On Thu, Oct 04, 2007 at 11:09:23PM +0200, Dov Feldstern wrote: > > > Andre Poenitz wrote: > >On Thu, Oct 04, 2007 at 10:23:25PM +0200, Dov Feldstern wrote: > >>Andre Poenitz wrote: > >>>On Thu, Oct 04, 2007 at 12:39:05AM +0200, > >>>[EMAIL PROTECTED] wrote: > >>>>Andre', could you give an example of a case where you'd like the cursor > >>>>to stop in between character styles? > >>>Anytime I want to start typing in one but not the other. > >>> > >>BTW, it's not exactly "one and not the other": it's "the outer and not > >>the inner". But if you have <b><i>hello</i></b> and now you want to add > >>something to <i> but not to <b>, you're no better off. With charstyles > >>as char attributes, you'd toggle off whichever one you don't want, and > >>type. > > > >With insets it is clear what is outer and what is inner, and you can > >even have bothe <i> in <b> and <b> in <i>. With current fonts it is > >more or less random what is in and what is out. > > > > With insets it may be clear, but still not match what I want in a given > situation, because I want to extend the inner and not the outer. So the > fact that it's clear isn't all that helpful...
All it'd take to rectify this is a 'swap-insets' lfun. > With current fonts I agree it's not clear. > > What I'm suggesting is continue using char-attributes (in an enhanced, > more general form, but the same basic concept) --- but to add visual > cues to make it clear exactly how far the range extends; and/or to make > it clear exactly what attributes the next font I type is going to have. > > Here's an idea, for example: in the status bar, clearly display the > "currently active" character attributes / styles. Or even better, this > could be a very unintrusive check-list of the currently active > attributes, allowing me to also uncheck any attributes that I want to > turn off. We are not talking about a fixed list of 'attributes'. We are talking about general semantic markup. Browsing through a random CS textbook suggests there are regularily several dozen things one might want to mark. > The point is, I think that with a little creative thinking, we should be > able to solve the specific problems that we have with font attributes, > without changing the entire metaphor, and without requiring changes in > other, already-complicated, parts of the code (e.g., cursor-movement). Thinking sometimes help. So I was told at least... Andre'
