[Arainach]

You don't have to claim that they're better, but making a big fuss about them in Zig - to the point of maintaining a compiler fork - and then not talking about it at all in the Rust doc (where it's apparent to anyone who's used Rust that compilation there is guaranteed to be slower) is dishonest or at least massively hypocritical.

[cowsandmilk]

The bun team made no mention of compile times in their post. The author of this post points to other work by bun that was hardly making a big fuss. It was just a Twitter post about parallelizing some things to make builds faster.

[woodruffw]

Where are they making a big fuss?

[Arainach]

I knew that LLMs had caused reading comprehension to tank, but I quite literally already directly answered that question - as did the original article under discussion here.

[woodruffw]

You haven’t answered it, and I can’t find a fair reading that would lend itself to that interpretation either. You can easily correct me by quoting the part of the Bun post that you find to be “making a big fuss.”

(Surely a relatively bland technical criticism of Zig - one that gets couched with praise - does not count as “a big fuss.”)

[Arainach]

> making a big fuss about them in Zig - to the point of maintaining a compiler fork

Literal quote (and again, appeared in the original article as well). Forking the compiler to try to get faster compile times is a significant effort that absolutely counts as raising a fuss.

[woodruffw]

The word “fuss” doesn’t occur in the original Bun article.

I also don’t think that forking the compiler is a big fuss. Large companies do that all the time, to carry whatever patches they haven’t upstreamed yet (or can’t). It’s really not that big of a deal.