Tests and Requirements, Requirements and Tests: A Möbius Strip

Over a year ago, Robert Martin and I wrote a paper on exploring and disambiguating requirements with acceptance tests. In it we formulate the equivalence hypothesis – as formality increases, acceptance tests and requirements become indistinguishable. At the limit, acceptance tests and requirements are equivalent. IEEE Software magazine published it in the January/February’08 issue (http://www.computer.org/portal/cms_docs_software/software/homepage/2008/s1mel.pdf).

As usual, comments are welcomed.

Comments (5)

  1. rshagisultanov@hotmail.com says:

    Good short article on the subject. We came hard and manual way to the similar conclusions with one of our projects. There is no automation tools used though. Any idea how FIT goes along with the VSTS/TFS testing functionality?


  2. Excellent article de Grigori Melnik et Bob Martin sur les tests unitaires, les tests d’acceptation, FIT

  3. As I was going through postings from our internal agile discussion forum (yes, we do have one at Microsoft

  4. Josh Gough says:

    This is a good paper. Thanks for writing it.

    Can you or anyone else point us to how we would implement this vision inside of TFS / MSF Agile?

    Our team is sort of struggling with these distinctions right now.

    Take care,