Re: console update

[Marcus Brinkmann]

On Tue, Jun 18, 2002 at 09:12:32AM +0200, Niels Möller wrote:
> I imagine that the console server wants to handle
> things like TABS and linewrapping, so that it needs to know the with
> of the glyphs. Is that wrong?

No, it is correct.  The console server definitely needs to know the width of
the glyph.  I am not sure what double-width characters in Unicode require in
the server and clients, I guess it won't be too hard to implement that.

> Hmm, now I realize that one assumption I've been making throughout
> this discussion is that we're using fixed width fonts. If one wants to
> get rid of that, using variable space fonts on the console, then a
> glyph matrix isn't so natural anymore.

This is an assumption you really have to make.  curses and other libs/apps
can not do the right thing with variable space fonts.

Is there a terminal emulator for X that uses variables space fonts?  I would
like to try it out.  But I can imagine a hell lot of display problems caused
by app-unawareness, starting from ASCII art to graphic characters (box
drawing) and overwriting text.  Eg,

echo -n 'Progress 1%'
echo `tput hpa 9`2

should yield "Progress 2%", while

echo -n `tput hpa 9`\
`tput smacs`\
`echo \`tput acsc l\` | sed -e 's/^\(..\)*l\(.\)\(..\)*$/\2/'`\
`tput rmacs`

should print the upper left corner into the box drawing position "10" of the
current line.  The former should be dependent on the width of the glyphs
making up the first nine cells of the line, while the second would be
something independent from screen output.  You could of course define all
sort of heuristic rules to determine which type is meant, but if libraries
like curses will follow the rules is another question :)

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' Debian http://www.debian.org brinkmd@debian.org
Marcus Brinkmann              GNU    http://www.gnu.org    marcus@gnu.org
Marcus.Brinkmann@ruhr-uni-bochum.de
http://www.marcus-brinkmann.de