Jonathan Lindegaard Starup

Results 363 comments of Jonathan Lindegaard Starup

`xfuzzer` is basically a test - maybe that should be extracted into a test

`xprinttyper` and `xprintphases` could maybe be bundled into a Debugging/Info option

`XPerfFrontend`, `XPerfPar`, and `XPerfN` are only relevant to one command. We don't really have a way to describe that in Options. But either way, they could be bundled into a...

Yeah I also considered that after posting. hmm 🤔 What about a low tech, desugaring solution. `Someting#{...} @ r` just becomes `Something.fromList(r, ...)`. Magnus has expressed objections to pre-backend desugaring,...

I've implemented `List#{...}` but I still believe that desugaring `MyOwnThing#{...}` to `MyOwnThing.fromList(List#{...})` would be cool

@magnus-madsen I think we have the power to do this and should do it Pros: it simplifies the AST to one base literal (either List or Vector probably) and it...

> how will you handle Map|...|: ??? Is this an ascription? It's just desugared directly `((FromVector.fromVector($Vector$(...)): Map[_, _]): ???)` > will you resolve Map using the local scope? yes

> Just wondering if this will always work out neatly. It works for List, Set, Vector, Array, and Map at least. Which allow us to remove their ad-hoc constructors

> Only Map if we do extra magic for key -> value mappings Yeah isn't it just `Map|('a', 1), ('b', 2)|`? and `Map|1, ,2|` will give some instance error >...

> Yes, this is just arguably a downgrade in clarity. True, but `Map#{'a' => 1, ..}` is also magic syntax not found anywhere else. So maybe the tuples are better...