A new, er…, ‘method’
A Synthesis Concept
Let’s invent a new methodology. No, wait, if there’s any buzzword used by the Pointy-Haired-Bosses(PHB) that really does have a repeatable connotation, it’s methodology. Methodologies always mean a cross between religious fundamentalist zeal and a slightly good idea, but are inherently good at hiding the ‘good idea’ far in the back of the book. Let’s instead invent a new, er, approach that will attempt to be a synthesis of some of the better ideas of the better ‘methodologies’ around. We’ll call it PNAM, for “PNAM is not a methodology” to remind ourselves it’s just an approach. By the way, the P is silent.
I’ve mentioned combining MDA with something like test-driven-development before, and we’ll use that as the basis for our new approach. Many ‘agile’ design methods are good at the key ‘getting-it-done’ part, but are not so good at the ‘anticipating change’ part. That’s not to say that agile methods cannot react to change – quite the contrary, the whole basis of agile is to expect random change thrown at you, and have processes around that deal best with that environment. The actual changes that get thrown at you are more nebulous, requirements changes, or interfaces, or this or that. Traditional stodgy methods like waterfall are good at anticipating these changes, but terrible at dealing with them. Ask any software engineer from the 80’s what costs the most in software development and he’ll say changes to requirements. He could rattle off a whole bunch of examples, and then some. Waterfall’s huge arrogance in it’s ability to predict change is that it believes prediction == avoidance. Agile’s huge arrogance is that it believes that as long as it stays as light on its feet as possible, it will always outperform a project that attempts some ‘formal design’.
No comments yet.