when it comes to this extension in general lately, and this ended up pretty far down on my priority list when you in the PR mentioned that it wasn't a complete solution.

Frankly I'm becoming more tempted to just copy-paste Godot's own (equally strange) gizmo and call it a day. The fact that I deviated to begin with was somewhat questionable, given that the substitute joints are largely meant to just augment the functionality of the built-in ones. I figure if someone wants it to be different it should rightfully be changed/fixed in Godot itself as well in any case.

Thank you for the PR though. I'll see what I end up doing about this. I'll go ahead and close your PR for now though.

With this extension now having been merged into Godot in the form of an engine module (see README), this extension will be considered to be in maintenance mode going forward, where only bug fixes will be considered for merging.

While this issue could still be considered a bug, it's highly unlikely that I'll spend any time addressing this, so I'll go ahead and close this as wontfix.