Version 0.7.0 breaks previous behaviour about bypass HOT 4 CLOSED

Thanks for reporting this issue.

Unfortunately I'm unsure how to reproduce this, looking at our PSPDFKit Instant codebase we have test cases with exactly the same pattern that all do work.

Can you post a complete failing example, so I can reproduce and debug?

from bypass.

Sorry for the slow reply @MSch, turns out that I slightly mis-represented the failing test case in my issue description, that's why you haven't been able to reproduce it.

The codebase I refer to is actually using setup_all callbacks as opposed to setup ones.

Here is a before/after PR: swoosh/swoosh#149

I'm wondering if that should still be mentioned as a breaking change somewhere, or if setup_all were never expected to work nicely with Bypass.

from bypass.

No worries @stevedomin and I'm glad we figured it out and it's working for you now 👍

You're right, setup_all was never meant to be the right place for Bypass.open to be called, because Bypass needs to be able to verify the expectation, and that happens at the end of the test.

So in that case it's not a breaking change, since it never was expected to work.

I'm closing this, but please let me know if you think there's a way to improve the documentation to make these expectations clearer. A PR is welcome too :)

from bypass.

Thanks @MSch, opened a PR with a note in the README: #46

from bypass.