This is the first stage in tracking keyboard grabs in the common code.
This lets us handle destroying or unmapping virtual window with a
keyboard grab.
If we only convert the first then motion hint emulation won't
work since we don't see the next motion even until we've
fully handled this one.
However, this changes a behaviour that has been like this since
the mists of time. I don't know if it could cause other issues.
I haven't seen any yet though.
There was a performance problem with the old flicker fixing
approach. For moved windows we copied the window data to the double
buffer pixmap and then back to the window with the rest of the
expose data. In some cases the copy from window data to pixmap was
very slow because the pixmap was allocated in system memory and
the window in video memory.
The new approach is to delay all window moves and then replay them
after the expose has drawn to the double buffer pixmap but before
drawing it to the window. Furthermore, we remove all exposed areas
from the destination of the delayed moves so we won't copy something
just to then immediately draw over it.
This makes scrolling in firefox fast, and it makes tests/flicker not
show any (detectable) flicker.
We return the raw window drawable, so its likely the app will do some
weird stuff to it, like draw using non-gdk operations. We don't want
the app to see any half-drawn state, so flush everything.
This fixes a scroll issue in firefox at least.
This fixes a mismatch in save/release gstate for the CGContext. If
this function is ever used for pixmaps, we need to add support for
that, probably will have to do so soon.
The one that puts windows outside the screen and moves them in when
showing. It might not be needed anymore and the workaround doesn't
work with the client-side window branch anyway because the window is
already mapped when we get showed for some reason.
If we do this there is a short window where we think there is a grab, and
if we generate a client side pointer motion event in this time (e.g. from
a window move) we could be delivering that according to the grab. But in
the Xserver that grab is infinitely short as it sends Press and then
Release immediately.
To see a problem from this, try using the scrollwheel to quickly scroll
in the testgtk list of buttons.
This is basically the same fix as was done for
gdk_window_move_resize_internal. We make sure not to move the native
child window contents twice and we don't copy data that was overwritten
by the moving of the native child windows.
Whenever a native window is moved this causes an immediate change in
the window (the window content is copied). This change conflicts can
conflict with outstanding moves or other cached changed, so we need
to flush all outstanding moves in the related windows.
To simplify the code for window move/resize the toplevel version was
split out to its own function.
Move native windows after recomputing so that we get the right new
shape before moving (and the implied copy). This means we're not
copying too much data.
Take into account the area of a moved window that contains native
subwindows, as these affect things in two ways:
First of all we shouldn't copy the original window location, as that
is copied by the native window move.
Secondly, we can't copy things that would end up copying from the
native window move destination, as the data that used to be there is
now destroyed by the native window move.
It can happen that another native window is re-shaped over the region to
be moved, this will not destroy the data (since we're unsetting the background
when we reshape), but it will mean we need to read from this window.
We already used INCLUDE_INFERIORS, but that only handle subwindows. We fix
this by doing the copy on the toplevel, offsetting the copy to compensate for
this.
All the calls that unset private->parent failed if that was
not a native window (impl), instead we need to find the impl window
for the parent. Add some helper functions for this and use them.
For move/resize of child windows, we really need to recursively unset
on the parent, because moving the window could expose other native
children of the parent.
In do_shape_combine_region, only unset background if we're changing
the bounding shape (i.e. not the input shape)
There is no need to do all these computations when moving toplevels
as that can't really change any visible regions. Nor will it cause
any exposes we need to handle.
move_region_on_impl() - doesn't need to copy anything if dx/dy == 0
Ensure that we queue an update when invalidating an empty area but we have outstanding moves
Temporarily unset background when moving native child windows
The history before this was kind of twisted as several different
approaches were tested, so that was all squashed into this initial
commit to hide the uninteresting changes and files that were later
removed.
Revert commit f4090b5c84, as that
conflicts with the client-side-windows work. Furthermore, that
change is already applied as part of the client-side-windows work.
2009-03-27 Behdad Esfahbod <behdad@gnome.org>
Bug 437533 – Implement draw_shape PangoRenderer method
* gdk/gdkpango.c (gdk_pango_renderer_draw_shape),
(gdk_pango_renderer_class_init): Implement draw_shape
* demos/gtk-demo/rotated_text.c (fancy_shape_renderer),
(create_fancy_attr_list_for_layout), (rotated_text_expose_event),
(do_rotated_text):
1) Port to pangocairo
2) Also show a rotated label
3) Install a custom shape renderer on both. The second one goes
through gdkpango and hence tests the patch above. It's working.
svn path=/trunk/; revision=22592
2009-03-17 Tor Lillqvist <tml@novell.com>
Bug 575700 - Crash when moving a GtkWindow
* gdk/win32/gdkevents-win32.c (gdk_event_translate): Fix
embarrassing crasher bug introduced on 2009-03-11. A variable was
assigned a value only inside a GDK_NOTE but still used outside
that.
svn path=/trunk/; revision=22568
2009-03-11 Tor Lillqvist <tml@iki.fi>
Bug 573067 - Intra-app dnd of text behavior on Windows is wrong
* gdk/win32/gdkdnd-win32.c (local_send_motion) (gdk_drag_motion):
The default action should be move, not copy. Tweak how the
GdkDragAction fields in GdkDragContexts are changed. Seems to help
the problem.
Add more debugging printout for --gdk-debug=dnd to many functions.
* gdk/win32/gdkmain-win32.c
* gdk/win32/gdkprivate-win32.h: (_gdk_win32_drag_protocol_to_string)
(_gdk_win32_drag_action_to_string): New functions for use in
debugging printouts.
svn path=/trunk/; revision=22522
2009-03-11 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdnd-win32.c: Use G_DEFINE_TYPE. Some debugging
printout changes. Hacking on the OLE2_DND code, which still
doesn't work, though, and is not normally compiled.
svn path=/trunk/; revision=22517
2009-03-11 Tor Lillqvist <tml@iki.fi>
Bug 570896 - gdkevents-win32.c(2947) : error C4053: one void
operand for '?:'
* gdk/win32/gdkevents-win32.c (gdk_event_translate): Fix
compilation error with MSVC. While at it add clarifying comments
and fix a typo in the debugging output.
svn path=/trunk/; revision=22515
2009-03-11 Tor Lillqvist <tml@iki.fi>
Bug 574794 - Can not resize shaped windows on Windows
* gdk/win32/gdkevents-win32.c (gdk_event_translate): On handling
WM_WINDOWPOSCHANGED, don't do anything if the position and size
hasn't changed. I am not exactly sure I understand why this helps
this particular bug.
svn path=/trunk/; revision=22511
GtkNotebook tabs
* gtk/gtkwidget.c (_gtk_widget_synthesize_crossing): Be a little
more careful when navigating up the parent chain.
* gdk/gdkwindow.c (_gdk_window_destroy_hierarchy): Reset the
pointer window thats stored on the screen when it is destroyed.
svn path=/trunk/; revision=22420
* gdk/x11/gdkevents-x11.c:
* gdk/x11/gdkscreen-x11.[hc]: Another attempt at getting Xrandr
screen size changes right. Actually handle XRRScreenChangeNotify
events, and ignore root ConfigureNotify events if we are using
Xrandr. Only emit size-changed signals if the screen size changed,
and only emit monitor-changed signals if the monitors changed.
svn path=/trunk/; revision=22398
2009-02-19 Ray Strode <rstrode@redhat.com>
* gdk/x11/gdkscreen-x11.c
(monitor_compare_function), (init_randr13):
After using randr to get a list of monitors,
sort the list such that the biggest output
of "cloned" outputs comes first in the list.
This helps apps that don't generally handle
overlapping outputs to work better in randr
clone mode.
svn path=/trunk/; revision=22386
2009-02-19 Sven Neumann <sven@gimp.org>
* gdk/directfb/gdkwindow-directfb.c
(gdk_directfb_window_move_resize): looks like being called with
negative width and height should be interpreted as a move. Fixes
scrolling and makes popups work to some extent.
(_gdk_directfb_move_resize_child): formatting.
svn path=/trunk/; revision=22384
2009-02-19 Sven Neumann <sven@gimp.org>
* gdk/directfb/gdkdisplay-directfb.c (gdk_display_open): move
the
call to _gdk_windowing_window_init() up a little. Fixes a crash
when running on pseudo-color visuals. Thanks to Masse Nicolas
for
pointing this out.
svn path=/trunk/; revision=22382
2009-02-19 Sven Neumann <sven@gimp.org>
* gdk/directfb/gdkwindow-directfb.c
(gdk_window_foreign_new_for_display, gdk_directfb_window_new):
initialize private->impl. Thanks to Masse Nicolas for pointing
this out.
svn path=/trunk/; revision=22380
2009-02-18 Sven Neumann <sven@gimp.org>
Bug 533456 – Memory leak when window is destroyed
* gdk/directfb/gdkdirectfb.h
* gdk/directfb/gdkprivate-directfb.h
* gdk/directfb/gdkcursor-directfb.c
* gdk/directfb/gdkevents-directfb.c
* gdk/directfb/gdkwindow-directfb.c: removed all #ifdef checks
for
DIRECTFB_MAJOR_VERSION >= 1. We depend on DirectFB >= 1.0.0 now.
Fixes a potential build issue that would lead to a memory leak
as
described in bug #533456.
svn path=/trunk/; revision=22367
2009-02-18 Sven Neumann <sven@gimp.org>
Bug 513422 – Unknown meaning of translatable messages
* gdk/directfb/gdkdisplay-directfb.c: removed broken and
commented
code and the translatable messages that belonged to it.
svn path=/trunk/; revision=22365
2009-02-18 Sven Neumann <sven@gimp.org>
Bug 560671 – the client_message API doesn't work in DirectFB
* gdk/directfb/gdkevents-directfb.c: applied a slightly cleaned
up
patch from Steve Lhomme that adds this functionality using
DirectFB's DFBUserEvent.
svn path=/trunk/; revision=22363
2009-02-18 Sven Neumann <sven@gimp.org>
Bug 543710 – Dead keys do not work anymore with DirectFB
* gdk/directfb/gdkkeys-directfb.c (gdk_directfb_translate_key):
applied patch from Josselin Mouette that removes special
handling
of dead keys as that is done by DirectFB for a long time
already.
svn path=/trunk/; revision=22361
2009-02-18 Sven Neumann <sven@gimp.org>
Bug 554407 – directfb backend does not implement GdkWindowImpl
* gdk/directfb/gdkgeometry-directfb.c
* gdk/directfb/gdkprivate-directfb.h
* gdk/directfb/gdkwindow-directfb.c: fix the build of the
DirectFB
backend and resurrect basic functionality. Based on a patch from
the Ubuntu bug-tracker and work by Masse Nicolas.
svn path=/trunk/; revision=22357
* configure.in: Check for Xrandr 1.3
* gdk/x11/gdkdisplay-x11.[ch]: Rename have_randr12 to have_randr13
and set if if we have randr 1.3
* gdk/x11/gdkscreen-x11.c: Go back to using randr for monitor
information, if we have XRRGetScreenResourcesCurrent.
svn path=/trunk/; revision=22327
2009-02-09 Cody Russell <bratsche@gnome.org>
* gdk/win32/gdkwin32.h: Add potentially missing #defines for
XBUTTON1, etc. Fixes the build that was broken in r21746.
svn path=/trunk/; revision=22297
2009-01-29 Tor Lillqvist <tml@novell.com>
Bug 559408 - Transparency lost when images are copied between GTK+
and the clipboard on Windows
The code in gdkselection-win32.c is simplified quite a lot. When
an image is stored on the clipboard from GTK+, only the "PNG" and
CF_DIB formats are advertised. PNG is our preferred format because
it can losslessly represent any image that gdk-pixbuf can handle,
even with alpha, unambiguously. CF_DIB is also advertised because
of its ubiquitous support in Windows software.
Delayed rendering is used as before, so for other data than text
nothing is actually stored on the clipboard until some app asks
for it.
It's pointless to advertise images formats like "image/png" or
"image/x-MS-bmp" that no other Windows application is going to
look for anyway. Likewise it's pointless to store text under
formats like "text/plain" or "UTF8_STRING". Just store
CF_UNICODETEXT. CF_TEXT will be available as a synthesized format
anyways.
Office 2007 offers and asks for images in PNG format using the
registered clipboard format name "PNG", so we do likewise. If the
"PNG" format is available that is preferred when copying to GTK+
from the clipboard.
Unfortunately Word 2007 still uses CF_DIB without handling
alpha. But PowerPoint 2007 uses PNG if available and handles alpha
correctly. Both Word and Powerpoint offer PNG with alpha nicely.
Firefox and IE7 offer images that contain alpha as 32-bit version
3 BI_RGB DIBs with the undocumented "extra" bytes actually being
alpha. Also, alpha is premultiplied into the RGB bytes, presumably
because that is how AlphaBlend() wants such DIBs. That is also
taken care of. At least for Firefox it easy to be sure that a
CF_DIB on the clipboard is from Firefox.
Also some general stylistic cleanup, comment improvements, and
improvements of debugging printout especially in the clipboard
handling. Those are not detailled below.
* gdk/win32/gdkprivate-win32.h
* gdk/win32/gdkglobals-win32.c
* gdk/win32/gdkgmain-win32.c: Move some globals that were used
only in gdkselection-win32.c to be local to that file.
* gdk/win32/gdkproperty-win32.c (gdk_property_change): Don't
bother checking if text to be placed on the clipboard consists of
only ASCII.
* gdk/win32/gdkselection-win32.c: Add static variables for a list
of GdkPixbuf-supported formats, well-known registered clipboard
formats for images, and for GdkAtoms for well-known image and text
formats.
(_gdk_win32_selection_init): Initialize above static variables.
(selection_property_store) (gdk_selection_property_get)
(_gdk_selection_property_delete): Don't use a FIFO of GdkSelProps
for a window after all, it breaks testtext. See bug #163844.
(gdk_selection_convert): When converting to the TARGETS format,
i.e. when the caller wants to know what clipboard formats are
available, if PNG is available we report just that and skip
CF_DIB, JPEG and GIF even if advertised.
If CF_UNICODETEXT is available, report only UTF8_STRING.
When converting to the UTF8_STRING format, i.e. when the caller
wants text from the clipboard, try just CF_UNICODETEXT. There is
no point in trying CF_TEXT as Windows will synthesize
CF_UNICODETEXT from CF_TEXT anyway, if some app has stored just
CF_TEXT.
When converting to the image/bmp format, i.e. when the caller
wants an CF_DIB image from the clipboard, we check if the DIB is a
version 3 32-bit BI_RGB one that is likely to actually contain
alpha in the "extra" bytes. Such a DIB is likely to have
premultiplied alpha even, at least in the case of Firefox 3 and
IE7. We then edit the DIB in-place into a version 5 one in
BI_BITFIELDS format and undo the alpha premultiplication.
In any case, prepend a BMP file header before letting go of the
data which will be fed to the gdk-pixbuf bmp loader by upper
levels.
(gdk_win32_selection_add_targets): If some kind of pixmap image
format is being added, actually advertise just PNG and
CF_DIB. Note that alpha won't be stored on the clipboard through
CF_DIB. This is because gdk-pixbuf's bmp loader doesn't save
alpha. Furthermore, few if any non-GTK+ Windows apps seem to
understand a version 5 DIB with proper alpha anyway.
(_gdk_win32_selection_convert_to_dib): Simplified muchly.
svn path=/trunk/; revision=22255
2009-01-29 Claudio Saavedra <csaavedra@igalia.com>
Bug 569435 – make maintainer-clean removes non-generated sources
* gdk/Makefile.am: Do not remove gdkprivate.h and gdkwindowimpl.h
during make maintainer-clean, as these files are not generated.
svn path=/trunk/; revision=22248
2009-01-26 Richard Hult <richard@imendio.com>
Bug 566628 – gdk_display_close always asserts on win32 and quartz
* gdk/quartz/gdkdisplay-quartz.c
(_gdk_windowing_set_default_display): Copy fix for bug #566628
from the win32 backend: Allow also a NULL parameter in the
g_assert().
svn path=/trunk/; revision=22223
* gtk/gtkwidget.c (gtk_widget_get_snapshot): Use a different approach
to snapshotting that is in line with what is done in the client-side
windows branch, and that works for widgets regardless if they are
double-buffered or not. Patch by Alex Larsson.
* gdk/gdkwindow.c (_gdk_window_calculate_full_clip_region): Clip
to the parent.
svn path=/trunk/; revision=22187