Optimization of Computer Programs in C

Very often I'm wondering whether my effort put into very specific things is realyl worth it and whether it actually saves time, whether other may be right by ignoring this stuff until the program's done and so on. This article brought me back to the fact that I definitely know what I do. I've already written programs and games with those cases I optimize for, I've already done my own part of reducing algorithm execution time and all the other things sploding your performance checking. Thus, I know what I do and can target specific problems that WILL happen when I used my own code. The linked article was very pleasing to read as it proves what's in my head for a long time. And it makes me feel better about my knowledge. Sometimes I feel bad for not having already worked with a billion of engines, a trillion of OOP patterns and a bazillions of varied projects. But most people who DID work on so many things are either old(er) (making it possible for to looking forward to) or limit them selves to upper levels where this is totally normal. But I prefer the bottom-up approach of learning since it makes my base and judgement over things more stable and closer to what's actually working more effective. I don't care about how old the article is (in fact, some higher indexed topics are sorta out-of-date since the article dates back to 1997), but programs barely change what they are made of - nobody rewrites compilers for every new CPU coming it due to the established instruction set and computation behaviour. However, did anyone notice that Valve seems to be porting Steam to Linux? Fabulous!

No comments: