Martin Pool

> I’m very enthusiastic about this feature and its potential. If there’s any way to see it implemented sooner, that would be _fantastic_. 🚀 Thanks for the sponsorship! I dropped...

@moodmosaic @asuciux @wileyj You could try out the code in #223 if you like: to me that seems to run reliably and to find some potentially interesting coverage gaps in...

I see there is also a [library interface](https://docs.rs/nextest-runner/latest/nextest_runner/). Potentially we could call that rather than spawning a subprocess, but, I'm not sure if there would be any particular advantage... It...

@sunshowers For mutation testing it's important to distinguish "failed to build" (unviable) from "tests failed" (caught the mutant). (It's not completely critical to distinguish them since neither is actionable by...

Oh, good to know, thanks!

That makes sense, thanks. I think the main thing I might want to do for richer control, but not right away, would be to give it clues about running tests...

Oh right, I put some extremely early thoughts on that in https://github.com/sourcefrog/cargo-mutants/issues/194.

Yes, see https://github.com/sourcefrog/cargo-mutants/wiki/Compared I'm not aware of any better mutation testing system for Rust.

Yep, having an option to ignore seems fine, but maybe it should be the wrong default.... Martin On Wed, Jul 20, 2022 at 7:58 AM Ed Page ***@***.***> wrote: >...

Also, just to confirm, this hinges on the version of `cargo` that's being invoked as `cargo metadata`, not on the version used to build the program, right?