Simon Jakobi

@TristanCacqueray Thanks for the link! Is this mainly about union equality then? My impression is that the amount of boilerplate saved by using wildcards is only a very small part...

OK, I concede that there are cases (like predicates) where wildcards seem to be the appropriate tool. I still worry that we'll see wildcards used a lot in scenarios where...

> Or, perhaps we could require that all wildcard pattern matches must ignore all but one constructor (e.g. `{ A = x, B = y, _ = z }` is...

> Alternatively: _all_ alternatives of the enum type must be empty (not just the ones being compared) Wouldn't that be necessary anyway, so you can write something like ```dhall \(x...

Can this issue be closed now that we have both `toMap` and [the `JSON` utilities](https://github.com/dhall-lang/dhall-lang/tree/da02ab2faf81b3e627f01cc5f4bb37403a15ab0e/Prelude/JSON) in the Prelude?

I just came across https://github.com/haskellfoundation/volunteering and wondered whether [this volunteer](https://github.com/haskellfoundation/volunteering/issues/3) might be a good fit for this project. Should we ask them?

This is an issue with the `dhall-haskell` implementation, so I've transferred it.

I wonder whether there's a connection to https://github.com/dhall-lang/dhall-haskell/issues/1592 and whether anything can be improved here without radically overhauling the parser.

The reports in #2264 and #2265 are very similar: ``` Error: Invalid input (input):1:14: | 1 | { a = "foo", b = "\e" } | ^ unexpected 'b' expecting...

> I'd be interested to see real-world examples where this is noticeable. Me too! @fizruk linked to a sample in https://github.com/dhall-lang/dhall-lang/issues/552#issuecomment-515115041, but it's sadly already normalized and rather convoluted.