Render thread blocking with larger schematics about objtoschematic HOT 4 CLOSED

Once the /refactor branch has been merged, improving the UI responsiveness is my next goal. Currently, the work of the program is performed in the same thread as the UI so when heavy work is performed it blocks the UI from being responsive.

In addition, complicated meshes with thousands of triangles will always be a performance issue, I have plans to add the ability to simplify a mesh to lower the number of triangles in it and hence reducing the time to voxelise the mesh.

from objtoschematic.

It's possible I'm reporting the wrong thing, but mostly the issue with this is that it's not just the rendering that crashes-- it will be attempting to export the schematic but hangs there and never actually spits out the file. I've found that this isn't usually just the case with large objects but rather specific meshes that have weird parts to them.

If it were just the UI then it wouldn't be as much of an issue because I could just reload it.

from objtoschematic.

This should be resolved once /worker is finished and merged.

from objtoschematic.

Now all meshes no longer use the old slow RenderBuffer code. On my PC, a mesh of 3 million triangles is loaded in 2.5s since c58a03e, whereas previously it would crash after 10s. This whole process still runs in the same process and still blocks the render thread which is not ideal but that's what /worker will fix but is a large refactor.

from objtoschematic.