My background is the book BDD in action.
How one goes from Feature to Stories? More specifically I would like to understand the following:
1- When does one provide the decomposition into stories? Do you do it while spotting the examples that illustrate your feature ?
2 - If so how do you proceed in your development then: Do you first start to on work on the stories while leaving your feature pending ? Do you right the feature scenario first and then, move on the stories that comes out of it, while leaving the feature test pending ?
The key here is the process, if i do a decomposition of a feature into stories, before writing my actual scenario that spot those stories, i might be writing stories that are not relevant to the feature isn't it?
This though make me think that stories must be fleshed out in the same way as unit test/integration test are fleshed out of a user stories?
However if one does that well it is difficult to plan a feature for scrum iteration.
I understand that it is just a planning tool, but it would be interesting to understand how this planning is actually used in the context of dealing with a feature.
I believe one should not do too much upfront planning, not commit to things without being sure, but in the mean time, it seem to me that the decomposition into the different process (stories) involve into the feature process requires some upfront planning.
I would appreciate a clarification on that point.