On Fri 14.Nov'08 at 17:14:52 +0100, Carlos R. Mafra wrote:
> On Fri 14.Nov'08 at 13:57:06 -0200, Renato Botelho wrote:
> > On Fri, Nov 14, 2008 at 1:47 PM, Carlos R. Mafra <[EMAIL PROTECTED]> wrote:
> > > On Fri 14.Nov'08 at 9:03:22 -0200, Renato Botelho wrote:
> > >> On Fri, Nov 14, 2008 at 8:58 AM, Carlos R. Mafra <[EMAIL PROTECTED]>
> > >> wrote:
> > >> >
> > >> > I will probably be ashamed by asking you this stupid question, but
> > >> > let me try...
> > >> >
> > >> >> I'm using a Sun keyboard for a while, and i found a possible bug
> > >> >> on "Focus next window" function, let me try to explain.
> > >> >>
> > >> >> My keyboard have a left Alt (Mod1), right Alt (Mod5) and another
> > >> >> couple of Mod4 keys, on a common keyboard it could be compared
> > >> >> with the Windows key.
> > >> >>
> > >> >> Since this Mod4 key is present on both sides of keyboard, i decided
> > >> >> to change my shortcut to this function from Mod1+Tab to Mod4+Tab.
> > >> >> After changed it on WPrefs, it worked but not fine, when I have some
> > >> >> windows opened and press Mod4+Tab it opens the blue box with
> > >> >> icons in the center of screen and change the window, but, when
> > >> >> I release the keys, the blue box stay there, don't disapear.
> > >> >>
> > >> >> I need to press Mod4 again to make it disapear, now i back to Mod1
> > >> >> and everything back to normal. I believe it's a but, don't you?
> > >> >
> > >> > Did you check with using 'xev' if this Mod4 key generates both
> > >> > KeyPress and KeyRelease events when you press the key only once?
> > >>
> > >> KeyPress event, serial 27, synthetic NO, window 0x5800001,
> > >> root 0x41, subw 0x0, time 670255524, (-165,624), root:(871,640),
> > >> state 0x0, keycode 115 (keysym 0xffeb, Super_L), same_screen YES,
> > >> XLookupString gives 0 bytes:
> > >> XmbLookupString gives 0 bytes:
> > >> XFilterEvent returns: False
> > >>
> > >> KeyRelease event, serial 27, synthetic NO, window 0x5800001,
> > >> root 0x41, subw 0x0, time 670255804, (-165,624), root:(871,640),
> > >> state 0x40, keycode 115 (keysym 0xffeb, Super_L), same_screen YES,
> > >> XLookupString gives 0 bytes:
> > >> XFilterEvent returns: False
> > >
> > > After discussing privately with Renato I managed to reproduce his
> > > problem with the Windows key in my keyboard, which also has keycode = 115
> > > and is recognized as Mod4.
> > >
> > > After some time I noticed that this piece of code in src/cycling.c is
> > > responsable for dealing with the KeyRelease event:
> > >
> > > case KeyRelease:
> > > #ifdef DEBUG
> > > printf("Got key release\n");
> > > #endif
> > > for (i = 0; i < 8 * keymap->max_keypermod; i++) {
> > > if (keymap->modifiermap[i] == ev.xkey.keycode &&
> > > wKeyBindings[WKBD_FOCUSNEXT].modifier
> > > & 1<<(i/keymap->max_keypermod)) {
> > > done = True;
> > > break;
> > > }
> > > }
> > > break;
> > >
> > > The problem happens because no keymap->modifiermap[i] among all
> > > 24 possibilities (max_keypermod = 3 here) is equal to 115.
> > >
> > > By using a few printfs I discovered that
> > > wKeyBindings[WKBD_FOCUSNEXT].modifier == 64
> > > for the Mod4 key so if I set keymap->modifiermap[18] = 115 then I get the
> > > correct behaviour when i = 18 in the above 'for' loop.
> > >
> > > I don't know how to fix this properly, but as a workaround this patch
> > > works
> > > for me here.
> > >
> > > Renato, can you test it too?
> >
> > Hello Carlos, It worked fine with Left Mod4 key \o/, but there is
> > another one on right side, its keycode is 116, I suppose it's simple
> > to have this fixed too.
>
> Right, so my theory is ok.
>
> _But_ that patch is not a fix, it is a ugly workaround just to prove what
> the problem is! Even the position which I chose to set keymap->modifiermap[]
> is not a good one, because it will get set each time you release the key.
>
> I hope somebody else with more X knowledge can come up with a fix, I just
> tried to help you get a fix more quickly by pointing out the problem.
>
> Of course you can "fix" the keycode 116 too by doing a similar thing, but
> we better get a proper fix instead.
Researching a bit more about this issue, I found out that this is a X bug,
which is explained here:
http://lists.debian.org/debian-x/2007/02/msg00285.html
I used the suggested modification to .Xmodmap
remove mod4 = Super_L
keycode 127 = NoSymbol
add mod4 = Super_L
and after loading it with 'xmodmap .Xmodmap' wmaker recognizes
the release of Mod4 key, without any ugly workaround patch.
--
To unsubscribe, send mail to [EMAIL PROTECTED]
