I have a question on the use of innerHTML vs. DOM methods. If you want to work object-orientated, it sometimes seems so cool to keep the DOM element you get from the appendChild method. And even more, its very handy to add an event listener like “click” and attach a function that is valid inside the object it is surrounded with. In that way, the function that is called will be aware of so much more then the event that comes as an argument of the call.

I would really like to know how you would judge this in terms of performance, since its so hard to compare..

Plus, I wonder how you see XHTML play in this topic. Of course you can’t really use XHTML (with correct mime-type) if you want to cover IE, but what if you don’t have to? Would it make a difference, performance-wise?

Thanks,
Daniel

BTW, I must say I find it sad to see how all these flaws of IE seem to dictate lot of your great work, and I’m very happy not having to worry much about IE personally.

that was a nice talk indeed and I completely agree with every single statement. In addition to it occurs to me - having watched several of those talks on yui theater - there is a common sense building up regarding best practices in JS coding.

However there is one objection I have to make: .innerHTML processing, if done correctly, is still much more performant than JS DOM level 2 processing. Also the code is much slimmer and easier to read.

I clearly understand why most frameworks stick to what is common average JS coding. I rather stick to the sheer performance boost.

Cheers from Berlin Germany,
and have a nice christmas time,

Frank Thuerigen

PS: you have mail

Can’t view the video properly (won’t cache it so stops every 2-3 seconds…). Tried a couple of times on FF and Safari…