I wasn’t aware that on non-x86 platforms long double is often implemented with quadruple precision. I had assumed it was an x87-specific hack. On ARM64 windows/macos long double is apparently 64-bits which could be a problem.
Personally something about that solution is unsatisfying. Feels like it’d be slow, even though that wouldn’t matter 95% of the time. I’d rather have 128-bit integer of nanoseconds.