Greg Haskins

@A11oW The [Gherkin DSL](https://github.com/greghaskins/spectrum/blob/1.2.0/docs/GherkinDSL.md) (given/when/then) may be helpful for your use-case. Each sub-step inside a `scenario` only runs if the previous step has succeeded.

I'm glad Spectrum has been helpful for you, @tjarratt . Thanks for the feedback, it helps motivate us to keep working on it. Spectrum doesn't currently support randomization, you're right...

Sounds like you've got some great ideas. A few thoughts: - It feels like defaulting to declared order (making randomization opt-in) would make sense, at least at first. That would...

I hear ya, and I do agree overall. Just trying to balance existing users and expectations from the RSpec/Jasmine world. I'd be open to flipping the default for `2.0`. At...

Oh, and I think a system property is sufficient for a first pass at configuration. It's likely something people will turn on in their build script and forget about. And...

@ashleyfrieze 's suggestion should get part of the way there @zifnab87, and it works out of the box. I've used it myself to only run a subset of tests at...

Very interesting idea. Can you describe more detail about the workflow the Cucumber provides, for those of us who don't use it often? For example, what's the flow when adding...

My gut feels like time spent in `beforeEach` and `afterEach` should count towards the test-level timeout (like JUnit). However, I would really prefer to match any precedent set by other...

Certainly, I'd say JUnit5 is in the longer-term roadmap for Spectrum. I'd like to generalize Spectrum a bit so that it can run on multiple backend running/reporting frameworks like JUnit4,...

@gajwani Thanks for reaching out (and sorry for not responding sooner!). I touched on this a little bit [Design Philosophy wiki page](https://github.com/greghaskins/spectrum/wiki/Design-Philosophy). Overall, it might look like: - Start the...