André Videla
André Videla
With #3450 I started moving some of the logic that the parser performs into the desugarer, with three goals in mind: - Create a clear separation between _surface syntax_ and...
I think you got it but just for completeness here is what's going on with linearity 1 There are two modes: 1. Match on the constructor and use any variables...
> Do we agree that the current behaviour that moves unrestricted types to erased types through a @-pattern is undesirable? at first glance it looks like a bug to me...
good point, thanks for bringing attention to it
This is fixed on current HEAD
Nope, not fixed actually
@joelberkeley Thanks for your comments. For your concerns about 2 and 3, please refer to the "Scope" section: > The syntax sugar does not help with indexing dictionaries, dictionary update,...
I think the pairing operator is a very strong alternative, and it has two great benefits over this proposal: - It does not require changing the language - It already...
I've updated the PR template with information for maintainers and beginners. We can't force people to read, but we can hope
yes this is worse, i'll benchmark my huge PR that refactors the trees using this and come back with a concrete number to measure the difference. I didn't notice anything...