This is a question for "authoring tools" like eText and Pages, that may want to
differentiate HTML/HTML3, not the browsers (which are, as pointed out earlier,
forward and backward compatible).
The scenario is, say, a graphical table object that may become
* a gif of the table
* a PRE space-aligned ASCII table
* or an HTML3 TABLE
I think that this "selective processing" is rare enough that it shouldn't
confuse the spec with alternate handling for L2 and L3, but I just wanted to
note it. Can anyone think of any other reasons to possibly keep HTML3?
Rohit
PS: thanks to Bruce Gingery for his thorough tratement of defaults... I haven't
digested it yet though :-)