Barney (anyone interested in the subject), I understand you’re passionate and are familiar with the subject, so I simply caution to brush away preprocessor-skeptic advice, and to underestimate traditional code optimization.
CSS optimization is a complex issue. The example I gave is powerful because it shows how a fantastically effective technique is difficult to be employed automatically (to my knowledge no preprocessor has even tried, for good reasons): “using every declaration just once” changes the order of rules, and can have undesired side-effects if applied blindly. (I’m coming from a career of exploiting style sheets.) However, it’s just *one* technique. The point is that there are more such techniques. I don’t think we identified all of them—that’s why I urge to be careful with preprocessors (garbage in, garbage out), but also wish the CSS working group to focus more on the side effects of all the new features [1]. Now, there’s of course use for preprocessors. But there can’t be a blanket yes or no: use of them depends on the project, the size and skill of the team, the goals and ambitions, &c. Somebody just voting yes (or no!) without considering these points doesn’t give sound advice. As for small projects, I recommend to use and—cherish them. Small projects are great to build experience and to test theories. Try *not* to use CMSes, libraries, preprocessors, and such in small projects. The size—small :)—should permit that. There are many valuable things to learn in small projects. One is how to avoid problems that are annoying in a small project, but devastating in a big one. And another is to employ techniques that improve code in a small project—and also in a big one. This all assumes people to be interested in tailored, optimized code. I view this as the hallmark of a professional web developer, but I understand that some of us are so used to being “pragmatic” that anything that gets the job done faster wins. Alas, getting the job done doesn’t mean getting the job done well. And I think that’s also important to consider, at least be aware of, in this debate. Otherwise preprocessors still mean but one thing: garbage in, garbage out. I hope this helps. [1] http://meiert.com/en/blog/20131205/spec-fragmentation/ -- Jens O. Meiert http://meiert.com/en/ ______________________________________________________________________ css-discuss [css-d@lists.css-discuss.org] http://www.css-discuss.org/mailman/listinfo/css-d List wiki/FAQ -- http://css-discuss.incutio.com/ List policies -- http://css-discuss.org/policies.html Supported by evolt.org -- http://www.evolt.org/help_support_evolt/
