SCM and Planning

Often one reads that CM recommends, or enforces, Planning. One even considers the CM Plan as an essential document.

This is paradoxical, and should be examined with suspicion under an SCM point of view:

• Only simple things require planning...
• Complex ones require experimentation and management: feed-back, testing of alternatives, recording of paths explored already, with the reasons for which they were abandoned, and the conditions under which they should be re-examined in the future, etc.

Planning is only based upon prior knowledge, whereas knowledge is supposed to grow as one goes, so that one shouldn't get bound to early decisions.

Plans are only useful as theories of what is to be done, and are likely to be changed. One shouldn't ever waste too much time and effort in the planning phase, since one is not well equipped then to make wise decisions. The best thing one can do in advance is to set up boundaries, and triggers to detect when these will be crossed: this is the testing philosophy of Extreme Programming.

A relevant quote from General MacArthur: Plans are useless, but planning is essential.
Let's take it to understand that the value of the planning activity doesn't lie in its product (the plan), but in its side-effects (the model of understanding built in people's mind). SCM should do better than build upon knowledge in people's mind. But this implies recognizing that a document such as the CM plan is not an adequate artifact.

-- MarcGirod - 14 Feb 2007