Thomas Hurst

This feels a bit weird for my taste. I'm not sure I see the utility in arbitrary slicing of results, and it's a remarkably unintuitive way of reversing sort order...

Separate flags would be my preference, and they all need to be able to justify themselves. e.g. for reversing sort order, `ioztat -s operations` in a terminal puts all the...

I made a similar effort to use dataclasses shortly after 2.0's release, but decided against it because it reduced performance for a pretty arguable code improvement. Your approach seems to...

Results are generated with [hyperfine](https://crates.io/crates/hyperfine). I tried `__slots__` on baseline and found no measurable difference.

Our Python MSV is 3.7, so Dataclass slots is a non-starter for now. These long positional argument lists are quite nasty - they're begging for an easy-to-miss mixup and I'd...

`CPU: Intel(R) Xeon(R) CPU L5639 @ 2.13GHz (2133.53-MHz K8-class CPU)` They're teenagers now. I knew they sucked, but I didn't quite expect a similarly-clocked 15W laptop part to beat them...

Looking like a 5-10% performance uplift on my hardware, plus a 0.5-2% drop in memory use.

Seems specific to lightweight engineering - easily reproduced, just Lightweight a hull reinforcement module, reload the page and the value drops until you re-apply it. The behaviour is not seen...

eza's using [uzers](https://crates.io/crates/uzers), a derivative of [users](https://crates.io/crates/users) which is what exa was using. Doesn't look like it's changed much since I last poked at it. It's just calling `getpwuid_r` -...

`gcc test.c` is probably giving you a glibc build - can you try with [`musl-gcc -static`](https://www.musl-libc.org/how.html)?