>I’m not entirely happy even with this level of performance regression, but for now I have run out of ideas for improving it further. The new borrow checker is a lot more sophisticated and tracks a lot more data, so strict performance parity is a tough ask. Nonetheless, given how bad performance was a few months ago, I’m happy that we’ve got it down to a level where most people probably won’t notice any difference. Given that the new borrow checker makes Rust a significantly nicer and easier language to use, I hope it’s an acceptable trade-off.

Hmm. For me, fast compile times have vastly more ergonomic benefit than NLL.

>Hmm. For me, fast compile times have vastly more ergonomic benefit than NLL.

Well we weren't getting those anyway :(

foldr

Yeah, I'm sure that this specific performance regression will have a minimal impact on overall compile times. But slow compilation is often a case of the death of 1000 cuts. Small performance regressions are tolerated over a long period, and eventually add up.

pjmlp

Until cargo gets proper support for binary libraries, and stops re-compiling common crates, things won't get better.

Although I still hope they will eventually come around to fix it.

saghm

I've been using sccache (https://github.com/mozilla/sccache) lately, and it mostly solves this problem for me (although I agree with what others have mentioned that the issue isn't that much of a deal-breaker for me given how infrequently I run `cargo clean`.