Re: Unicode and C++

[Owen Taylor <otaylor redhat com>]

Derek Simkowiak <dereks@kd-dev.com> writes:

> ->    while (*p)
> ->      {
> ->        if (g_utf8_get_char (p) == wc)
> ->          n++;
> ->        p = g_utf8_next_char (p);
> ->      }
> 
> 	Where are these g_utf8_*() functions defined, and will they be a
> standard part of Gtk+ 2.0?  (And is Gtk2 the official new name for Gtk+
> 2.x)?

GLib. Yes. No, it will be GTK+ 2.0. But I'm experimenting
with using gtk2 as a short abbreviation.
 
[...]

> -> I very much doubt I'll have a chance to do this for Pango-1.0. 
> 
> 	I do have a question about UCS-4 encoding: I thought that the
> 32-bit encoding was only used with special-purpose, "non-standard" fonts
> and characters.  I also thought that UCS-2 could encode any (registered)
> natural script.  (Otherwise, how could Java get away with a 16-bit char?)

Many people, myself included feel:

 a) Feel that confining yourself to a practically full character
    space is not a good idea.
 b) A character is a character, even if it is an Egyptian 
    heiroglyph. 

But yes, the 16-bit BMP (basic multilingual plane) is good enough
for almost any non-scholarly use.

> 	So, what benefit is there to giving Pango a 32-bit interface?
> Wouldn't any characters not in UCS-2 need custom fonts and/or glyphs for
> rendering, thus making Pango unusable as the layout/rendering engine?

Pango generally does not emphasize rendering with unicoded encoded
fonts, and is just as suitable for use with a font encoding 
characters outside of the BMP as one encoding characters within
the BMP. And there will be standardized assignments outside of the
BMP in the future.

                                        Owen