I'd be inclined not to. For one thing, I *strongly* doubt that the folks
doing the spec'ing are going to accurately guess what the popular objects
are -- one beauty of the Web is the organic way in which things can get
popular, quite suddenly and unexpectedly. Also, we want to allow for
very simple, low-weight browsers. So unless these objects were *extremely*
basic (in which case, they'd probably be pretty dull), I don't see the
point. The number of objects that a browser wants locally could easily
vary by several orders of magnitude from system to system...
On the other hand, if we wound up implementing this idea, I would
*definitely* want to create a browser that can read from CD-ROM, with
an eye towards eventually issuing a disk chock-full of useful toys...
-- Justin
Random Quote du Jour:
"Honk if your horn is broken."