Andre Poenitz wrote:
On Wed, Oct 03, 2007 at 01:05:38PM +0200, Helge Hafting wrote:
I, for instance, actually prefer having two separate positions at a font
boundary.
For the sole purpose of supporting dissolving?
No, for the prupose of being able to work with it. Right now I get the
wrong behaviour in 50% of the cases and I have to do strange stuff to
move from 'before the font change' to 'after the font change'.
Getting a simple way to disolve insets is just a bonus.
I can see a problem - what if some word is wrapped in several insets?
I.e. it is both "foreign" (needs special spellcheck and that blue underline)
and also a "programming language construct" so it is set with a
special font. Finally, it might be emphasized too due to the sentence it
is in.
Do you want 4 cursor positions in front of this word?
In case there are four insets, sure. And each of them would have a
meaning.
Apart from that _four_ would be really rare.
And what inset will
dissolve when the user press backspace? "Do you feel lucky?"...
Guess what. Insets are nested in a very specific order. There always at
most one choice which inset to dissolve on "backspace".
Or should "many styles applied to the same range" be handled by a
single inset applying all those styles?
No.
Now there is only one extra cursor position, and dissolving removes
everything.
Of course this would emphasize structure and would not be acceptable
by the finger painting faction as that's not what they used to.
Do "structure" have to mean "hierarchical structure" that disallow
overlapping ranges?
Yes.
We need that for the external format anyway.
`
Funnily enough though, the people bitching at finger painting at this
list are also the ones that want non-structured fonts on screen.
This is so illogical that I basically gave up participating in this
kind of debate...
I do not mind that LyX is different in many ways. I just hope we won't
get limitations that feel very artifical, and that gets in the way of
normal work.
I find mathed usable, and it uses fonts-as-insets.
I too find mathed very useable. But it edits math, not text.
I think text have different needs than math.
Having several cursor positions in front of a word will be
both confusing and cumbersome to move through. It may not
be a problem if it happens rarely, but can we be sure of that?
If charstyles becomes useful & popular, then they will be used.
Having at least two styles beginning at the same point will
happen, with 3 cursor positions in front of a word. If you
click there and want to delete the previous word, will
you remove styles as well? Or if you want to remove styles, will
LyX give a clear indication of what style will go _if_ the user deletes?
Can these clear indications be provided without breaking up the
text too much?
Helge Hafting
