Loading ...
Sorry, an error occurred while loading the content.

Agile requirements traceability redux

Expand Messages
  • J. B. Rainsberger
    Everyone: I apologize for starting this again without doing all the proper research, but I want to get this question out and do not have the patience to read
    Message 1 of 6 , Jun 2, 2004
    • 0 Attachment
      Everyone:

      I apologize for starting this again without doing all the proper
      research, but I want to get this question out and do not have the
      patience to read /all/ the relevant threads.

      SUPPOSE we use Fit/FitNesse to write customer/acceptance/story tests for
      all our requirements (or even a large portion of them). Suppose there
      was a FitNesse coverage test that computed which parts of the code
      (class, lines, methods, whatever) are exercised by the FitNesse suite.
      Does this effectively provide adequate requirements traceability?

      Thanks.
      --
      J. B. Rainsberger,
      Diaspar Software Services
      http://www.diasparsoftware.com :: +1 416 791-8603
      Let's write software that people understand
    • Brad Appleton
      ... I think there were about 3-4 different aspects of traceability that came up in the thread (instigated by me :-). - One of them was simply traceability
      Message 2 of 6 , Jun 2, 2004
      • 0 Attachment
        On Wed, Jun 02, 2004 at 10:06:15AM -0400, J. B. Rainsberger wrote:
        > SUPPOSE we use Fit/FitNesse to write customer/acceptance/story tests for
        > all our requirements (or even a large portion of them). Suppose there
        > was a FitNesse coverage test that computed which parts of the code
        > (class, lines, methods, whatever) are exercised by the FitNesse suite.
        > Does this effectively provide adequate requirements traceability?

        I think there were about 3-4 different aspects of "traceability"
        that came up in the thread (instigated by me :-).
        - One of them was simply traceability from requirements to tests
        & their (successful) results. Your approach does that, and more.

        - Another was traceability of requirements thru code (and tests),
        such that one could (ideally) identify which parts of the code
        are impacted by which requirements. Your proposal would show
        which parts of the code have test-coverage. If it does that
        by also indicating which parts of the code are covered by which
        tests/requirements that cover it, then your approach would meet
        this goal. If your approach indicates which parts of the code
        are "covered" but being able to map requirements into code,
        then it wouldn't meet this goal. (but if its not really your
        goal - then maybe it is "good enough")

        Ultimately, I believe "adequate" will depend upon the specific
        traceability "requirements" your particular project is trying
        to meet. Does your project have such requirement(s)? And do
        you know what the tests/audits are that you will have to pass?

        --
        Brad Appleton <brad@...> www.bradapp.net
        Software CM Patterns (www.scmpatterns.com)
        Effective Teamwork, Practical Integration
        "And miles to go before I sleep." -- Robert Frost
      • Ron Jeffries
        ... I hate, as always, to agree with Brad, but I do. It depends on the meaning of adequate requirements traveability , which seems to be a notion that s all
        Message 3 of 6 , Jun 3, 2004
        • 0 Attachment
          On Wednesday, June 2, 2004, at 10:06:15 AM, J. B. Rainsberger wrote:

          > SUPPOSE we use Fit/FitNesse to write customer/acceptance/story tests for
          > all our requirements (or even a large portion of them). Suppose there
          > was a FitNesse coverage test that computed which parts of the code
          > (class, lines, methods, whatever) are exercised by the FitNesse suite.
          > Does this effectively provide adequate requirements traceability?

          I hate, as always, to agree with Brad, but I do. It depends on the meaning
          of "adequate requirements traveability", which seems to be a notion that's
          all over the map.

          Do you have a specific need in mind?

          Ron Jeffries
          www.XProgramming.com
          It is not because things are difficult that we do not dare,
          it is because we do not dare that they are difficult. --Seneca
        • Phlip
          ... I think traceability means, for any given block of code, you can ask what requirements caused this code? I m not sure what that gives you; I suspect
          Message 4 of 6 , Jun 3, 2004
          • 0 Attachment
            J. B. Rainsberger wrote:

            > SUPPOSE we use Fit/FitNesse to write
            > customer/acceptance/story tests for
            > all our requirements (or even a large portion of
            > them). Suppose there
            > was a FitNesse coverage test that computed which
            > parts of the code
            > (class, lines, methods, whatever) are exercised by
            > the FitNesse suite.
            > Does this effectively provide adequate requirements
            > traceability?

            I think "traceability" means, for any given block of
            code, you can ask "what requirements caused this
            code?"

            I'm not sure what that gives you; I suspect that
            planned design methodologies institute traceability
            because it gives good side-effects.

            But if we need traceability and don't need whatever
            the side-effects are, then:

            http://groups.google.com/groups?selm=5Tuvc.895%24w35.405%40newssvr15.news.prodigy.com


            =====
            Phlip
            http://industrialxp.org/community/bin/view/Main/TestFirstUserInterfaces




            __________________________________
            Do you Yahoo!?
            Friends. Fun. Try the all-new Yahoo! Messenger.
            http://messenger.yahoo.com/
          • J. B. Rainsberger
            ... No, I don t, but Brad s reply was enough. If the goal is trace requirement to code, then my suggestion appears to work. If the goal is enable impact
            Message 5 of 6 , Jun 3, 2004
            • 0 Attachment
              Ron Jeffries wrote:

              > On Wednesday, June 2, 2004, at 10:06:15 AM, J. B. Rainsberger wrote:
              >
              >
              >>SUPPOSE we use Fit/FitNesse to write customer/acceptance/story tests for
              >>all our requirements (or even a large portion of them). Suppose there
              >>was a FitNesse coverage test that computed which parts of the code
              >>(class, lines, methods, whatever) are exercised by the FitNesse suite.
              >>Does this effectively provide adequate requirements traceability?
              >
              >
              > I hate, as always, to agree with Brad, but I do. It depends on the meaning
              > of "adequate requirements traveability", which seems to be a notion that's
              > all over the map.
              >
              > Do you have a specific need in mind?

              No, I don't, but Brad's reply was enough. If the goal is "trace
              requirement to code," then my suggestion appears to work. If the goal is
              "enable impact analysis," then my suggestion also appears to work, at
              least in some cases: make the desired change in the simplest way, then
              run the tests and see what fails.
              --
              J. B. Rainsberger,
              Diaspar Software Services
              http://www.diasparsoftware.com :: +1 416 791-8603
              Let's write software that people understand
            • Brad Appleton
              ... I ask similar questions anytime on on a project that requires traceability. I say something like: what exactly does traceability mean/entail for the sake
              Message 6 of 6 , Jun 4, 2004
              • 0 Attachment
                On Thu, Jun 03, 2004 at 06:21:20AM -0400, Ron Jeffries wrote:
                > I hate, as always, to agree with Brad, but I do. It depends on the meaning
                > of "adequate requirements traceability", which seems to be a notion that's
                > all over the map.
                >
                > Do you have a specific need in mind?

                I ask similar questions anytime on on a project that
                'requires' traceability. I say something like: what exactly
                does traceability mean/entail for the sake of this particular
                project? How do we know when we've got it? What can we do to
                verify that for the folks requiring it?

                At the time, it didn't necessarily occur to me that what I was
                asking was essentially the same as "tell me your traceability
                story, then write the acceptance test for it" (I just did it cuz
                I was lazy, and wanted to automate it, but with minimal effort :)

                --
                Brad Appleton <brad@...> www.bradapp.net
                Software CM Patterns (www.scmpatterns.com)
                Effective Teamwork, Practical Integration
                "And miles to go before I sleep." -- Robert Frost
              Your message has been successfully submitted and would be delivered to recipients shortly.