Closed xavierhaurie closed 7 years ago
We have seen issues like this before and in all cases the reason was a dependency between tests. Unlike karma, that always runs all tests together, wallaby runs tests in parallel, so neither the execution order, nor the set of test files that will be executed together can't be guaranteed.
From what you've described, some of your tests may fail because they somehow depend on other tests. When you edit a specific files, wallaby runs just that file, so nothing influences its tests and it passes.
I'd start with setting both workers
count to one to see if the issue is still there and remove the parallelisation factor before it's sorted out.
Then I'd try to modify tests
setting to only include the spec where the fantom error shows up. Running the config with just that spec should clear the error. Then I'd go ahead and include some other specs, while restarting wallaby to see if the issue shows up again (like bisecting by the other way around). The idea is basically to find out what other spec is influencing this one.
If you don't see any issues with your test dependencies, please create a sample GH project where the issue can be reproduced, I'm happy to have a look.
Closing the issue for now, please let me know if it's still valid and you can share a sample project so I may have a look.
Issue description or question
Test code fails, and then is made to pass by simply adding a console.log() statement. Restart the runner and it fails again. Remove the console.log() and it passes... etc.
Wallaby.js configuration file
Code editor or IDE name and version
IntelliJ IDEA v15
OS name and version
OSX