Stream: general

Topic: rust semantics, scheme/lisp syntax ?


zeroexcuses (May 23 2020 at 05:31, on Zulip):

rust semantics, scheme/lisp syntax

Assuming that things ilke lifetime / generics can be made to work nicely with scheme/lisp syntax, anyone else want a rust-semantics, scheme/lisp syntax language ?

In particular, I believe it will loser the barrier to entry for writing macros (both lisp's defmacro & scheme's hygienic macro system seem more powerful than macro_rules! and easier to use than procedural macros)

I suspect there are ways of using Rust that I am currently not using because procedural macros require lots of work.

Josh Triplett (May 23 2020 at 17:31, on Zulip):

If you were going to go there, I would suggest just a straightforward translation to the Rust AST.

Jake Goulding (May 23 2020 at 23:41, on Zulip):

"Rust++"

RalfJ (May 24 2020 at 07:25, on Zulip):

"(Rust())"

Lokathor (May 24 2020 at 07:48, on Zulip):

I wonder what the straightforward translation of generics/traits to a lisp-like language is

RalfJ (May 24 2020 at 07:52, on Zulip):

we'd have to look at a typed lisp dialect for that (AFAIK most of them are untyped)

pnkfelix (May 27 2020 at 12:14, on Zulip):

FX is one of my favs: https://apps.dtic.mil/dtic/tr/fulltext/u2/a256798.pdf

pnkfelix (May 27 2020 at 12:14, on Zulip):

But obviously samth’s Typed Scheme is way way way more successful

pnkfelix (May 27 2020 at 12:15, on Zulip):

Or sorry, Typed Racket

pnkfelix (May 27 2020 at 12:35, on Zulip):

At the same time, I’m with @Josh Triplett : I’d “just” add an S-exp variant of the current AST. Unless you’re going to design a new language variant (with new semantics rather than merely new syntax), I don’t think there’s all that much to do to accomplish that.

isHavvy (May 27 2020 at 21:42, on Zulip):

You can even implement the S-exp variant syntax as a macro.

Last update: May 29 2020 at 15:55UTC