Right now we create a single type variable placeholder for these things, but we could infer their structure at least similar to what we do for function declarations. This would allow us to refer to an object properties or a tuples elements before the object/tuple has been fully inferred.
Right now we create a single type variable placeholder for these things, but we could infer their structure at least similar to what we do for function declarations. This would allow us to refer to an object properties or a tuples elements before the object/tuple has been fully inferred.