利用者:Sebbas/GSoC 2017/Reports/Week 1

提供: wiki
移動先: 案内検索

Week 1

Part one of my particles goal for this GSoC is to get FLIP particles into the Blender viewport. This week I worked on the data exchange and wrote some API functions.

What's been done

  • There is now a helper function to read particle data from Mantaflow .uni files. I approached it the same way I did when creating the mesh reading function (.bobj.gz files). The load() functions from Mantaflow were also very useful here. (rB8810490a2c23)
  • Several generic API functions needed to be written to connect C code (from pointcache.c and particle_system.c) with C++ code (FLUID.cpp which in turn calls the Mantaflow Python scripts) (rB9cb43962a7a8)
  • Inside the particle system code there is a function that currently reads Elbeem particle data into an actual particle system. For Mantaflow particles we need that as well and so I wrote a new wrapper for that. (rB5468cf836efb)
  • All in all, what you can do now is toggle "FLIP particles" in the cache options and it will save particle data instead of mesh data. Displaying them is still a TODO however.
Preview of the new option for FLIP particles in the cache tab.

Next up

  • Since I am now able to save and load Mantflow FLIP particle data, the next step is to get all the particles actually show up in the viewport. When I saw that the particle import/export worked, I internally hoped that they somehow would just appear. Yeah, that was opportunistic. It couldn't be that easy and so I'll look into displaying FLIP particles next week ;)
  • Next Friday I have meeting with Nils. Hopefully FLIP particles will work by then so that we can discuss secondary particles (tracer, drop, float).
  • Also, I think finishing all particles (FLIP and secondary particles) and then moving on to fluid guiding makes more sense. So basically what I'll do in my schedule is swap "secondary particles" and "fluid guiding" (TBC).

Questions

  • No