Ad absurdum that would make assembly a better choice — we can just as well reason about better register allocation strategies. I think the fundamental abstract

Ad absurdum that would make assembly a better choice — we can just as well reason about better register allocation strategies. I think the fundamental abstraction level is very important to get right. Of course it depends on the problem domain, and it might just make sense for Zig to expose explicitly allocators everywhere, but maybe some implicitly passed allocator when not otherwise specified could be a better trade off (correct me if it is a thing)