You are reading a single comment by @bq and its replies. Click here to read the full conversation.
  • I'm pragmatic about these things. The code i need to write depends on what i need to achieve. Which is exactly why boilerplates are unsuitable. For example I will use jquery if it would save a lot of coding, run faster in less memory, and lots of the code would make use of it, but I wouldn't use much else. Jquery leads you into a lot of highly inefficient coding practices that stem from its use of the DOM as a data store, for example, but sometimes it's unbeatable.
    With Modernizr, you're loading a library to do what exactly? Test whether a browser supports a certain feature? You should already know what your target browsers support, and in the real world you'll only be using one or two of these things. It's invariably quicker to include only the code for those features, or simply work around it.
    With reset CSS, after your project is complete go and remove everything you're not using and you'll probably end up with a third of the CSS, if that. And then remove everything you redefined elsewhere and you'll remove half of that. And then there is the HTML itself. Do you really need all those containers? And do you need to use

    so badly that it's worth deploying a load of JavaScript and CSS just to make IE work, when you could remove it all and go back to a
    ?
    The first rule of performance websites is only deploy what you need. Boilerplates take that rule and throw it out of the window.

About

Avatar for bq @bq started