[gtk+] Created tag 2.17.3

[Matthias Clasen <matthiasc src gnome org>]

The unsigned tag '2.17.3' was created.

Tagger: Matthias Clasen <mclasen redhat com>
Date: Tue Jul 7 01:06:58 2009 -0400

    GTK+ 2.17.3

Changes since the last tag '2.17.2':

Alexander Larsson (252):
      Revert patch that conflicts with client-side-windows work
      Initial client-side-windows work
      Add offscreen testcase
      Add testcase showing Gtk+ flickering a lot
      Create native window when requesting xid
      Export gdk_window_set_has_native
      Really create a native window when getting xid
      For foreign parents, use the real parent not the root (which is what we list as parent in the APIs)
      Never set shapes on foreign windows
      Make synchronized_configure work for child windows (used by in-process plugs) and ensures a native window exist
      Change assert to warning with some information
      Make sure we change the event mask on native window if they change between toplevel and child
      Grab pointer on toplevel window, not on nearest native, so that we always get all events on toplevel
      Fix up keyboard grab handling
      Ensure we have native window before reparenting to toplevel
      Add g_warning to not yet supported code
      Avoid crash by not passing in NULL mask pointer to _gdk_windowing_get_pointer
      Actually report x/y_root against the root window
      Fix coordinates reported from gdk_window_get_pointer (had sign reversed)
      Move clip offsetting into _gdk_gc_intersect_clip_region
      Move drawable clip tracking into GdkGC
      Move _gdk_gc_remove_drawable_clip from after each time its set to before each pixmap set and when adding a new different region
      Track clip_mask in GdkGC. Fix gdk_gc_add_drawable_clip to handle clip masks
      Don't convert foreign windows to child when reparented to non-toplevel
      gdk_pixmap_draw_pixbuf - gc may be NULL
      gdk_pixmap_new - drawable may be NULL
      gdk_bitmap_create_from_data - drawable may be NULL
      Fix draw_drawable being offset when using window as source (testgtk text test)
      The double get_composite_drawable case can happen for pixmap targets too, so move the code to gdk_draw_drawable
      Move gdk_window_get_deskrelative_origin to the GdkWindowImpl and make the common code handle abs_x/y
      Clear to end of window if passing zero width or height to gdk_window_clear_area (fixes clist/ctree clearing)
      Handle gdk_window_get_impl_window on root window
      Fix some return value warnings
      Don't sent native MotionHint masks on pointer grabs as then we won't get enought motion events reported.
      When reparenting, remove from old parent before setting the new parent
      Store colormap on impl window (necessary for e.g. foreign windows)
      Don't draw on input_only windows
      Ensure native window if composite requested
      Don't include composited childen in window clip region, update region when setting composited
      Move flushing code to non-implicit-paint only call.
      Don't actually clear background when mapped, as that causes ugly flashed. Just queue a repaint.
      Share the code to copy impl window regions w/ update_area
      Don't access event after freeing it
      Add _gdk_pixmap_set_as_backing which is a way to get GraphicsExposures on pixmaps reported to the window they are backing store for
      Implement flicker free handling of window moves/resizes
      gdk_window_draw_points() - make sure we use the offset points
      New approach to motion event handling
      Fix crossing event order and some details
      Implement cursors
      Handle cursors right with grabs
      For implicit grabs, we grab the *event* window, not the pointer window
      Remove unnecessary clear_area method in GdkWindowImpl
      Make _gdk_windowing_window_queue_translation a GdkWindiwImpl call
      Make gdk_windowing_window_destroy a GdkWindowImpl method
      Create _gdk_drawable_create_cairo_surface to ugly typechecking
      Add _gdk_windowing_window_get_shape
      Implement shaped windows
      Remove shape_combine_mask from backend. Make X11 backend use XShapeCombineMask directly to unset region
      Implement input shapes
      Handle mask == NULL in shape_combine_mask
      Don't remove composited childrens area when processing updates
      Ensure that we're properly hiding impl window if one of its parent non-impl windows are hidden
      Move all native children (recursively) in gdk_window_scroll
      Update abs_x/y before moving native children
      Various fixes for native windows:
      Ensure that we always invalidate newly exposed areas, as X may not (due to temp bg none)
      Move native window showing/hiding out of is_impl, we may need to do this for emulated windows too
      Ensure we show native window when creating them for already visible windows
      Fix gdk_window_get_root_origin and gdk_window_get_frame_extents when called on non-toplevels
      Don't destroy NULL region
      Recompute abs_x/y before reparenting windows due to new native window parent
      Avoid flickering by using more temp unset bg
      Need to split up impl changing and reparenting so we can recompute abs_x/y inbetween
      When creating native window, set the native background
      Correct type checks on gdk_window_set_group entry.
      Ref imp even in non-native windows and don't NULL out in destroy so that get_screen works after destroy.
      tests/testwindows - play with GdkWindows
      Set bg none while changing so we don't draw the bg color
      Add  _gdk_drawable_get_subwindow_scratch_gc as a way to get cached gcs with GDK_INCLUDE_INFERIORS
      Include inferiors when moving window parts
      Handle GDK_WA_CURSOR in non-native window creation
      Make xfree gdkinput code compile
      Force native windows for all windows used with input events
      Send expose events to exposed area when changing the window shape
      Make sure we handle clipping in gdk_window_draw_pixbuf if gc == NULL
      Move all backend specific raise/lower code into _internal helpers
      Add restack_under method to GdkWindomImpl
      Fix up gdk_window_raise and native window creation to not move native window above windows outside the non-native parent
      Fix up native window handling in gdk_window_lower too
      Restack native window after reparent put it on top
      Avoid lots of unnecessary computation when moving toplevel windows
      Always set the shape when creating a native window.
      Make testwindows support multiple selected windows
      Don't flicker when unsetting shape (using bg none)
      Avoid extending implicit paint size due to empty regions not working well with gdk_rectangle_union on clipbox
      Fix various problem with temporarily unsetting background
      move_region_on_impl - shortcut if region is empty
      do_move_region_bits_on_impl needs to copy from overlapping native windows too
      Make moving subwindows correctly handle native windows
      Make _gdk_x11_window_tmp_unset_parent_bg not change above the toplevel window
      Call backend move_resize with the right coords/size
      Fix resize of native children
      Fix performance issue with find_native_sibling_above
      Ifdef out x11 specific code
      Add support for scrolling to testwindows
      Handle native children correctly in gdk_window_scroll
      Don't expose a temporary grab during a scrollwheel button press/release cycle.
      Don't send expose events if region is empty
      Fix typo in pixmap window backing
      testwindows - support for manually selecting window pos and size
      Don't allow native window > 65535 pixels
      Clip exposed area to the visible region.
      Flush any outstanding stuff in the window in non-buffered get_internal_paint_info case
      Split out the flushing of outstanding moves
      gdk_window_process_updates_internal only needs to flush outstanding moves
      New flicker flicker fixing approach
      Implement GDK_POINTER_MOTION_HINT_MASK support
      Convert all pending X events in _gdk_events_queue
      Move keyboard grab info to common code
      Move keyboard grab state tracking code to common code
      Ungrab keyboard if virtual window is hidden or destroyed
      Keep track of the impl windows wrapper directly with a ref
      Revert "Convert all pending X events in _gdk_events_queue"
      Switch motion hint emulation to be serial based.
      Only set cursors on toplevel.
      Fix gdk_window_has_no_impl typo
      Add _gdk_x11_roundtrip_async
      Track pointer grabs on Xserver time
      Don't ignore native CROSSING_GRAB/UNGRAB events
      Fix typos in X11 details of new OSX APIs
      Don't look for EXPOSE mask set to disable bg none setting
      Fix setting of private->shaped when setting input shape
      Avoid warnings about non-handled GDK_EVENT_LAST enum
      Remove unused label
      Fix warning due to typo
      Fix warning about wrong type in assignment
      Remove unused code
      Fix warning about XChangeProperty argument type
      Remove_gdk_pixmap_set_as_backing etc, as this is not used anymore
      Simplify do_move_region_bits_on_impl by removing unused arguments
      Don't recursively set bg none on the parent when e.g. moving children
      Don't crash if window->parent is NULL
      Don't mess with any events on the root window.
      Make gdk_window_get_toplevel() for CHILD window with root as parent
      Fix typo in gdk_window_input_shape_combine_region
      Update clip region etc, when creating a foreign window
      Don't warn about unexpected events on foreign windows
      Sync the display when creating a native window.
      Only sync the display if creating a native window implicitly
      Move bitfield in GdkWindowImplX11 to make struct more compact
      Fix up indentation
      Ref windows during _gdk_x11_roundtrip_async
      New approach for grab tracking code
      Don't send crossing events to destroyed windows
      Move implicit grab tracking totally to common code
      Report the steady state in gdk_display_pointer_is_grabbed
      Make sure we report grab broken on window destroy
      enable motion hints - use last sent request
      Handle crossing events with subwindows unknown to gdk
      Return ignoring of native grab/ungrab events when we don't have a grab
      Filter out all native grab/ungrab events with detail INFERIOR
      Enable custom event masks for native windows
      Workaround X11 clipping bug
      Fix assert, it was checking the wrong window
      Don't set the background for input only window
      Only translate native motion events to motion events
      Destroy native children when recursing from a destroy on a virtual window
      Be more explicit about when to flush moves
      Refactor background clearing code so we can clear a general region
      Call backend show() in gdk_window_show even if window is mapped
      Clear background on expose with no EXPOSURE_MASK
      Don't touch private->parent after its been change
      Don't event process updates for foreign windows
      Don't clear background on no exposure mask if NULL background pixmap set
      Make outstanding window moves work with the new model
      Fix order of outstanding moves in queue
      queue expose translations after the actual pixel copy
      Use clearer names for GdkWindowRegionMove members
      Move GdkWindowRegionMove construction/destruction to separate functions
      Remove invalid source area from copy in move_region_on_impl
      Move unnecessary window copies to the right place
      Free region also if its empty (we're taking ownership now)
      Try to combine consecutive window moves into one
      Fix x vs y typo
      Make offscreen windows not be in hierarchy
      Remove offscreen hooks support
      Move the new parts of GdkWindowObject to a private header
      Ensure offscreen windows are not used as non-toplevels
      Fix coordinates for background clearing when redirecting
      Merge branch 'master' into client-side-windows
      Handle flushing implicit paint while there is a paint outstanding
      Add gdk_region_rect_equal to compare a region with a rect
      Add support for visibility notification
      Add _gdk_gc_get_exposures
      Add support for client side graphics exposures in gdk_draw_drawable
      Add some test for draw_drawable graphics exposures
      Initial version of input support
      Make disabled xinput work again
      Rename gdk_window_set_has_native to gdk_window_ensure_native
      Whitespace cleanup gdkwindow.c
      Add more GdkWindow implementation docs
      Fix up gdk_window_set_has_native rename issues
      Further clean up window drawing redirection/clipping
      Fix some warnings added by the input handling
      Fix too many args warning
      Update GdkOffscreenWindow to new draw_drawable prototype
      Set the right parent root window for the offscreen windows
      window_get_pointer should return the direct child
      Make window_get_pointer a window impl virtual
      Add signals for offscreen window embedding
      Implement offscreen get_pointer with offscreen signals
      Base pointer grab on get-offscreen-parent
      Implement the new offscreen signals in GtkOffscreenBox
      Update event emulation to handle offscreen children
      Add gdk_window_get_root_coords
      Use gdk_window_get_root_cords to get popup position for combobox
      Add gdk_window_offscreen_children_changed call
      Send crossing event due to geometry change in idle
      Handle the window being destroyed in do_synthesize_crossing_event
      Handle gdk_window_beep on offscreen windows
      Always return FALSE from idle callback to avoid loop
      Remove ununsed variables
      Don't reset window hints when showing multiple times
      Merge branch 'master' into client-side-windows
      Fix spelling of synthesize
      Simplify _gdk_display_get_last_pointer_grab using g_list_last
      Rename accumulate_get_parent to accumulate_get_window
      Use g_signal_emit instead of g_signal_emit_by_name when possible
      Remove old obsolete comment
      Factor out common code to sync_native_window_stack_position
      Fix typos
      Fix g_signal_emit calls
      Ensure that we always calculate clip regions for root windows
      Clip cairo drawing against client-side child windows
      Implement gdk_window_clear_area natively for foreign windows
      Better implementation of native clear_area
      Don't hide foreign children when emulating visibility
      Move destroyed check to common code for get_origin & get_root_coords
      Add gdk_window_get_root_coords to gdk.symbols
      Add missing symbols to gdk.symbols
      Don't return value from void function
      Don't use GDK_WINDOW_SCREEN, as its an X11 specific macro
      Make win32 backend build and minimally work
      Fix win32 GC set_region handling of reset_region
      Some win32 cleanups
      gdk_win32_blit, don't get the size of the impl
      Remove debug spew, add TODO_CSW
      Clean up embedding api
      Move new draw_drawable argument to end to make it more backwards compat
      Merge branch 'client-side-windows'
      Make GdkDrawable draw_drawable backwards compat
      Hack that works around firefox plugin crash

Bastien Nocera (1):
      Bug 461944 â?? pressing the volume icon in full screen shuts down the sound

Benjamin Otte (8):
      fix compile warnings
      remove unused variables
      unconditionally set right alignment for the size column
      clarify the init value for the default sort column id
      check for allowed NULL inside g_return_if_fail() statements
      make sure the "Create folder" button doesn't get shown accidentally
      make the size column resizable
      clean up code in gtk_list_store_set_n_columns()

Cody Russell (7):
      CSW Win32 work in progress - builds but does not work yet.
      Misc stuff
      Get width/height from GetSystemMetrics()
      Updates
      grabs
      Grab changes
      fix wrap width calculations that caused item height to be wrong

Daniel Nylander (1):
      Updated Swedish translation

David Zeuthen (2):
      Bug 586691 â?? Better GtkInfoBar defaults
      Bug 587485 â?? GMountOperation::show-processes support

Federico Mena Quintero (3):
      Don't set the sort column when there is no model
      Show the size column by default in the file chooser
      bgo#586315 - gtk_file_chooser_list_shortcut_folders() was crashing

Hans Breuer (10):
      Conditionally define WS_EX_COMPOSITED
      Updated msvc build
      gtk_status_icon_get_title() must return a value
      Undef USE_MMX for msvc build
      Use G_PI rather than M_PI
      Fix includes for missing prototypes
      Initialize 'private' before use
      Enable window creation with GDK_WINDOW_CHILD again
      Fix condition to deliver root coords
      Updated msvc build

Hendrik Richter (1):
      Updated German translation.

Ivar Smolin (1):
      Updating Estonian translation

Jorge Gonzalez (2):
      Updated Spanish translation
      Updated Spanish translation

Marco Túlio Gontijo e Silva (1):
      Updated Brazilian Portuguese translation

Matthias Clasen (37):
      Bump version to 2.17.3
      Don't use deprecated api internally
      Forgot one use in libgail
      More instances of those deprecated functions
      More uses of the get_cell_renderers functions
      Fix an oversight in mnemonic handling
      Correct a couple of typos
      Add a title property to GtkStatusIcon
      Reduce roundtrips
      Update some links in the tutorial
      Fix 'Add to bookmarks' initial sensitivity
      Remove extra padding from comboboxes
      Don't force toggle size for tabular menus either
      Forgotten header file
      Add a way to turn off toggle spacing in menus
      Demo setting an output name for print to file
      Whitespace cleanups
      Improve handling of print-to-file uri
      Improve printing example
      Add a testcase for use-stock/use-underline in buttons
      Make stock buttons have mnemonics again
      Update for 2.17.3
      Add new symbols
      Add doc stubs
      Add another bug ref
      Small documentation fixes
      More small documentation fixups
      More documentation fixes
      Fix the doc build
      Fix redraw issues in GtkInfoBar
      Use gdk_display_sync instead of XSync
      Update NEWS
      Fix a typo
      Make distcheck pass
      Temporarily turn of tests
      Fix some distcheck issues
      2.17.3

Michael Natterer (1):
      Fix warning about uninitialized variable plus spacing and indentation cleanup

Nguy�n Thái Ng�c Duy (1):
      po/vi.po: updated Vietnamese translation

Richard Hult (93):
      Add reset_origin argument to _gdk_windowing_gc_set_clip_region
      Include check for offscreen windows in WINDOW_IS_TOPLEVEL
      Don't init size in gdk_window_impl_quartz_init
      Remove get_size and get_visible region
      Remove width/height from GdkWindowImplQuartz and use those in GdkWindowObject
      Update _gdk_window_impl_new and add private variables where we changed impl to private
      Rename _gdk_windowing_window_destroy to _gdk_quartz_window_destroy and add to the impl iface
      Update signature for gdk_window_quartz_show and adapt the implementation
      Update gtk_window_quartz_hide
      Update gdk_window_quartz_reparent
      Remove gdk_window_quartz_clear_area
      Update gdk_window_focus
      Update gdk_window_set_type_hint and gdk_window_get_type_hint
      Update set_urgency_hint/geometry_hints/modal_hint/set_skip_taskbar_hint/set_skip_pager_hint
      Add more toplevel window checks
      Update gdk_window_quartz_set_background and gdk_window_quartz_set_back_pixmap
      Update gdk_window_get_deskrelative_origin, gdk_window_get_root_origin, gdk_window_get_frame_extents
      Add mask argument to _gdk_windowing_window_at_pointer, and update more checks
      Update more toplevel checks and update set_background and back_pixmap
      Remove gdk_window_set_child_input_shapes
      Remove gdk_display_pointer_is_grabbed and gdk_pointer_grab_info_libgtk_only
      Implement _gdk_windowing_create_cairo_surface
      Add stubs for queue_translation/antiexpose
      Add quartz window check macro
      Setup iface for queue_translation/antiexpose
      Change window check to quartz window check in gdkinput
      Offset subviews with the parent's abs coordinate
      Fix bug from first patch round, get private from window, not from impl
      Handle NULL event mask in _gdk_windowing_window_at_pointer
      Remove the old workarond for broken tracking rects
      Update the window size after getting windowDidResize
      Rework the event translation to look more like X11
      Make the various fill_*_event functions consistent
      Clean up the scroll support a little
      Add enter/leave notify to get_event_mask_from_ns_event now that those are handled like other events
      Only send map events for toplevels
      Add fill_crossing_event to be used soon
      Clean up some whitespace
      Rename _gdk_quartz_events_send_map_events to ..._event
      Get rid of all the synthesized crossing events except for ANCESTOR one on the toplevel
      Don't use drawable_get,release_context in _gdk_windowing_create_cairo_surface
      Only setup tracking rect for toplevels, we only want enter/leave for those
      Remove autoreleasepool in drawRect, not necessary
      Fix typos that broke loading without lazy symbol lookup
      Remove unused function _gdk_quartz_window_scroll
      Attempt to implement draw_drawable with src = dest = window
      First cut at getting events working
      Add serial argument to _gdk_windowing_got_event
      Remove gdk_keyboard_grab_info_libgtk_only which moved to the common code
      Add gdk_windowing_window_get_next_serial
      Setup ->impl_window on the root window
      Add a window argument to begin_paint_region in the paintable interface
      Expose _gdk_window_process_updates_recurse for backends
      Remove invalidate_maybe_recurse and process_updates from the paintable interface
      Call _gdk_window_process_updates_recurse() in drawRect instead of creating our own expose events
      Update for changes in the paintable interface
      Get the correct root coordinates for events
      Use event_type we already have instead of regetting it
      Leave AppKit events unhandled
      Get implicit grabs working
      Remove unused remainders from the old implicit grab code
      Fix coords returned by _gdk_windowing_window_get_pointer, fixes scribble demo
      Apply non-click-through policy on unfocused windows even if the app is active
      Use isKeyWindow instead of isMainWindow for the non-click-through check
      Use the common pointer grab code instead of tracking it ourselves
      Remove unused variable
      Cast to avoid warning
      Replace quartz specific keyboard grab code with common code
      Use cooca to convert coordinats from event window to grab window instead of homegrown code
      Use _gdk_display_unset_has_*_grab in break_all_grabs()
      When breaking grabs on deactivation, the unset should be implicit
      Ignore all events if the app is not active
      Update for latest changes in the common code for grab tracking
      Remove old cursor setting and make it work with client-side windows
      Remove tracking of "current mouse window", this is handled in the common code now
      Button press and release can share the same fill_button_event call
      Remove unused code and fix some indentation
      Remove re-declared variable
      Remove workaround for missing enter events on newly popped up windows
      Relax the check for ignoring events above the content view
      Don't ignore all events when inactive, only clicks
      Make owner_events grabs work again, use the pointer window, not the event window
      Try to reduce risk of triggering the "beam sync" penalty in quartz
      Get the right event window for non-grabbed windows too
      Disable/enable screen updates around process_all_updates to speed up rendering
      Limit manual flushing so we don't get hit by quartz' fps limiting
      Only add up native subwindow offsets for gdk_window_quartz_get_origin
      Improve explicit drawing and flushing (fixes GtkRuler)
      Tweak the click-through handling so active apps get it (but not inactive ones)
      Rename gdk_window_beep to _gdk_windowing_window_beep
      Update gdk_window_quartz_show, ..._get_origin, ..._get_pointer
      Remove duplicate definition of GdkInputWindow
      Update _gdk_windowing_pointer_grab

Ryan Lortie (1):
      Use g_mapped_file_unref()

Stanislav Brabec (1):
      bgo#529908 - Sanitize handling of motion events in GtkNotebook

Tristan Van Berkom (1):
      Accelerators failed for submenus (GNOME bug 582025)

Vincent Untz (1):
      Use g_return_val_if_fail instead of g_return_if_fail

Yaron Shahrabani (1):
      Updated Hebrew translation