возможно

Yeah, correct.

Yeah, tank game. Will see how it goes

It was just a small quick thing, I am still working on the tanks.

It's not a shader, but cpu doing the procedural texture based on angle between faces.

​@@reasonnotfound yes you build an edge texture i know, this allows to build custom normal maps for the edges like an edge shader. Do you plan to make it open source?

Probably not, it has it's problems (highly dependent on uv maps). I might if I figure out a better solution which would be usable also in my little game thing.

Yeah, it was a "couldn't sleep" idea, so it probably won't be used 😅

A lot of lines... youtube compression does not like that apparently :P

Thanks!

Yes, you can feed in start and end points.

Not efficient at all :P

Not yet, but I hope it would turn into one. Now I need a cannon to fire the shells.

​@@reasonnotfoundwhen you model armor i suggest you check out the blog post of the game children of a dead earth titled "raw steel", and check out the replies of the channel "actionscripter"'s videos

😎

Prob cause it's physically attached to the part of the tank that wiggles

Thanks!

Not entirely sure how you mean it. It generates the texture on the cpu, the texture then can be used for whatever. I can even save the generated texture as png and edit it later. So far there are two catches, the model needs to be autosmoothed and the uvs need to be non overlapping. It's not perfect.

@@reasonnotfound I meant saving a drawing of what will be behind the lines. It's not that important, just a little idea.

Maybe, but tanks are a lot of physics stuff. Last time I had quite a trouble getting physics things stay stable, but now there are stuff like jolt and might be interesting to experiment with it a bit. Will see.

@@reasonnotfound "but tanks are a lot of physics stuff. Last time I had quite a trouble getting physics things stay stable" YES, I know it´s hard. That is why Im asking . Simulation of aerodynamics (planes) is much simpler, since in most cases the path of a body in air is described by a smooth continuous curve. That allows to use effective and most importantly fast approximations to calculate the position. They are known for a long time and are extremely simple to calculate. Those. Even simulating realistic train movement on rails is more complicated than simulating an airplane. In contrast, the movement of ground equipment like tanks goes all the time in a collision with ground, i.e. the trajectory has clear points of sharp extremes, so the approximation is difficult and has to be considered a difficult way. This is one of the reasons why tank _simulators_ are many times few than aircraft simulators, and tank simulators WITH working chassis practically do NOT exist. We can also look at the real world and robots. And we see that flying robots (drones) appeared a very long time ago and are widely used everywhere, but robots moving on the ground are an order of magnitude few. And this is all because the laws of movement of ground units and, as a consequence, their control are orders of magnitude more complicated than flying ones. The main difficulty is not that it is necessary to simulate the suspension springs - this is an elementary matter. But the fact that a ground-based object is constantly in collision mode, with the ground / other objects. This creates unpredictable effects that cannot be approximated and complex calculations have to be used to resolve this cases in each tick of physics simulation. But airplanes are not in this mode, but quite the contrary, they fly along a smooth path without extremes. Therefore, their modeling is much simpler. Any even the most significant twists that can occur for an airplane are orders of magnitude smoother than, for example, the ordinary situation of hitting a tank on a vertical wall or a tree. And when they are on the ground, they are also on a flat ground and they basically have three points of contact - the easiest case for ground equipment. And as soon as the aircraft goes beyond these conditions (and becomes like a tank ), in all flight simulators that I know, it behaves extremely unnaturally and incorrectly. Many drones are also fixed wing aircrafts. The equations in both cases are nonlinear, for helicopters/rotors it is also much more complicated. And the equations are everywhere non-linear and for tanks and for planes and even for trains. In the airplane simulators you can apply simplified approximations ( "look-up tables") that will give a realistic result. But in the simulators of ground vehicles, you can not do this - you need to calculate as needed. That is why they are more complicated.

I see. I tried a little test where I tried to simulate whole suspension using rigidbodies (kzbin.info/www/bejne/nYqZZ4Gjl9B3Z8U). Few physics based vehicles with fully simulated suspensions might be possible, but as you said, quite annoying to make stable. There is one guy on youtube who did fully rigidbody tank tracks, which was absolutly insane.

@@reasonnotfound Actually there are 5 (five) games with physics based vehicles. You can check them, T-72 balkans on fire, Steel Armor: Blaze of War, Graviteam Operation Star, Graviteam Mius front and Tank Warfare Tunisia. All available on Steam. BUT they weren´t made with Godot hehe. I wonder if it would posible to port FlightGear to Godot. Since planes are easier than tanks. And also DaggerFall Godot lol. BTW the light tank you modelled appears in Tank Warfare Tunisia.

Not sure, I don't have the project files anymore. I had more then 10 of these use rigidbody for each moveable part, where parts were connected with constraint joints. I would be surprised if raycasted solution would struggle with performance.

It's based on plane equation and pythagoras I think. I solved it a long time ago so I don't remember the exact solution.

I had information about proximity to the hole, so I moved vertexes in the direction of their normals.

that is brilliant! thank you! @@reasonnotfound

;)