David F. McClinton wrote a paper on The Unwritten Laws of Systems Engineering. I thought of these during the recent discussion on Critical Path, in the context of "what other activities are taking place on a software project outside the context of writing code?
- Everything interacts with everything else
- Everything goes somewhere
- There is no such thing as a free lunch
- Configuration Management - never confuse change with progress
- Functional Analysis - never be afraid to start over
- Failure to Commit - better is the enemy of good enough
- Engineering documentation - if it is not written down, it never happened
- Planning documentation - never be above plagiarism
- Allocation of resources - a thing not worth doing is not worth doing well
- System synthesis - There is no shelf, to get the "off the shelf" part
- System integration - any interface left to itself will sour
- Work planning - plan your work, work your plan
- Contingency planning - we don't have time to do it right, but we have time to do it twice
- Assignment of tasks - nothing is impossible to the man who doesn't have to do it
- The problem of make work - don't keep polishing the cannon ball but do get the caliber right
- Design validation - any analysis will be believed by no one but the analyst conducted it
- Performance validation - one test is worth a 1,000 expert opinions
- Test planning - never conduct a test if you can't live with the result
- Time management - after all is said and done, a lot is said and little is done
- Briefing chart complexity - never use a word chart when a picture will do
- Systems risk management - never go in with the first wave
- More risk - never go in with the second wave either
- Management style - have the heart of a child, but keep in a jar on your desk
- Legal proof - deny everything, admit nothing, demand proof, and reject the proof
Mr. McClinton was for many years a Chief Systems Engineer at Lockheed Missiles & Space Company
