Full Changelog: v0.16-beta...v0.17-beta
- MULTITHREADING! YAY!! But it's done in the easiest way possible: Launch N amount of sparkles.pyw and divide number of particles by number of threads. Hehe. Why didn't I come up with that earlier?:
- Performance now SOMETIMES multiplied by N of threads! SOMETIMES = Low amount of particles: 2 threads best; High amount: 4 best. Maybe the Windows desktop window manager craps on it?
- Can now spawn particles in the neighborhood of 300 - per frame - other with no FPS reduction. (Yea, I still don't understand time dilation. I'm not a physicist.)
- Performance is now only limited by how big the rectangle is that is being updated every frame. Need to find some way to lower area, but not with a huge amount of rectangles. Spacial hashes?
- Finally explained what is going on with each setting: There are two vectors. Sometimes they're added. You can manipulate them
- Sparkles now with rotation! Random or controlled, bilateral and for both vector.
- The strength of all slider and other options is now somewhat normalized and within practical bounds.
- Interpolation now no longer leaves out one "slot" of 1/3s of "moved distance" ... It's much more smooth looking!
- Velocity limit not global and not hard. It's soft! Flacid even. No need to harden it, too: The particles rub it juuust right. :P
- Internal: Variables now have much better names
- Should be all.
Current issue:
- My multithreading solution causes n subprocesses of sparkles to spawn. This happens too fast when it's all compiled to a executable. Therefore I had to add a 1.5 second delay between launches of the sparkles subprocesses.
- This is not an issue when run from code! (.pyw file)