February 2006 Archives
I was greeted the other morning by an email from a colleague to one of our local mailing lists inquiring as to whether this artfully crafted conference web page parody was my Machiavellian Invention. Anytime something a tad puckish comes along, they blame me. Were this really a "Machiavellian Invention", the authors would have loaded it up with AdSense tiles and tried to get it Slashdotted.
Having somehow missed the CFP, I informed one of the organizers that I'd just take Salmon Ladders: Adding Feedback to the Waterfall Model Saves Fish, and Keeps your Project from Spiraling Down the Drain somewhere where it was wanted. I received this rather brusque reply:
A talk about feedback ("going upstream") would be entirely inappropriate at this conference. Feedback is only needed if there's a chance you might get things wrong. If you do things right, you won't get things wrong. All it takes to do things right is to Follow The Bloody Process. Already, we see that not enough of the speakers realize the importance of this simple principle. We don't need any more apostates.
This all said, the mirth and glee on the part of the agile insurgency associated with the ritual mocking and flogging of this dead horse is sadly misplaced. This horse is very much alive and with us. Once again, commando coders working in teams of sizes 100 and 101 go about their work oblivious to the machinations of the infantrymen in the trenches of the 102 and 103 world, where callous officers are still re-fighting World War I. Herein lies a tale, albeit for another day.
To paraphrase George Bernard Shaw, the big-iron SEI worlds and the XP Agile cultures are two peoples divided by a common tongue. I'm amazed how many times over the last twenty years I've heard commando and infantry people us the same language to describe radically different realities, completely oblivious to these differences in style and scale. At each of these scales, its denizens take it as implicit that their reality is the reality. Such assumptions are largely unspoken, and surprisingly pervasive. When one's context changes slowly, if at all, the temptation to postulate that which is stable and persistent in that context must hold universally can be hard to restrain.
This kind of context sensitivity is one of my favorite things about patterns.