The idea here was to compare different styles of writing CSS and see how they affect the page rendering. So using all ID’s, using all classes, using OOCSS techniques, etc. I gave some best practices advice on this a little while ago, but had no tests to back anything up. So the most interesting part of it all to me was how the tests were being measured. The test idea came from a Jon Sykes article where he writes:
The results concluded that OOCSS is way fast compared to declaring each selector with property and values separately. Even to the point being noticeably slower. Also, Opera is the fastest and Firefox is the slowest (IE not tested). But is this the perfect test? Doesn’t having all the repetitive property and values increase the size of the CSS and maybe that’s why it’s slower? It’s hard to test only the speed of the selector, as far as I know.