Post #2578189
2026-05-13 18:11 UTC
@joe@f.duriansoftware.com If I'm remembering correctly, I think Zig's old stackless coroutine implementation appeared used something similar to this, backed by a whole-program analysis.
Replies (1)
-
@joe@f.duriansoftware.com 2026-05-13 18:17
@zwarich@hachyderm.io interesting, do you know why they switched it up? it seems like you could turn the whole-program aspect into a linker problem