Re:Pony: Combining safe memory sharing with Erlang-like actors
by GolDDranks on Sunday July 02, 2017 @07:50AM
(#14682092)
I just had a thought that made me ridiculously excited. Like in Pony, doing FFI in Rust is unsafe for obvious reasons. But would it be possible to define a subset of the typesystems of both languages to have a possibility of calling Rust code safely from Pony? (I doubt it would work another way around, since Pony needs its runtime.) That would create super cool synergy with ergonomic and ultra safe Pony and ultra performant and safe Rust.
Re:Pony: Combining safe memory sharing with Erlang-like actors
by spooneybarger on Sunday July 02, 2017 @07:51AM
(#14682095)
I suppose it would be possible although I'm not sure it would be worth the effort. Only one way to find out though.
Re:Pony: Combining safe memory sharing with Erlang-like actors
by user2283 on Saturday July 08, 2017 @05:39PM
(#14727317)
I think that would be really great. You can already call Rust code via FFI, but it looks like that languages have a lot in common with the LLVM backend, and memory- and type-safety. Imagine the boost to productivity Pony users would get if they could call Rust code from the crates.io ecosystem from Pony, with next to no glue code?
We love when folks notice us, hopefully a couple people who read this end up in the Pony community. That said, the "better-than-Rust" makes me uncomfortable. Pony and Rust have different goals and make different tradeoffs with regard to memory safety. We don't think one is better than the other. They are different.
Anyway, back to my Sunday. Y'all enjoy.