What kinds of activities will trigger reflow of web page with DOM?
It seems there are different points of view. According to http://www.nczonline.net/blog/2009/02/03/speed-up-your-javascript-part-4/, it happens
- When you add or remove a DOM node.
- When you apply a style dynamically (such as element.style.width="10px").
- When you retrieve a measurement that must be calculated, such as accessing offsetWidth, clientHeight, or any computed CSS value (via getComputedStyle() in DOM-compliant browsers or currentStyle in IE).
However, according to http://dev.opera.com/articles/view/efficient-javascript/?page=3, taking measurement triggers reflow only when there is already reflow action queued.
Does anybody have any more ideas?
This often solves those incomprehensible layout bugs.
Both articles are correct. One can safely assume that whenever you're doing something that could reasonably require the dimensions of elements in the DOM be calculated that you will trigger reflow.
In addition, as far as I can tell, both articles say the same thing.
The first article says reflow happens when:
The second article states:
I take this to mean the same thing they said earlier. Opera will try its hardest to cache values and avoid reflow for you, but you shouldn't rely on its ability to do so.
For all intents and purposes just believe what they both say when they say that all three types of interactions can cause reflow.
Cheers.
Look at the "Rendering triggered by Property Read Access" section of Understanding Internet Explorer Rendering Behaviour, where the following code in IE will cause rendering activity.