Saturday, May 26, 2007

The case for McMansion architects (part 2)

So I think I might even have reasonably established that handing off a specification to an architect (no matter how good) is necessarily going to be nothing but an unmitigated disaster if you're expecting something arable the first go-round.
But these McMansions, these big, generic, samey houses. These enterprise software applications, this big, onerous, pay a million and pay a million more for consultants to get something nominally useful that no one likes using programs. Why do they get built in the first place?
I keep coming back to a PBS Frontline program called The Persuaders and in particular, the interview with Clotaire Rapaille, a French cat who isn't afraid to pose like he's in a rap video in front of his mansion with his cars (seriously, check the picture on the front page of his internets site) and proffer a bastardization of Jungian archetypes and Lacanian linguistic dynamics for mad profit.
As he explains in the program, the car companies came to him, looking to get in touch with their consumers and give them something that would jump off the lots, so he started interviewing people and came to a revelation - bigger automobiles, taller off the ground, with tinted windows were "code" for "wealth", "power" and "domination."
Rapaille talks about drilling through the mammal brain and speaking directly to the lizard brain. John Coltrane put it more eloquently - "The emotional reaction is all that matters. As long as there is some feeling of communication, it isn't necessary that it be understood." Rapaille sure was on to something with making them SUVs bigger and taller and whatnot, because they provoke gut responses in people. You ask people who drive them why they drive them and they'll tell you "because I feel safer in them" and it really doesn't matter if you were to put the crash tests that say otherwise in front of them. They're safer and that's all there is to it. You ask me, who would just as soon drive a goofy European-small car, about them and I hate them because they're gaudy monuments of conspicuous consumption.
But underneath my upturned nose, I'll admit it - there's the seeds of jealousy in there. I can't explain why, either. I've driven them and hate driving them. I don't like being up high, they handle like tanks and jesus christ the gas milage. But still, when I'm driving behind some jackhole and I can't see through the blacked-out rear window to the car in front of them, wouldn't it be nice to be able to see the guy in front of them?
Why the McMansions? In this context, I think it's easy to see where the infatuation with more house and less soul than you need comes from. For the same money you're putting out on a McMansion (and probably more because deviation from the crowd is gonna cost you), you're ultimately getting "less" house. There's safety in numbers and everyone else is doing it, so it takes an extra dose of courage to get it done.
You can't define soul, but square footage is easy to quantify.
Think about it - deep down inside, you've got that sense of unease gnawing at your stomach. You're spending more money for ultimately less house. The garish abomination that they're building is more than they need, isn't built for them but rather some generic approximization of them, they can't afford it, but... they're getting a whole lot of square footage and maybe I could make it my own by decorating it nicely and choosing the fixtures I'll have the place built out with? It's easy to slide back to the road more travelled.
The lowest common denominator is hard to argue with for a reason. IBM may not be the greatest, but nobody ever got fired for buying it. Budweiser and McDonald's suck, but won't you come off like a snob if you tell your friends you want to get something better?
I'd like to think that I have the courage to admit to myself and others that I'm OK with smaller and better. You probably have the same courage... but do we have the courage?
Product development comes up with a giant specification. You may have the courage to question whether a service-oriented n-tier architecture implemented in .Net 2.0 and redundant Oracle databases communicating with your Oracle Financials OLTP system with the data cached in XML for quick retrieval and the requisite smattering of patterns talk makes any sense whatsoever for this hypothetical product, but how convincing of an argument can you make to the contrary
Product development probably isn't really interested in hearing that these dozens of features that they've spent weeks and months painstakingly realizing and documenting probably aren't needed. Management isn't going to be sold on you telling them that less being more - they see their career riding on hundreds of pages worth of specification and who is this programmer to tell me that less is more and we don't need these features that product development does? We paid that architect good money for a good reason.
McMansions, SUVs and enterprise class software are all about us being sold what we want instead of what we need. It awfully hard to argue with the inner child that's having problems rationalizing away the fact that we're paying 15 percent more money for 15 percent less square footage. Peopleware costs a boatload and wow look at all of the things that it can do (I've tried to figure out what it is and near as I can figure, they've managed to assemble what I can only describe as a Rorschach application), so it can obviously do what we want (nevermind that we need to use the thing and that may not be possible). It's hard enough to argue with our inner child; you get a room full of inner children together and the odds of common sense rearing its pretty head rapidly approach zero.
A quote I've heard that's apt - "in a community of saints, we are all sinners."

No comments: