I attended GR8 in the US (was in Minneapolis, MN - my home town) and attended a Cuke4Duke session where the steps class was written with a Groovy Cuke4Duke DSL.
I have the following ideas for features for JBehave and wanted to share. If there is value in these, then I will enter them in as enhancement requests. 1. A Gherkin story parser - being able to write plain-text stories for either Cuke4Duke or JBehave has tremendous value for users. Gherkin is the grammar for stories in Cucumber. 2. Follows on item 1 - support for tags - the ability to have @tag1 @tag2 ... for a story or a scenario. This can then be used in @BeforeStory, @AfterStory, @BeforeScenario and @AfterScenario with a tag attribute that would contain a syntax on tags excluded or included (@BeforeStory(tag="@database ~...@wip")) where the '~' is a negate. Likewise, the Ant, Maven and command line runners will need a tag execution option as well (tagsIncluded, tagsExcluded?) 3. Earned vs Burned - this is not in Cucumber but could really help get buy-in from managment. Imagine that you could add Value: and Cost: numbers to your stories or scenarios. The Cost is the estimate in story points that dev and test think a story or scenario will take. The Value is what the business gets from this. Assigning business value as points is sort of new in the agile world. The beauty of it is that on every run of JBehave, you could show Earned Value for scenarios/stories that pass. Show the Estimated Cost vs Actual Cost (another field) and you could do alot of management reporting directly from JBehave. Thoughts? Questions? Concerns? Record them or not? Thanks, Brian --- Brian Repko LearnThinkCode, Inc. email: [email protected] phone: +1 612 229 6779
