https://issues.dlang.org/show_bug.cgi?id=22481
Issue ID: 22481 Summary: Thread.sleep should be pure Product: D Version: D2 Hardware: x86_64 OS: Linux Status: NEW Severity: enhancement Priority: P1 Component: druntime Assignee: nob...@puremagic.com Reporter: cont...@lsferreira.net Conceptually, a sleep is not a side effect and can be implemented with unoptimized assembly code, e.g. running n noops instructions. Since we are using the kernel implementation for precision sleep and it provides a well documented behaviour we should trust it and fake pure it. --