Re: Occupy selector...

[mc]

[EMAIL PROTECTED] (Michael George) writes:

> Okay, I've checked out the patch proposal and I have the source for 3.6.
> However, where exactly does this patch go?  At the bottom of AddWindow()?  Or
> in the place of:
>       tmp_win->group = tmp_win->w/* NULL */;

I did added this to add_window.c. Note the part if I have preprocessed
out and the line just before it.

    if (tmp_win->wmhints && (tmp_win->wmhints->flags & WindowGroupHint)) 
      tmp_win->group = tmp_win->wmhints->window_group;
    else
        tmp_win->group = 0; /* Group ID should be a long */
#if 0
        /* MC: Original. */
        tmp_win->group = tmp_win->w/* NULL */;
#endif
        
    if (tmp_win->group && tmp_win->group != tmp_win->w)
    { 
        /* 
         * GTK windows often have a spurious "group leader" window
         * which is never reported to us and therefore does not
         * really exist. Look for * that here. It is in fact a
         * duplicate of the WM_CLIENT_LEADER * property.
         */ 

        TwmWindow *groupleader; /* the current twm window */ 
        stat = XFindContext(dpy, tmp_win->group, TwmContext, (XPointer 
*)&groupleader); 
        if (stat == XCNOENT) 
            groupleader = NULL; 
        if (!groupleader)
        { 
            tmp_win->group = tmp_win->w; 
        } 
    }