> -----Original Message-----
> From: Robert Brady [mailto:[EMAIL PROTECTED]]
...
> > I don't know if anyone *really* cares, but they are there, and
> > they are "doublewidth". But: 1) UAX 11 does care about this for
> > compatibility reasons; and 2) this kind of "width" data really
> > belongs *in* (or associated with) the font. So by choosing an
> > appropiate font list (in your terminal emulator preferences, or
> > whatever) one could provide the "context" needed, even for PUA
> > characters. An FDCC-set is most certainly the wrong place (unless
> > you also consider putting fonts there...).
>
> And how on earth is the program running in the terminal
> supposed to know the widths of the characters?
By having the same list of "few-width" fonts (set by preferences
and the same actual fonts available in the system, either by
copy (if remote) or actually the very same files). The program
(via library routines, maybe wcwidth) would then query [the width
auxilliary files for] the font(s) (rather than any FDCC-set) to
get the width data.
The systems have to agree on character encoding used, by
preference settings, ISO 2022 aside. Similarly they have
to "agree" on fonts used in the same way (any font setting
protocol aside).
I admit that there may be versioning issues here, but I don't
think the width data for such "few-width" font will vary once
developed and properly verified. So I don't any problem *in
principle* here (though there may be practical problems).
Kind regards
/kent k
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
