Beyond what amounts to a packaging difference, what does that even mean?
How do you think zero-cost C ABI support (including fully working cross-lang LTO) works in lang implementations that have that? And how do you think that’s different from what Zig gives you?
Note: “Zero-cost” means zero runtime cost, not zero compile cost. The compiler does real work — monomorphization and inlining — to make abstractions disappear, and that work shows up as longer build times. See Reducing Compile Time for the trade-off.
most of the benefits come from a shared code base, one compiler (which is insanely fast with incremental rebuilds and can watch files for changes) means changes in the c code reflect in a single build step within seconds. the benefit is primarily for the developer as the compiler output should be relatively similar.
I’m not sure what you’re talking about then, in zigs case you don’t need an exposed ABI to use existing c code?
in terms of performance it will always be ambiguous as implementations are hard to compare across languages, however they use llvm as an alternative to support architectures as they build out their native compiler. in the transition for x86_64 they were seeing up to 70% increase in compile speeds. its still ambiguous as its dependent on their implementation of llvm, though its such a large increase it is still meaningful.
but i say all this as a rust and zig advocate where i would happily choose either over the the majority of alternatives.
it will always be ambiguous as implementations are hard to compare across languages
Correct.
in the transition for x86_64 they were seeing up to 70% increase in compile speeds
And that was a part of what I was hinting at, because you get >>70% speed-up with Cranelift in most Rust projects. But in either case, faster code generation is not free lunch, hence the mention of comparable runtime performance of generated binaries.
Beyond what amounts to a packaging difference, what does that even mean?
How do you think zero-cost C ABI support (including fully working cross-lang LTO) works in lang implementations that have that? And how do you think that’s different from what Zig gives you?
from rust docs:
most of the benefits come from a shared code base, one compiler (which is insanely fast with incremental rebuilds and can watch files for changes) means changes in the c code reflect in a single build step within seconds. the benefit is primarily for the developer as the compiler output should be relatively similar.
That has nothing to do with what we were talking about. And in any case:
Needs qualification vs. other languages for binaries with comparable runtime performance.
(Hint: you will be surprised.)
Not special or unique.
Not only not special, but literally exists in all workable build tools.
bindgen+build.rsis the Rust version of this.I’m not sure what you’re talking about then, in zigs case you don’t need an exposed ABI to use existing c code?
in terms of performance it will always be ambiguous as implementations are hard to compare across languages, however they use llvm as an alternative to support architectures as they build out their native compiler. in the transition for x86_64 they were seeing up to 70% increase in compile speeds. its still ambiguous as its dependent on their implementation of llvm, though its such a large increase it is still meaningful.
but i say all this as a rust and zig advocate where i would happily choose either over the the majority of alternatives.
ABI is not something that gets “exposed” or not.
Correct.
And that was a part of what I was hinting at, because you get >>70% speed-up with Cranelift in most Rust projects. But in either case, faster code generation is not free lunch, hence the mention of comparable runtime performance of generated binaries.