But how about ugrep?

https://github.com/Genivia/ugrep

I wrote the blog post before ugrep came out.

But I ran the benchmarks in the blog post a few months ago and included ugrep. ugrep isn't faster on any of them. https://github.com/BurntSushi/ripgrep/blob/master/benchsuite... (However, as I acknowledged in my reddit link below, ugrep is faster than ripgrep in some cases. If I were redoing my benchmark, I would include those cases in it.)

The full set of raw commands executed: https://github.com/BurntSushi/ripgrep/blob/master/benchsuite...

I've also contested ugrep's benchmarks in a few ways and have had unproductive conversations with ugrep's author: https://old.reddit.com/r/rust/comments/i6pfb2/ugrep_new_ultr...

Wow, I hadn't heard of ugrep, and I'm glad you mentioned it. The fzf-style fuzzy matching is really nice, and filtering with magic bytes is interesting too. I don't really have any massive files that render GNU grep slow, so the speed is almost irrelevant to me, but the other features look powerful.