Good discussion. Glen, is that certain large company generating dynamic content? Why not just add the body class on the server tier?
Mike
However, I do see your points about degradeability. I think it would be prudent for each developer to log (along wwith other relevant statistics) the availability of JavaScript. For instance, at a certain large company I know, they don't "test" IE 5.0. More than likely, the task completion rates would fall to zero for those users because the screen would be totally jacked. They make up under 1% of the audience. How many browsers in how many resolutions with CSS on/off, with JS on/off? You see what I mean? Additionally, you have to consider the purpose of each site. Take http://jquery.com/api/ Without JS, it doesnt have the examples. But does that matter? Who would look at that site without JS on? Or Yahoo Mail. They flat out redirect you to a page saying, "Sorry, you have to enable it". But who does that hurt? Is there really a non-JS audience out there who says, "It has to work perfectly without JavaScript and/or without CSS"? Its hard enough to test different browsers in different resolutions as it is. Regarding conditional comments to include other CSS: I find it annoying. Here is why. When I declare a class like div.header ul.global li.something span.other {margin-top: 10px} (or whatever) and realize that IE6 is 3 px off, and I cant figure out why. And my boss is pressuring me to move on. I want to define the hack or override for CSS -next- to the original class. Putting it in a ie6.css file doesnt allow me to keep the classes together and organized.
