How to input a Float64?

[oxinabox]

Literals seem to be Float32.
as far as i can tell the parser.dx code doesn't support Float64 (Double)

[danieldjohnson]

I believe you've linked to the prototype of pure-Dex userspace parsing, which isn't actually used when parsing normal Dex code. But indeed the Haskell parser implementation that parses Dex code seems to hard-code all non-integer numeric literals as floats:

dex-lang/src/lib/Parser.hs

Line 209 in a373663

<|> UFloatLit <$> doubleLit

(as does the type inference pass:)

dex-lang/src/lib/Inference.hs

Line 295 in a373663

UFloatLit x -> matchRequirement $ Con $ Lit $ Float32Lit $ realToFrac x

[oxinabox]

I believe you've linked to the prototype of pure-Dex userspace parsing, which isn't actually used when parsing normal Dex code.

Yes, as a work around to not being able to input them directly I thought I might use parse Double "3.1415926535".

But indeed the Haskell parser implementation that parses Dex code seems to hard-code all non-integer numeric literals as floats:

Yes that's the main issue