Rachit Nigam

Oh interesting! This is a bit funky because it'd require you to have both the `.skip` and `.expect` around for the same test. One style thing we use in our...

You have not! Tags are entirely a naming discipline in runt and don’t really exist in any other form right now. Regarding your broken golden test example, the workflow we’ve...

@TheZoq2 just gauging what you think about the ad-hoc tag nonsense and if the `skip-ignore` flag still makes sense. If it does, let's merge it.

Yeah, sounds like a reasonable change to me!

That's interesting! One possible concern is that it would pollute the working directories for the tests and the user would have to add `*.new` to the `.gitignore` to avoid the...

There are two problems here: 1. It gets harder to read complicated scripts in runt descriptions. 2. Runt captures both STDOUT and STDERR making it more annoying to redirect everything....

Yeah, another question I have with this is should the setup be considered atomic? If so, we'd need some way of enforcing that for each test suite. Right now, `runt`...

We could but that won't fix the whole `cargo build` before `runt` problem because you don't actually want `cargo build` before each test. Another option is just a universal setup...

Ah okay, that seems like a reasonable thing to start with in that case. Let me know if you're interested in implementing this and I can help get started with...

Yup! That’s exactly how rsmt2 does it as well. I have a use case for it so might try implementing it. One nice thing in this library is since everything...