Le 16 août 07 à 15:49, Jean-Marc Lasgouttes a écrit :
Mael Hilléreau <[EMAIL PROTECTED]> writes:
Then many things are already ugly, e.g. charstyles :) Storing a
spellchecker setting into a character or font is senseless, whereas
an inset is designed for functional purposes.
Well, calling "senseless" the method used by all other word processors
is a bit weird. I am not saying that the fact that they use it means
that there is no other way, though.
Perhaps it is used by other word processors... But isn't LyX supposed
to be better?! :)
Why "senseless"? Because semantically speaking, spellcheck deals with
words, not characters.
The real problem you mentioned may be that insets do not provide as
many display schemes as needed. New schemes would be useful for
other purposes as well, e.g. for charstyles; however that's not a
reason for putting charstyles properties into a font.
charstyles have been introduced because of some shortcomings of our
font
system, in particular the fact that fonts did not nest correctly. They
feel however as an afterthought, and my belief is that we have not
found yet how to integrate them fully.
In my vision, a charstyle is a more abstract entity than a font. A
font is only appearance, whereas a charstyle should have some
semantics too. IMO the same font can be used for very different
purposes, whereas a charstyle has a given use. I see this difference
as analogous to the one between '\textit' (font) and
'\emph' (semantics).
Moreover, I don't know if you very often need to mark as non
spellchecked more than one or two words (I mean outside e.g. a note)?
Not me, that's not very usual.
I do not write my notes in SMS language either :)
Yes, and I even mix other things with that! :)
Ok. As 3-box is a way to display insets, I thought that this could
have consequences on how it will be managed.
No, the current work is more about making configurable a lot of things
that used to be hardcoded.
Ok, it's always a benefice to unhardcode ;)
Mael.
