On Wed, Dec 03, 2003 at 05:48:57PM +0000, John Levon wrote: > On Wed, Dec 03, 2003 at 07:47:37PM +0200, Martin Vermeer wrote: > > > What I meant was that in light of the smallness (assuming good > > authoring practices) of these style instances -- as rare(-ish) > > exceptions to common rules -- the balance of pros and cons looks more > > favourable for the insets/objects paradigm. And the paradigm will tend > > to support and encourage these good practices. > > > > Of course it's possible to support semantic styling with ranges too... > > heck, you can practice structured programming in assembler if you want > > to (only partly tongue-in-cheek ;-) > > You're missing your justification of why one is somehow more structured > than the other. Feel free to take out the overlapping stuff if you like.
If the overlapping stuff is out and physical position == locical position is in, all-boxes is (a) natural, (b) simple. Andre'
