Replies: 1 comment 1 reply
-
I haven't given it much thought, but I don't expect it to be through feature cfg, nor by wrapping the API of `winit'. My preference is to let The |
Beta Was this translation helpful? Give feedback.
-
I haven't given it much thought, but I don't expect it to be through feature cfg, nor by wrapping the API of `winit'. My preference is to let The |
Beta Was this translation helpful? Give feedback.
-
Referring to #233
winit has quite a broad API surface. Creating an abstraction would result in a lot winit's public code to be copied to ribir_core and then somehow mapped between ribir_core and winit at runtime.
I also wonder, how easy or difficult it would be to map to a alternative implementations when the abstraction is based on winit.
An alternative I can think of is to annotate access to winit API with feature cfg, but then access to every alternative implementation would need to be prepared in
ribir_core
.The dependency to winit would not be removed entirely, but at least it could be made optional.
I haven't looked at
ribir_gpu
yet.Any thoughts?
Beta Was this translation helpful? Give feedback.
All reactions