On 3/31/06, Augusto Arcoverde da Rocha <agarobr.listas@gmail.com> wrote:
On 3/30/06, Peter Åstrand <astrand@cendio.se> wrote:
>
> On Thu, 30 Mar 2006, Wine Bugs wrote:
> > http://bugs.winehq.org/show_bug.cgi?id=1181
> >
> > ------- Additional Comments From Speeddymon@gmail.com  2006-30-03 10:26 -------
[cut]
I don't know the problem with the bug 4923 [1], but when we fix the
bug 1181 [2] and make Wine use KeySym, this race condition won't occur
since X11DRV_ToUnicodeEx function will not use AltGrMask or any other
global variable concerning the keyboard state, only a table like
"ksym2vkey". Besides, applications that use X server API directly
don't have this behavior.
---
[1] http://bugs.winehq.org/show_bug.cgi?id=4923
[2] http://bugs.winehq.org/show_bug.cgi?id=1181

I'm trying to solve the bug 2400 [3] and looking more closely the code
in dlls/x11drv/keyboard.c I started thinking in rewriting this code to
use directly keysyms sent by X server to translate XEvents to vkeys
instead of the currently
"xevent->keycode->keysym->keycode->vkey/character" way as implemented
from the X11DRV_KeyEvent to X11DRV_ToUnicodeEx functions.
---
[3] http://bugs.winehq.org/show_bug.cgi?id=2400

I think that rewriting the keyboard code in this way is a very good thing, and will be more than happy to help coordinate testing of it, although I can't test any myself.  Rewriting the code should allow us to close many of the keyboard related bugs in bugzilla, which will be a big BIG plus in my book.

Like I said, if you want me to, I will coordinate testing of new code.  I'll start a new thread called keyboard rewrite and ask for volunteers with different keyboard layouts to reply to that thread to test new code.  I'm also going to create an offspring of the main wiki page for this project, once given confirmation that you will do the rewrite.

Tom