There've been at least a half a dozen journal posts that have slipped through my fingers lately. I'm working. My project is in a crunch and while there's light at the end of the tunnel, it's going to be a few weeks before I'm posting regularly again.
It's funny, at times like these I often ask myself why the heck I still do software. What sort of a glutton for punishment am I? And yet, since accepting the tech leadership of this project I haven't gotten to think about code so much in months and that part feels great. Those muscles have felt constrained and cramped in the role I took this year and having to work in a close concentrated way can be its own reward. I'm bored out of my mind already of mining databases for test cases, and I'm champing at the bit to actually get to the part where I WRITE some CODE, but the architectural wranglings, the planning- those are good. If it weren't for the fact that this is all due to a completely avoidable problem, it'd all be fantastic.
We did some stuff wrong, I've learned a couple of hard lessons about ALWAYS standing your ground when you know you need sometihng and your client is trying to blow you off, but now we're doing it RIGHT and that feels really good.
When this is all over, I want about six months I'm never going to get to think about new stuff. I've thought about nothing but process and people for months, it's time to think about Java and widgets. Protocols and data structures and architecture OH MY!
I know why I do software. Because when you did it right, there's really nothing like it. Trouble is, doing it right seems to be more about ecology than anything else. A strong set of programmers is not enough. A strong set of managers is not enough. A reasonably engaged client is not enough. A lack of any one of the three can doom you.
But it gets worse. Even if you have the programmers who know their trade and the managers who budget intelligently and the client who actually knows what he or she wants, sometimes you're still doomed. What the client wants is not possible. Or the programmers are doing new things and that isn't in budget. There's a 1000 things none of which make the programmers stupid or the client unhelpful or the managers incompetent that can still spell disaster.
Still, it's a good game if you've got the nose for the right risks. If you know how to choose a spot and dig in and work, it's one of the best games in town.
Posted by karen at July 20, 2004 08:21 PM