It seems that fixing this is a research problem, which would lead to the holy grail of programming languages, i.e. an ultimate language that is as expressive as Idris and as efficient as Rust, and is thus essentially perfect.
It seems that fixing this is a research problem, which would lead to the holy grail of programming languages, i.e. an ultimate language that is as expressive as Idris and as efficient as Rust, and is thus essentially perfect.
Programming languages are not theoretical things. They're concrete, practical tools that _enable_ other stuff. Engineering, not science.
>Programming languages are not theoretical things. They're concrete, practical tools that _enable_ other stuff. Engineering, not science.
You can't escape theory, engineering is applied science.
> engineering is applied science.
Absolutely. But the metrics are different.
Otherwise, since Assembly provides the most control out of all, would you consider it the most expressive? ;-)
Maybe this definition is idiosyncratic, though. It's certainly not objective.
Depends on what you mean. Idris's notion of multiplicities essentially subsumes Rust's borrowing (there's some differences with affine vs linear types), so I can't think off the top of my head of things that you can ensure with Rust that you can't with Idris, but Rust has a lot more quality of life improvements that make things less clunky (also having a GC, Idris can get away with a lot less need for borrowing in the first place).
Are both Idris's expressiveness and Rust's efficiency (given stronger guarantees) perfect? Aren't theese languages really complex both to learn and to write? There are poblems without a solution, perfect and unique to all of them.