Also, it seems that the faster machines get, the less incentive people have to write code that is not enormously wasteful of resources, and they either get away with not having had formal software education, or they either forget what they spent a lot of money to learn or, as you suggested, they let laziness keep them from writing good codeĪ while back, I traced chrome to see why it was racking up so much cpu time when I wasn’t asking it to do anything, and I saw that chrome was issuing non-blocking polls that nearly always returned with none of the polled fds ready for io, but those polls were then followed by the unconditional issuance of non-blocking reads on those descriptors that of course weren’t ready for i/o. As to the effect of programmers not knowing the difference between order n and order n**2 (o more) algorithms, I first ran Mozilla on a 4 MB Vax 11/780 which was a one MIP (roughly equivalent to one megahertz) machine, and it ran just barely slower back then than it does now on Intel processors that execute code at least 4000 times faster.Ī bit of the difference is due to memory allocation issues, but the majority is due to code being written by multiple people who don’t understand what has already been coded and reinvent the wheel over and over. I would be very happy if programmers had taken little more than a “combinatorial algorithms” (math or C.S) class and the C.S. Too many people working as programmers today have not had formal training, and don’t know why they would need it.

You are not kidding! But it’s not all laziness. Not fixing memory leaks or allocating too much memory are the two biggest. There’s plenty of things developers can do that will suck memory fast if they’re lazy. The explanation is the memory leaks, as has been repeated in this thread several times.

Why does firefox take up so much memory