(12-02-2019, 06:44 PM)StudioTatsu Wrote: Over the past 2 weeks, I have really started to enjoy all of the obi physics products, and working them into my pipeline has been great.
So, I wanted to ask, is there a roadmap? What do we have to look forward to?
A few things I would hope to see in the future.
- Fracturing and slicing of Meshes. Runtime Destruction.
- Softbody Tearing (similar to cloth tearing)
- Particle Editor with Manual particle positioning override
- Ability to use translate tools in editor while in playmode on gameobjects with softbodies and cloth components
Not expecting any of these anytime soon - I just wanted to add to the wishlist.
Thanks. Looking forward to the future.
Hi there,
Thanks for your words! Being a one-man (sometimes two-man) team makes things progress slowly, but I do my best
. Currently there's several development fronts open:
- Developing a toy version of Obi using ECS. The objective is to rewrite the underlying engine in C#, for faster iteration times and instant support of all platforms. Still in the "figure out how to achieve certain things" phase, and evaluating performance. Basic XPBD constraints are almost done, the hard part is efficient collision detection/response.
- Better editor tools for cloth and soft bodies. Better particle selection using raycasting, ability to move/create/delete particles, and create/delete constraints. This would allow for manual placement of soft body clusters.
- Obi Plant. Too soon to talk much about it, but basically an interactive way to generate tress and climbing plants, that can either be used as-is (simulated at runtime), or baked to static meshes.
About the proposed points:
- Softbody tearing/fracture is definitely something we want to have, but remeshing seems a tad complex/messy. It is manageable in 2D (for cloth) but for softbodies it has to be volumetric, which adds another layer of complexity. Still researching and figuring out the simplest/fastest way to approach it.
- Using the translate tool in editor is already possible as long as any particle is fixed to the object's transform. Or, if you use local space simulation, you can transform the solver itself. If there is no fixed particle and you're in world space mode, the object will obey the laws of physics and it is not possible to move it using translation gizmos. Specially in the case of cloth, as it does not have a "center" that we could position/rotate/scale around. Softbodies would be easier since they have a rigid rest shape that we can use, but still...
yes, please. lol