Nick Marino
Nick Marino
This is a pretty minor nitpick, but I noticed an inconsistency between the docs and the behavior of the custom formatting code. The docs for lager_default_formatter say "any traditional iolist...
Instead of just returning `true`/`false` based on whether the row was inserted, this changes some insert-related code to now return `Some(n)` or `None`, where `Some(n)` means the row was inserted...
The shrinking algorithm we've implemented has worked well for the tests we've written so far, but it's worth noting that it is limited to removing a single test operation at...
Currently, we only shrink by removing entire operations, not by shrinking the operations themselves. Shrinking an individual operation can be quite tricky, however, because oftentimes later operations were generated based...
I'd like us to write more tests for the framework itself, and verify that stuff like shrinking works correctly. Not super critical, since the test suites that use `proptest-stateful` kiiiiind...
Currently you have to implement every function in `ModelState`, even if you don't care about things like post-test cleanup or postconditions. It's probably worth making default no-op implementations of those...
The current version of this library is tied to async Rust (and specifically to tokio). You can write tests that don't use any async code, and they'll work fine, but...
Would be nice if we could just call `ProptestStatefulConfig::default()` if we don't want to have to fill in all the struct fields by hand. Maybe that gives us min ops...