If you are building your own controls then you are all about building things in an OO manner because it helps you in the areas on manageability. You may or may not be testing your components with automated/unit test tools and quality is everything.. as well as browser compatibility. When doing this one of the major struggles in your world will probably be in the area of memory leaks (I’ve joined the coalition to kill IE6).
To be honest this is probably not you, but if you’ve come to one of my deep JS talks… this is my target (and my cardinal sin… sorry)
The way you probably use JS
Ok, the most common use of JS and the way that I think most people use it is to do minimal coding to wire up some validation or to wire up some interaction or even to include some Web 2.0 control (aka “Ajax Control”). Your mantra is “I want to get in and out quickly so that I don’t have to touch JS more than I need to…" We usually don’t take the time to test this stuff because it feels more declarative and that seems trustworthy (although you could test it)
This is exactly why jQuery is so popular. Not only does encourage this type of thinking (and even embraces it), but the re-usable plugins take on this same sort of thinking (the show/hide methods come to mind). That and the sheer fluency of the library makes it easy to use. The selectors are really a variation of the same selectors that CSS uses so you are re-using knowledge that should already be in your head as a web-developer/designer..
Anyway, while JS OO is cool.. most of us work the other way (and even those of us who use JS OO do most of our JS coding using the latter method)