Closed ZacLN closed 3 years ago
Just out of curiosity: how does this actually work? If we want to propagate return types, don't we need to know for which input types? Are we just using the same logic that precompile is using?
This is just turning on the get_return_types
option properly - making sure it gets passed to everything as we cache. Nothing fancy actually relating to the inference
We will want to turn this on once we have remote indexing up and running.