After months of hard work, we’ve got Cucumber 2.0 into a state where it can run its own tests and (usually) give us useful feedback. We’ve just merged this code into the master branch.
There’s still a lot to do. The specs all pass, but only approximately 50 / 150 scenarios are passing. The 100 that fail are tagged out with @wip-new-core while we get them into a passing state.
The decision to move this code into master was taken because we’ve been getting pull requests from kind people fixing things on code that’s going to be deleted for the 2.0 release. Having the 2.0 code on the master branch should help avoid this confusion.
We’ll continue to release bugfixes as needed off of the 1.3.x-bugfix branch, but we’ll concentrate our efforts on getting the 2.0 code ready.
I’d love some more help with this. Particularly:
- Taking individual @wip scenarios and making them pass.
- Refactoring the adapter that bridges between the new report API and the old formatter API.
Helping with (1) could be as simple as taking an individual @wip scenario, diagnosing the root cause, and creating a PR with the @wip tag removed (so that the test is failing) and explaining what needs to be changed. Even if you don’t feel confident to make the change, just doing the work to turn a statistic into a meaningful task on our todo list would be really helpful.
Of course if you want to try and fix the code to make the scenario pass, that would be even better! We’re always happy to give you some free BDD coaching.