Profpatsch
Profpatsch
> Of course you can transpile C to JS, but isn't it better to transpile Coq to JS? In practice no, because there’s a lot more industry work going into...
@Gabriel439 I’d be interested where you draw the line of adding new builtins symbols to the Dhall standard. I’m semi-torn about anything that is time/technology-specific, like IPv4 addresses. I’m afraid...
They are resolved in the importer though and don’t have a representation in the dhall AST.
How about ``` w'' foo bar '' ``` for “wrap” (like e.g. rust or python use `r"` to disable escape codes).
Super not on board with the idea of adding text parsing abilities. https://github.com/mozilla/nixpkgs-mozilla/blob/master/lib/parseTOML.nix ^ this is a nix toml parser. As you would expect, it is super slow, runs at...
> It's 14 MB in size. There must be some unnecessary parts that can be stripped in there.
`1.5M /nix/store/4zd34747fz0ggzzasy4icgn3lmy89pra-gcc-7.3.0-lib/lib/libstdc++.so.6` huh? `7.5M /nix/store/2cmvn0lp5jxqc3s7cxv0ws2rc6h67jqs-dhall-1.8.2/bin/dhall` The problem is that `strip` can’t do much with Haskell binaries[citation needed].
@nh2 use `nix why-depends` (from nix `2.0`)
Oh heck, please don’t add any Unicode processing to dhall
We definitely need support for importing dhall higher-order functions into Haskell