Profpatsch
Profpatsch
In general, dhall doesn’t aim to support all use-cases, but aims to give the user power to define their own semantics layer. There’s multiple ways a `lowercase` function could be...
`with` in nix breaks lots of things (and has confusing shadowing semantics), I don’t know how it would be in dhall. But in general imports that don’t explicitely bring symbols...
Same for `./share/man/dhall.1`
I’d like this to be possible as well. It looked strange to me that `{ a = Integer } : { a : Type }` but `{ a : Integer...
This is a particularly aggravating issue when you have a long file, example I had this typo:  on line 1238, and it caused this error message: ``` > dhall...
Morte … I just now realized that the Planescape naming scheme is older than Dhall :)
It seems the parser is backtracking too much (I’m bitten by this a lot when writing any “serious” dhall expressions). Can the earley parser handle that correctly? If so, what’s...
Can we make this general enough to not be only useful in some corner cases? Stuff like > we would, at typechecking time, evaluate the argument of fromMap, and verify...
I would argue that there are two parts to this feature, the internal and the external part. I agree that using numbered variables in the normalization pipeline is a huge...
I see. I’d like to see e.g. user tooling discourage the manual use of indexing nonetheless, encouraging the user to explicitely re-name variables at definition site instead.