Hi All
As we push to move the testing approach for OpenHIE forward we wanted to check in with the technical teams on the question of:
Are there any objections to using Gherkin as the test definition language? What are others using?
2 Likes
We’re using Gherkin for the OpenMRS test cases piloting with the OpenHIE test framework.
to Gherkin for readability of scenarios - OpenLMIS and PCMT both use it.
Though Gherkin and BDD can become a challenge for exactly the reason it’s useful: it separates the how from the why (of course - as do all abstractions - keep reading). I’ve seen that often results in scenarios being written in more of an imperative style rather than a declarative one when those over-focused on the how steer, and less frequently in it being so terse that it might as well have step 2 be “magic happens here” - usually when no one really understands an existing workflow or there’s disagreement about what’s important.
Getting it “right” can be a challenge, for most teams in my experience, and there’s really no way around the challenge: balancing the input to create the right abstraction for most.
For myself and the supply chain group I think we’re still wrapping our heads around the bigger picture of this testing within the OHIE architecture, do we have a resources page that starts with goals/vision/etc and dives in from there?