Practice Trumps Theory

     Happy Cinco de Mayo to all you Mexicans out there among my Gentle Readers. To the rest: don’t go out for Mexican food tonight. Trust me on that.

     Now have a snippet of dialogue from an unnamed story:

"Why is there a herring duct-taped to the ground wire?"
"Sir, I don't question your methods."
"That's not a method, that's madness. Witchcraft."
"Look, sir. Do you want to be right, or do you want results?"
"Well, results, of course."
"Then don't touch the fucking fish."

     That made me howl with laughter, not just because of the exceptionally clever phrasing, but because I’ve been there.

     In my years in engineering, I was often responsible for meeting a tight, rigid deadline: one which allowed for no slip-ups. On a couple of occasions there was a large pot of money at stake. Once it was in the tens of millions of dollars. Defense engineering can be like that.

     No one in his right mind would commit to such a deadline without the certainty that he can meet it. In the world of military procurement, there are no second chances. If Company A fails to meet the time and budget targets, Company B will be ready to step forward.

     But once you’ve committed, the watchword becomes No Experimenting! You must insist on proven methods only. The development team leader must resist any attempt to insert an attractive but unproven method with immovable firmness. Yes, the attractive but unproven method might later be proved better, faster, cheaper, or some combination of the three. But you can’t risk it with all those bucks on the line.

     That’s akin to blasphemy to a bright young engineer, only a year or two out of college, who’s sure he has a silver bullet chambered and ready to fire. The team leader was educated in the Sixties or Seventies; he’s hopelessly out of touch with what’s been happening since then. And boy, can that bright young engineer pout! He’ll also talk to his colleagues about his old fuddy-duddy of a boss. “Isn’t engineering about doing things the best way?” he’ll protest.

     It’s a sad lesson, but it’s one that must be administered and driven home with as many hammerblows as necessary. No: engineering is not about doing things “the best way.” It’s about meeting the specification within the given constraints, especially the constraints of time and budget.

     Getting that across to subordinates has been among the toughest and most thankless tasks of my career. I’m sure I’m not alone in that.

     But there’s worse: that bright young engineer might not be a subordinate. He might be a “compliance officer” assigned by the customer. Again, that’s common in defense engineering. You have to listen to him; he can throttle the money flow at his whim. And it’s amazing how intrusive and extensive his whims can be.

     There was one such occasion where a compliance officer wanted to have my team develop a very large program in a language none of us knew. I fought him off, but it was a memorable tussle. We delivered on time and within budget, but I never got another polite word out of him. Fortunately he was reassigned to another defense contractor after that.

     In such situations, when you have a love of knowledge and technology, the temptation to “go along to get along” is amplified by your own predilections. Here there be tygers! What a victory it would be, your subconscious whispers, to improve on the prevailing state of the art – even if no one knows you did it! Maybe the shiny new method will work!

     Well, yeah… but what if it doesn’t? What if you can’t get it debugged in time? You’re the point man; when you miss the deadline, the avalanche will fall on you. You won’t get to point at anyone else and say “Well, he said it would work.”

     So consider engraving this exquisitely concise and pointed motto on a nice piece of mahogany, inlaid with mother-of-pearl:

Don’t Touch The Fucking Fish.

     Hang it where all your people will look upon it daily. And do have a nice day.