Mature optimisation
So as I’m waiting for a nearly full rebuild to complete, I thought I should write up a blog entry. Previously I’ve not been writing much because of the upheaval due to moving and generally been busy sorting out a whole bunch of things. This last few days however, I’ve not been writing much because I’ve had my head deep in coding work, which is much more enjoyable! Of course, from your end-user perspective, it doesn’t really matter why.
Anyway, this week has been particularly busy with coding, because we’ve finally got things visible and running properly on the PSP, almost as well as they were on the PC, which is immensely satisfying. Unfortunately, it has highlighted all of the things that we need to fix because the PC let us get a way with a lot, performance wise, and still ran well. But that in itself is good, because it means we get to wade in and tackle some nice solid problems like “the texture format is too fat”, or “we’re making too many state changes”; for which we can see solutions, apply them and instantly see results in the form of improved performance. It’s a standing rule that you shouldn’t optimise too early, because you’ll end up wasting effort. However, when optimisation is appropriate, it’s definitely a satisfying task. As problems go, those are the ones programmers like the best.
Although we should have moved into our new office by now (by last week in fact), BT have been dragging their heels sorting out getting a phone line installed, and while my new landlords have been very understanding about it all, it’s been quite frustrating. I would have liked to have been in and set up by now, so that it was no longer nagging at the back of my head that we have to move in the future. Still, I’m sure the wait will be worth it, and instead I code away here, surrounded by still packed boxes of kit, all still waiting to move to the new office before I unpack them.
Right, build’s done – back to palletisation.