The most disqualifying thing about Zig is the semantics around aggregate assignment. github.com/ziglang/zig/... For performance, sometimes Zig doesn't copy structs on the stack, so refactoring to introduce temporaries can cause Zig to just miscompile code. It happens a lot and still isn't fixed. - ThreadSky

pcwalton.bsky.social • 25 days ago

The most disqualifying thing about Zig is the semantics around aggregate assignment.

https://github.com/ziglang/zig/issues/3696

For performance, sometimes Zig doesn't copy structs on the stack, so refactoring to introduce temporaries can cause Zig to just miscompile code. It happens a lot and still isn't fixed.

Comments

thoth.ptnote.dev•24 days ago

Issues like this are why I never want to use Zig. It feels like the language is actively out to get you sometimes.

goech.bsky.social•24 days ago

Are you german? wouldnt be proposing/contributing a solution a better way than complaining?

pcwalton.bsky.social•24 days ago

I'm not interested in using or contributing to a language that has no intentions of becoming memory safe. The more popular Zig gets, the less secure the industry gets.

jsoo.refl.club•25 days ago

Ah yes - the most optimal language

jacko.io•25 days ago

Do you know if the decision to get rid of "parameter reference optimization" will also fix aggregate assignment? https://github.com/ziglang/zig/issues/5973#issuecomment-2380332493

pcwalton.bsky.social•25 days ago

Not sure, but that's weird that they say they want to use function purity as a way to determine the semantics? The usual way you would do this would just be to always copy semantically and then have a MemCpyOpt like pass that cleans up unneeded copies. This is more precise than function purity.

lonjil.xyz•25 days ago

If you pass something "by value" and it no longer turns into a hidden reference, I suppose yes. Though if you use an explicit reference then I guess no. And I wonder what the criteria for determining "pure functions" that can safely have hidden references will be.

lonjil.xyz•25 days ago

Would be pretty funny if a parameter got hidden pass by reference'd because it's never written to, and then you accidentally write to it due to the return location semantics 😄

mcy.gay•25 days ago

completely unserious. they do not care about correctness, let alone safety

mcy.gay•25 days ago

not to mention that their lowering of various operations is weirdly 90s brained. it's like they don't know what we consider basic optimizations in this century

mcy.gay•25 days ago

all structs are passed by reference, when you actually want to pass small structs as a bunch of registers. and, they do not perform a callee copy of things passed by synthetic reference.

pcwalton.bsky.social•25 days ago

Ugh, I didn't realize it was that bad.

The struct parameter passing thing is a complete blocker for shipping things in Zig. The compiler is just broken. And I assume they aren't fixing it because they don't want to regress performance.

pcwalton.bsky.social•25 days ago

The moment I realized Zig was unserious is when they claimed they were going to prevent stack overflows with static reachability analysis. They're not even at the point of being able to argue about whether to use e.g. k-CFA--they don't even understand the problem domain yet.

mcy.gay•25 days ago

lol. lmao.

Comments

Posting Rules

Reply