Information Technology Reference
In-Depth Information
how the CMMI can help us reason about our processes. It helps us ask questions
that can in turn help us manage our work more effectively. Often the questions
that result from using the CMMI are ones we might not think to ask otherwise.
As an example, organizations that rely heavily on product reuse should also
be relying heavily on reusing the end product such as the software code, and
reusing requirements, test cases, and test results. In other words, if I am
reusing a product that has already been verified, I should be able to reuse
that verification to gain the benefits of that effort. I will still need certain lev-
els of verification and validation in the new environment where I am reusing
the product, but the potential exists to “skip” certain lower levels of verifica-
tion. To help us reason about where in our processes it makes sense to allow
one to “skip” certain verification steps, SP 1.3 reminds us that we should
have verification criteria. This leads to the question:
What are the criteria we use to determine when a verification level can be
skipped?
It should be apparent that the creation of criteria can be a powerful aid to
help an organization and its processes become more agile in making
dynamic work-related decisions. However, criteria can only help if they have
been created and personnel are trained in their use.
19
4.24 Do I Need to Make Sure the Steps in My Processes
Are in the Right Order?
MYTH
CMMI-compliant processes require a sequence of steps.
I have observed numerous Technical Working Groups wasting valuable time
arguing about the steps in a process and the order in which those steps occur.
First, the CMMI defines processes as “activities that can be recognized as
implementations of practices in a CMMI model.” It doesn't say the order in
which those activities occur must be specified.
It has been my experience that when first developing process documentation,
any order dependencies should be one of the last items we worry about. I have
found that TWGs can spend incredible amounts of time discussing sequence
19. The power of criteria in helping an organization make more rapid real-time decisions is discussed at
greater length in the GEAR case study in Chapter 7.
