Waffle will eventually be a generic table datastructure, but is now just an experiment, kicked off by Practicing Ruby Issue 2.2.
I am trying to build out an open source library via "Use case driven development", in which only features that fulfill a demonstrated need will be implemented.
Each feature request / patch to Waffle must come with an example use case. This use case should answer a particular question about a particular realistic data set.
If you want to suggest a feature, please follow the examples in the use_cases folder, and at a minimum, check in your dataset and the question you want answered. Ideally, also add some imaginary code for how you would like to see the problem solved.
The example code should only demonstrate the use of essential features, but can contain a comment about nice-to-have features. Once you've done this, submit a pull request so that your feature can be considered before you spend too much time working on its implementation.
You do not need to implement your feature, or write formal unit tests for it, but you are of course welcome to do so.
Your tests should (when possible) work against the same data set that the use case was written against. Your implementation code should only satisfy the needs of that use case. This doesn't mean you should write trivial stubbed out methods, but it does mean you can cut corners when it comes to things like validations or edge cases. We can wait for another use case to drive those features.
Contact [email protected]