> ... further down, they rip out the remains of unicode handling ...

Ah. Well, that makes it a little unfair, surely.

> Looking at a single byte at a time, it presumably only handles the "C" locale ...

Again.

> --max-line-length? no. But I'm guessing GNU wc isn't benchmarked with that option on

I wonder if wc does the work anyway, and only reports it if asked, or if it actually changes the code path if it's not needed.

So this entire post feels ... intellectually dishonest. personally I'm all in favour of Haskell, and I wish I had the chance to use it "in anger" rather than just doing the occasional toy thingie that I do. But this post doesn't do it or its community any favours.

Disappointing.