I didn't think that image fading was all that new. What's interesting about the image fading script though, in relation to this thread, is doing this sort of thing in the code:
Code:
op_obj[fade.prprt]=op;
Which demonstrates code branching via predefined terms which vary by the objects available.
For instance, in IE 6 +, the op_obj will be an image object's filters object's 0 property, fade.prprt will be 'opacity', and op will be a number from 0 to 100.
In FF, op_obj will be an image object's style property, fade.prprt will be 'opacity', and op will be a number from 0 to .99.
In Safari 2, op_obj will be an image object's style property, fade.prprt will be 'KhtmlOpacity', and op will be a number from 0 to .99.
None of this is determined by sniffing the browser identity though, but by what objects are available when the code runs.
Bookmarks