I found DWOP_SHAPED but thats not quite it.
There is I'm pretty sure a option to not recieve events if your 100% clear.
Dangit.
Anyway I gotta go to bed. I hope this info helps.

Mike


On 1/17/06, Mike Emmel <[EMAIL PROTECTED]> wrote:
> On 1/16/06, Alex Lau <[EMAIL PROTECTED]> wrote:
> > Mike Emmel wrote:
> > > On 1/16/06, Zhenghui Zhou <[EMAIL PROTECTED]> wrote:
> > >> I have try at alex's suggestion, it works now. Thanks very much.
> > >> But I'm not familiar with the directfb, so I am not sure is my fix right 
> > >> or
> > >> not.
> > > Alex I'm not sure what you fixed here its also broken on the simple
> > > window manager.
> > > Maybe we need to patch both. I know you and David were talking about
> > > these issues.
> > I think so ...
> > I see the tabbing bug now ...if I get the time ...I will look into it
> > The reason it is not able to tabbing over I think is the gdk widget list
> > having that :) non-visible or decorated surface again, or worst ... it
> > get lost in space  ( I only hack the mouse click on non-visible surface
> > not being getting the input channel but not the keyboard )
> >
>
> Hmm these invisible windows that are input only are there normally
> only when a window loses focus. There job is the first time a window
> gets focus or regians focus is to redirect the first click to the
> previous focus owner it catches this event which would otherwise be
> lost. Then the should be either hidden or destroyed. Hmmm
>
> > Reference  . Refs  Windows
> > --------------------------
> > 0x00000005 :  17    13
> >
> > -----------------------------------[ Windows
> > ]------------------------------------
> > Reference  . Refs     X     Y   Width Height Opacity   ID     Flags
> > ----------------------------------------------------------------------------------
> > 0x0000000a :   4    392,  351     12 x   18    0xff     1  -  VISIBLE
> > 0x00000058 :   4   1025,  769    168 x   36    0x00     7  ^  DECORATED
> > 0x000000fd :   4   1025,  769    168 x   36    0x00    13  ^  DECORATED
> > 0x00000170 :   1   -100, -100     10 x   10    0x00    15  -
> > 0x00000144 :   3    409,  320    206 x  128    0x00    14  -  DECORATED
> > 0x00000019 :  11    294,  201    435 x  365    0xff     3  -  VISIBLE
> >  DECORATED
> > 0x000000c0 :   1    357,  325    310 x  118    0x00    11  -  DECORATED
> > 0x000000b5 :   1    357,  325    310 x  118    0x00     9  -  DECORATED
> > 0x00000082 :   3    409,  320    206 x  128    0x00     8  -  DECORAT   ED
> > 0x00000035 :   2   1024,  768    112 x   30    0x00     6  -  DECORATED
> > 0x0000002f :   8   1024,  768     74 x   48    0x00     5  -  DECORATED
> > 0x00000023 :   2      0,    0   1024 x  768    0x00     4  -  DECORATED
> > 0x0000000d :   2      0,    0   1024 x  768    0x00     2  -  DECORATED
> >
> > see the -100x-100 10x10 window?
> > I'm not sure how does it get created... any idea? (I will look)
>
> That looks very much like either a windo left from debugging window or
> one created
> and then sized to a default size. I don't think it should really be there.
>
> As I said earlier these input only windows should not be hanging around.
> Also I have the windows that are clear because there not shown yet to
> not except input till alpha is greater than zero is there maybe a bug
> here ?
> Its getting late and I can't find the code but there is a option to say a 
> window
> does not receive events when its Opacity is 0x0 maybe I deleted the code ?
> I'm worried that windows are not getting destroyed also..
>
> Eeeeek,
>
> Mike
>
>
>
>
> > But what default directfb handle the input channel will pass everything
> > into that window, so you can't tab back ( because that window doesn't
> > handle tab with the gdk widget listting ) and mouse can't get out of it
> > ... since that layer is not even on surface
> > Beside the tab bug you get new widget bug also ... now if you keep
> > clicking Alt-N for new directory and Cancel it you will see alot of
> > surface being created .. but not reusing the old one.. is that right?
> >
> > >
> > > Mike
> > >
> > >> May I ask your help, Mr Mike?
> > >>
> > > Sure not a problem, I created  the bugs.
> > >
> > >> BTW, you may test with close the cursor with the directfb param 
> > >> "no-cursor"
> > >>
> > > I don't quite understand what your asking here. Can you explain more
> > > or send me and example. Are you saying disable the cursor ?
> > >
> > >
> > > I'm not sure if there is a way to simply turn off the cursor  Davide ??
> > >
> > >> Regards,
> > >>
> > >> Zhenghui Zhou
> > >>
> > >> On 1/16/06, Mike Emmel <[EMAIL PROTECTED]> wrote:
> > >>> On 1/16/06, Alex Lau <[EMAIL PROTECTED]> wrote:
> > >>>> In my Unique update already fix this problem ... but I'm not sure that
> > >>>> is a good fix ... just let you know
> > >>>>
> > >>>> fix gdk_window_set_keep_above
> > >>>> and
> > >>>> fix gdk_window_set_keep_below
> > >>>>
> > >>>> will give you what you wanted
> > >>>>
> > >>>>
> > >>>> Zhenghui Zhou wrote:
> > >>>>> Yes, I'm sorry, I just forgot to give the example before I sent it.
> > >> The
> > >>>>> "filesel" example following with the GTK+ can show.
> > >>>>>
> > >>> Hmm you may be right Alex. I ran testfilechooser and it shows problems
> > >>> similar to what your describing. But the filesel example is different
> > >>> the text box gets focus but you can't tab through it. I think
> > >>> testfilechooser is showing bugs because of keep above and below while
> > >>> filesel is some sort of focus traversal bug. Alex can you test
> > >>> examples/filesel with you patched window manager ?
> > >>>
> > >>> Mike
> > >>>
> > >>>>> Thanks,
> > >>>>> Zhenghui Zhou
> > >>>>>
> > >>>>>
> > >>>>> On 1/16/06, Mike Emmel < [EMAIL PROTECTED]> wrote:
> > >>>>>> Can you send a simple test program that illustrates the problem.
> > >>>>>> Also in gtk under test there are quite a few test case one may
> > >> already
> > >>>>>> show the bug.
> > >>>>>>
> > >>>>>>
> > >>>>>> Thanks.
> > >>>>>>
> > >>>>>> Mike
> > >>>>>>
> > >>>>>>
> > >>>>>
> > >>>>>
> > >> ------------------------------------------------------------------------
> > >>>>> _______________________________________________
> > >>>>> directfb-dev mailing list
> > >>>>> [email protected]
> > >>>>>
> > >> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
> > >>>>
> > >>
> > >>
> > >
> >
> >
>

_______________________________________________
directfb-dev mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev

Reply via email to