The return of the acute-cedilla problem

[Alexandre Oliva]

On May 29, 2004, Z <zleite at mminternet.com> wrote:

> Now would you (or the list) happen to know why xorg does not follow the
> usual usual us_intl composition rules?

Err...  Usual?  If it's international, and some languages have
legitimate uses for ć, then it wouldn't be, erhm, international to map
'c to ç, would it?

So I would think xorg is doing the right thing in this regard, and so
is gtk.  The difference is that gtk has custom compose rules for pt_BR
locales, and xorg doesn't.

The other difference is that the us-acentos keyboard layout, used for
the text console when US International configuration is selected, also
happens to generate ç from 'c.  Nor surprising, since us-acentos was
created in Brazil, and never adjusted to match the improved (?) X11
Compose settings.  Unfortunately, if this change was made, there'd be
no way (AFAICT) to generate ç in the text console.  Oh well...

> As it is now, gtk and, more
> importantly, openoffice follow the old style.

?!?

I've just started OOo within a en_US.UTF-8 locale and US International
keyboard compose rules, and 'c does generate ć, like everywhere else.
Ditto if I start it as LANG=pt_BR.UTF-8 ooffice within my en_US.UTF-8
X session.  This if FC2, BTW.

> Also, how come the console won't show accented charachters on
> en_US.UTF-8?

IIRC the text console is limited to 256-character fonts, so it can't
display full Unicode character sets.  That's a shame, but I think this
limitation comes from the PC BIOS, so it's a bit hard to overcome
without going for bitmapped text consoles :-(

-- 
Alexandre Oliva             http://www.ic.unicamp.br/~oliva/
Red Hat Compiler Engineer   aoliva@{redhat.com, gcc.gnu.org}
Free Software Evangelist  oliva@{lsd.ic.unicamp.br, gnu.org}