Nicolas Boichat

@mjguzik great catch! I wanted to cover the integer case with this issue, but you're totally right, and it does help reducing system time in all cases. Just adding this...

Err, `seq` with reasonable integer parameters does work correctly right? (e.g. `seq 0 1 1000000`) I don't think we should just throw the towel because it is buggy in some...

Implemented a fast path in #7564. It's quite a bit of added code, but it does provide significant performance benefits, and gets within 10% of the GNU implementation in those...

No problem with me, but it might be best to base your changes on top of #7458, since that heavily refactors the formatting code.

My understanding is that it's totally ok to open PR, _especially_ for such small changes (I read it more as caring for the time of the contributor: if you care...

Maybe I'm missing something, but making that simple change (remove `enable_pipe_errors`) doesn't work for me. From `strace`, Rust decides to ignore `SIGPIPE` very early on in the startup process: ```...

For the latter part of the original report, #8353 prevents this bug from being hit in `trap_sigpipe_or_skip_`: ``` (trap '' PIPE && timeout 10 yes |:) 2>&1 | grep 'Broken...

Prototype in 133b7cb564f4af3d71b1030d19c72c747fb7d267, updating `uu_date` only. Basic stuff seems to work, apart from 2 tests failing because of missing format support (see above): - Fixes #7497, #7498, #7659. - Helps...

Converted `ls` as well, same branch https://github.com/drinkcat/coreutils/tree/jiff-dirty . Size is "okay": ``` cargo build -r -p uu_ls && ls -l target/release/ls -rwxr-xr-x 2 drinkcat drinkcat 3769336 Apr 27 13:57 target/release/ls...

Maybe! For reference, `chrono` implementation without "pre-scan" is 4% faster than jiff's ("pre-scanning" saves another 10%). -- all within that `ls` command above, raw performance gap will be larger of...