David Chelimsky

/cc @philomates @nubank/eng-prod-test

@philomates if we don't add support for predicates, we should at least document that fact. Maybe even an error or warning? WDYT?

Thanks for this @pitalig. I agree that we should just be using set logic for sets. Are you interested in making a PR, or would you prefer that somebody else...

@pitalig FYI, here's the same test using `matcher-combinators.standalone` (no test framework necessary). Just pointing this out because `matcher-combinators.standalone/match` is a relatively new addition. ```clojure (require '[matcher-combinators.standalone :refer [match]]) (reduce (fn...

I think that `(missing 9)` and `(unexpected 11)` is easier to grok. @philomates you said > I don't think it would be too difficult to adapt the behavior to be...

> and as long as we're 100% sure Alarm bells ringing! I strongly recommend avoiding anything that backs us into any corner on this. How about releasing alphas, betas, RCs,...

> Note that Travis build may fail due to timing It's happened every release that I've done, and it's been a frustration that I've simply accepted. Does anybody know if...

> We don't have the step where it happened like we do with failed assertions. Instead we just have the flow name. The stack trace [in the gist](https://gist.github.com/caioaao/0505aef68992f5bd6da88aba3e27a6f8) points directly...

/cc @markomafs

> Capturing match results is coupling state-flow to matcher-combinators implementation, and I don't think that's a good thing. [That is already the case](https://github.com/nubank/state-flow/blob/b90bb160b2e9bffc2c68a56507925f477b406d43/src/state_flow/cljtest.clj), so perhaps we need to think about...