Fix hints on PhysicsMaterial bounce/friction. #42203

rcorre · 2020-09-19T22:46:24Z

These values are only meaningful in the range 0 to 1.
Make sure the editor enforces reasonable values.
Fixes #42202.

These values are only meaningful in the range 0 to 1. Make sure the editor enforces reasonable values. Fixes godotengine#42202.

akien-mga · 2020-09-20T06:56:08Z

Are you sure that the values outside 0..1 can't be used for non-physically correct behaviours? Like bounce > 1 to increase bounce height each time (breaks energy conservation, but we're making games :)).

rcorre · 2020-09-20T12:52:31Z

Interesting, I'll have to try that 😆
FWIW I stole the hints from PhysicalBone. Whatever the contraints are (even if there are none), should those two be in sync?

rcorre · 2020-09-20T13:25:20Z

Unfortunately values greater than zero don't seem to make a difference (at least in Godot 3.2 with Bullet Physics).

Xrayez · 2020-09-20T14:09:22Z

scene/resources/physics_material.cpp

 ADD_PROPERTY(PropertyInfo(Variant::BOOL, "rough"), "set_rough", "is_rough");
- ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce"), "set_bounce", "get_bounce");
+ ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce", PROPERTY_HINT_RANGE, "0,1,0.01"), "set_bounce", "get_bounce");


You can use or_greater hint while still restricting the slider to 0..1 range, which should be common for most use cases anyway, but still allow for specific use cases (for values greater than >1, can be manually typed via keyboard), see #36450 for instance.

Suggested change

ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce", PROPERTY_HINT_RANGE, "0,1,0.01"), "set_bounce", "get_bounce");

ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce", PROPERTY_HINT_RANGE, "0.0,1.0,0.01,or_greater"), "set_bounce", "get_bounce");

Interestingly, the new behavior in master is that you can still drag the slider past the max value, but the slider will be snapped to the specified range instead, this is not the case for Godot 3.2, IIRC.

Good to know! Given that > 1 didn't seem to be meaningful here after all, I think hard limits make sense.

Hmm I guess I just happen to have some physics bodies in 2D with custom physics, it's possible to set values above >1 just like Akien mentioned, so in theory the physics material could be used for non-engine classes, since it's agnostic to the physics backend used. It's a small nuance, but that's something which I'd do myself, but the engine development philosophy is that everything must be kept simple to cater the most common use cases, so I can only comply. 😛

The physics material itself probably just clamps the value to the range of 0..1 anyway.

No, it doesn't 🎉:

godot/scene/resources/physics_material.cpp

Lines 62 to 65 in 9a463cb

void PhysicsMaterial::set_bounce(real_t p_val) {

bounce = p_val;

emit_changed();

}

Makes sense. Maybe or_greater would be appropriate here then. I'll leave that to you to PR if you think its useful though :)

The physics engines do clamp the calculated bounce to 0..1:

body_pair_2d_sw.cpp 214:real_t combine_bounce(Body2DSW *A, Body2DSW *B) { 215: return CLAMP(A->get_bounce() + B->get_bounce(), 0, 1);

(at least in 2D)

Sure you can set a negative bounce on one body of -5.0 and the other one +5.5 and they should amount to +0.5 and be within bounds... But why? We don't cater to undefined use cases, and if there's no implemented physics engine that actually supports values outside 0..1 (as in, it's designed in a way where those values actually make sense, not that they can incidentally produce an outcome which is not outright a bug), there's no reason to future proof.

If you have a very non-standard use case, you can use a script to set those values to arbitrary values. Or create a UnlimitedPhysicsMaterial resource in a script :)

We can discuss this in https://github.com/godotengine/godot-proposals, but I don't see the point in allowing values that don't make sense currently in the engine.

Thanks, I'm changing CLAMP to 0, 10 in my fork...

As mentioned, if you have a use case for unclamped bounce or friction, please open a proposal on https://github.com/godotengine/godot-proposals

This PR did not restrict anything further than what the codebase already does - even if you could input 10 before this change in the inspector, the engine would clamp it to 1 in body_pair_2d_sw.cpp as I've demonstrated above.

akien-mga · 2020-09-23T07:07:03Z

Thanks!

akien-mga · 2020-09-24T10:21:43Z

Cherry-picked for 3.2.4.

TetraGenius · 2020-12-02T14:11:04Z

Various physics engines allow the user to use bounce > 1. Maybe a paddle or block that increases speed and difficulty. I use it to simulate gas particles hitting a hot wall (increases their speed/pressure). But if someone can point me to making a UnlimitedPhysicsMaterial resource I'll try that...

TetraGenius · 2020-12-02T14:14:55Z

You can also check superelastic restitution in research.

TetraGenius · 2020-12-09T09:20:21Z

Thanks, I'm changing CLAMP to 0, 10 in my fork...

Fix hints on PhysicsMaterial bounce/friction.

6c18bae

These values are only meaningful in the range 0 to 1. Make sure the editor enforces reasonable values. Fixes godotengine#42202.

Calinou added bug cherrypick:3.x Considered for cherry-picking into a future 3.x release topic:editor topic:physics labels Sep 19, 2020

Calinou added this to the 4.0 milestone Sep 19, 2020

Xrayez reviewed Sep 20, 2020

View reviewed changes

akien-mga merged commit cadba26 into godotengine:master Sep 23, 2020

rcorre deleted the physicsmat_hint branch September 23, 2020 12:23

akien-mga removed the cherrypick:3.x Considered for cherry-picking into a future 3.x release label Sep 24, 2020

This was referenced Apr 30, 2021

Friction in PhysicsMaterial is too low even at 1.0 (the max) #48298

Closed

Revert "Fix hints on PhysicsMaterial bounce/friction." #48321

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix hints on PhysicsMaterial bounce/friction. #42203

Fix hints on PhysicsMaterial bounce/friction. #42203

rcorre commented Sep 19, 2020

akien-mga commented Sep 20, 2020

rcorre commented Sep 20, 2020

rcorre commented Sep 20, 2020

Xrayez Sep 20, 2020

Xrayez Sep 20, 2020

rcorre Sep 23, 2020

Xrayez Sep 23, 2020 •

edited

Loading

Xrayez Sep 23, 2020 •

edited

Loading

rcorre Sep 23, 2020

akien-mga Sep 23, 2020

TetraGenius Dec 9, 2020

akien-mga Dec 9, 2020

akien-mga commented Sep 23, 2020

akien-mga commented Sep 24, 2020

TetraGenius commented Dec 2, 2020

TetraGenius commented Dec 2, 2020

TetraGenius commented Dec 9, 2020

	ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce", PROPERTY_HINT_RANGE, "0,1,0.01"), "set_bounce", "get_bounce");
	ADD_PROPERTY(PropertyInfo(Variant::FLOAT, "bounce", PROPERTY_HINT_RANGE, "0.0,1.0,0.01,or_greater"), "set_bounce", "get_bounce");

	void PhysicsMaterial::set_bounce(real_t p_val) {
	bounce = p_val;
	emit_changed();
	}

Fix hints on PhysicsMaterial bounce/friction. #42203

Fix hints on PhysicsMaterial bounce/friction. #42203

Conversation

rcorre commented Sep 19, 2020

akien-mga commented Sep 20, 2020

rcorre commented Sep 20, 2020

rcorre commented Sep 20, 2020

Xrayez Sep 20, 2020

Choose a reason for hiding this comment

Xrayez Sep 20, 2020

Choose a reason for hiding this comment

rcorre Sep 23, 2020

Choose a reason for hiding this comment

Xrayez Sep 23, 2020 • edited Loading

Choose a reason for hiding this comment

Xrayez Sep 23, 2020 • edited Loading

Choose a reason for hiding this comment

rcorre Sep 23, 2020

Choose a reason for hiding this comment

akien-mga Sep 23, 2020

Choose a reason for hiding this comment

TetraGenius Dec 9, 2020

Choose a reason for hiding this comment

akien-mga Dec 9, 2020

Choose a reason for hiding this comment

akien-mga commented Sep 23, 2020

akien-mga commented Sep 24, 2020

TetraGenius commented Dec 2, 2020

TetraGenius commented Dec 2, 2020

TetraGenius commented Dec 9, 2020

Xrayez Sep 23, 2020 •

edited

Loading

Xrayez Sep 23, 2020 •

edited

Loading