Setting things up

[almarklein]

Setting up

• Use git-filter-repo to clone wgpu-py, but only the gui stuff (wgpu/gui/*.py and tests/test_gui_*.py).
• Create repo at Github 👋 and push the reduced clone.
• Move wgpu/gui to rendercanvas.
• Populate repo with readme, gitginore, pyproject etc.
• Make it work with from rendercanvas import WgpuCanvas
• Add the examples.
• Activate CI.
• Activate RTD.

Moving forward

• Add docs enough so we can do the below.
• See if I can get the tests running without changing them too much.
• Open Refactor wgpu.gui for scheduling wgpu-py#618 as a pr here: Implement scheduling #6
• Rename WgpuCanvas to RenderCanvas.
• Better grouping methods in the backends.
• Fix tests, and activate in CI.
• Get rid of other references of "wgpu" and "gui"
• Make docs work and clean them up.
• Rename and clean up examples.
• Merge the two internal utils modules.
• Use __all__ in the modules.
• Make that Canvas subclasses only need to implement private methods?
• Check todos.
• Open up get_context() so this is not wgpu-only: Refactor inter-context-canvas api #21
• Can we move present() out of CanvasContext()? Nope, and that's ok.
• Remove asyncio.get_event_loop() (its being deprecated).
• Implement gallery.
• First release.

More short term work

• Run examples on CI
• Handle sigint.
• Allow async event handlers.
• Animate events.
• Register the draw function as a draw event handler instead of with request_draw()?
• Still thinking about canvas.add_event_handler() vs canvas.events.connect()
• Allow requesting a draw by setting a field in the event object?
• Store methods as a weakref-to-object, plus method name. See pysignal
• Tracking statistics (track time spent on different things). Plus visualizing them. Maybe later.

Adoption

• Move issues from wgpu-py repo to here.
• Add rendercanvas version to diagnostics of wgpu-py.
• Use rendercanvas downstream.
• Deprecate wgpu.gui.
• Remove wgpu.gui.

[almarklein]