2003-09-02 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdisplay-win32.c (gdk_display_open): Don't call
CloseHandle() on the HMODULE returned from GetModuleHandle().
Didn't cause any harm, but didn't do anything useful either. When
running a GTK+ program under MS's debugger, the CloseHandle() call
would cause a "first-chance exception" in ntdll.dll to show up.
(The exception is caught appropriately if you let it proceed, but
it confuses the person using the debugger). Thanks to Bruce
Hochstetler for noticing.
Mon Sep 1 19:39:14 2003 Soeren Sandmann <sandmann@daimi.au.dk>
* gdk/gdkwindow.c (gdk_window_get_bg_gc): Fix bug where the
background pixmap was being offset twice, once by the ts_origin in
gdk_window_get_bg_gc(), and once by gdk_draw_rectangle().
Tue Aug 26 09:57:38 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkkeys-x11.c (update_keymaps): Fix variable
declaration not at start of block. (#120371, Damien Carbery)
Mon Aug 25 16:15:41 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkkeys-x11.c: Fix up CapsLock and
NumLock handling for non-XKB. Add a special case
hack for NumLock on Sun servers. (Patch from
Robert Basch, #115819)
Thu Aug 21 15:17:42 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkkeys-x11.c: Change the interpretation
of consumed_modifiers so that it contains:
- Modifiers combinations actually found in
state.
- Single modifier modifier combinations.
But not multi-modifier combinations that aren't
in event->state. Document. (#100439)
Wed Aug 20 22:04:47 2003 Soeren Sandmann <sandmann@daimi.au.dk>
* gdk/gdkwindow.c: Make the layers in the paint stack have their
own pixmap instead of sharing one.
* tests/testgtk.c (create_get_image): Fixes to make the window fit
on an 800x600 screen
Tue Aug 19 13:58:50 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkgeometry-x11.c (gdk_window_scroll):
Don't use impl->position_info.clip_rect for toplevels.
(#107068, reported by Thomas Mirlacher)
Mon Aug 18 11:48:51 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdrawable-x11.c (convert_to_format):
Add a fallback for unaligned source data for
FORMAT_ARGB_MASK. (#117217)
Sat Aug 16 10:34:49 2003 Owen Taylor <otaylor@redhat.com>
* modules/input/imcedilla.c: Make the list of default
languages more comprehensive. (Suggestion of
Fco. Javier F. Serrador)
Sat Aug 9 12:47:11 2003 Owen Taylor <otaylor@redhat.com>
* gdk/gdk.c (gdk_arg_context_parse): Fix broken
parentheses when skipping args.
Fri Aug 15 12:34:04 2003 Owen Taylor <otaylor@redhat.com>
Fixes for X/Cygwin builds of GTK+ (#107623, Masahiro Sakai):
* configure.in (GDK_DEP_LIBS_FOR_X): Make sure that
we link libgtk against X explicitely, since we
make GTK+ calls for plug/socket.
* gtk/queryimmodules.c: Use USE_LA_MODULES and
G_MODULE_SUFFIX here.
* gdk/x11/gdkdrawable-x11.c (gdk_x11_draw_image): Protect
XShmPutImage with #ifdef USE_SHM.
2003-08-09 Noah Levitt <nlevitt@columbia.edu>
* gdk/gdkevents.h:
* docs/reference/gdk/tmpl/windows.sgml: Elaborate a bit more in the
docs on GdkFilterFunc and GdkFilterReturn.
2003-08-08 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c (print_event): Print also the root
coordinates for events that have such. Print coordinates for enter
and leave events.
(gdk_event_translate): Don't use event uninitialixed in the
return_exposes branch of the WM_PAINT handler.
* gdk/win32/gdkwindow-win32.c (gdk_window_new,
_gdk_windowing_window_get_pointer, _gdk_windowing_window_at_pointer):
Must offset top-level window coordinates here, too.
2003-08-07 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkinput-win32.h
* gdk/win32/gdkinput-win32.c (_gdk_input_configure_event,
_gdk_input_enter_event): Drop the GdkEvent* parameter, it wasn't
used.
* gdk/win32/gdkevents-win32.c (gdk_event_translate): Adapt caller
accordingly, in fact an uninitialised variable was dereferenced.
[Win32] Add support for multiple monitors.
* gdk/win32/gdkprivate-win32.h
* gdk/win32/gdkglobals-win32.c: New global variables for
multiple-monitor info: _gdk_num_monitors, _gdk_monitors, and
_gdk_offset_x and _gdk_offset_y.
* gdk/win32/gdkdisplay-win32.c (count_monitor, enum_monitor): New
functions, enumeration functions passed to EnumDisplayMonitors().
(gdk_display_open): If the EnumDisplayMonitors() and
GetMonitorInfo() API is present (on Win98, Win2000 and newer), use
if to find out monitor info.
Calculate the offset between Win32 coordinates (relative to the
primary monitor's origin (and thus negative on monitors to the
left of or above it), and GDK's (visible coordinates should be
non-negative).
* gdk/win32/gdkscreen-win32 (gdk_screen_get_n_monitors,
gdk_screen_get_monitor_geometry): Use information collected above.
(gdk_window_move, gdk_window_move_resize_window_get_geometry):
Subtract _gdk_offset_{x,y} from GDK root window coordinates.
(gdk_window_get_geometry, gdk_window_get_origin,
gdk_window_get_frame_extents): For top-level windows, add
_gdk_offset_{x,y} to GDK root window coordinates
Still need to handle multiple monitors in
gdk_window_fullscreen(). Probably should make the window
fullscreen on the monitor where the cursor is?
* gdk/win32/gdkevents-win32.c: Add _gdk_offset_{x,y} to all GDK
root window coordinates in GdkEvents.
[Win32] Fix geometry hint handling. Add support for resize
increment and base size, and aspect ratio geometry hints. The
"gridded geometry" test in testgtk now works beautifully.
* gdk/win32/gdkwindow-win32.c (gdk_window_set_geometry_hints):
Turns out this function shouldn't actually ever modify the
window's size, just store the hints. (Old code kept for a while
inside #if 0.)
(gdk_window_set_hints): Remove presumably broken code that handles
the position hints, this function is obsolete anyway.
* gdk/win32/gdkevents-win32.c: Drop the current_{x,y}_root
variables, not used.
(adjust_drag): New function, used to implement resize increment
hints.
(gdk_event_translate): Handle WM_SIZING, implement resize
increment and base size, and aspect ratio geometry hints here. The
WM_GETMINMAXINFO handler takes care of the minimum and maximum
size hints as before. Fix the WM_GETMINMAXINFO handler to take
into account window decorations. No need to modify the
ptMaxPosition and ptMaxSize fields in the MINMAXINFO struct,
the defaults are fine.
* gdk/win32/gdkprivate-win32.h
* gdk/win32/gdkwindow-win32.c (_gdk_win32_adjust_client_rect,
_gdk_win32_get_adjusted_client_rect): New helper functions.
2003-08-07 Matthias Clasen <maclas@gmx.de>
* gdk/x11/gdkevents-x11.c (gdk_check_wm_state_changed): Reset the toplevel->have_* flags before
iterating over the _NET_WM_STATE atoms. (#119217)
2003-08-07 Hans Breuer <hans@breuer.org>
* gdk/win32/gdkevents-win32.c : finally use TrackMouseEvent
(only if the new window not known to Gdk) to get proper
leave notification, and get rid of the wrong placed
tooltips, bug #102283
(gdk_event_translate) : small code reordering to not get
GDK_MOTION_NOTIFY for still mouse and get back tooltips on
menus, bug #117367
2003-08-04 Hans Breuer <hans@breuer.org>
* gdk/win32/gdkwindow-win32.c(gdk_window_set_modal_hint) :
set HWND_NOTOPMOST for modal==FALSE, fixes bug #118435
* gdk/win32/gdkwindow-win32.c : implement gdk_window_[un]fullscreen
(gdk_window_{move,resize,move_rezize})) : noop if FULLSCREEN
* gdk/win32/gdkwindow-win32.c (gdk_window_set_icon) :
against all expectations it is basically ony a call with WM_SETICON
needed (the visual effect is there, though colors or mask are
still somewhat wrong).
* gdk/win32/gdkwindow-win32.h : place to store the HICON
* gdk/win32/gdkdrawabke.win32.c : readded the conditional end
pixel drawing as it was fixed at 1999-08-23
gtk/gtk.def : updated
Mon Aug 4 14:53:46 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkimage-x11.c (_gdk_windowing_image_init):
Fix problem where use_xshm was getting set wrong if
MIT_SHM wasn't found. (George Lebel)
* gdk/win32/gdkevents-win32.c (apply_filters): Fix braino:
actually use the passed-in filter list. The function also needs a
GdkWindow parameter, as filter functions expect
GdkEvent::any.window to be valid. (#119034, Hans Breuer)
2003-08-03 Matthias Clasen <maclas@gmx.de>
* gdk/x11/gdkscreen-x11.c (gdk_screen_get_n_monitors):
(gdk_screen_get_monitor_geometry): Improve wording of the docs, avoid the
term "virtual screen". (#119030, Tor Lillquist)
Sat Aug 2 12:53:16 2003 Owen Taylor <otaylor@redhat.com>
* configure.in: Requires glib-2.3.0, pango-1.2.0.
* configure.in: Require Xft version 2,
remove code for handling older versions of pango and Xft.
Many miscellaneous improvements to X checks
* acinclude.m4: Add GTK_ADD_LIB() macro for adding
a library to a variable, avoiding dups.
* gdk/x11/gdkfont-x11.c (gdk_font_from_description_for_display):
Always load "fixed"
* gdk/x11/gdk*-x11.[ch]: Remove support for Xft1 and
for pangox.
2003-08-02 Tor Lillqvist <tml@iki.fi>
Fix gdk/win32 window geometry handling again. The window position
in a GDK_CONFIGURE event should be that of the client area, not of
the window decorations. (I was confused by the term "window
border" in X11. It does *not* mean the window manager
decorations. There are no X11-style window borders in Win32.)
Also, this time do take the geometry hints into account
appropriately when moving windows. Now testgtk's gravity test's
move buttons work OK. There are stil problems with taking gravity
into account when showing a hidden window.
* gdk/win32/gdkwindow-win32.h: Keep a whole GdkGeometry as hints
instead of separate fields.
* gdk/win32/gdkevents-win32.c (handle_configure_event): Don't
adjust for decorations.
* gdk/win32/gdkwindow-win32.c (get_outer_rect,
adjust_for_gravity_hints): New functions.
(gdk_window_move, gdk_window_resize, gdk_window_move_resize): Use
above functions, take geometry hints into account.
(gdk_window_set_geometry_hints): Size hints specicy client area,
not including decorations.
Fri Aug 1 16:30:13 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdrawable-x11.c (_gdk_x11_have_render):
If the check for missing depths, try actually creating pixmaps
of the depths to deal with Xinerama not reporting
all the depths it should. (#115822)
Fri Aug 1 15:06:25 2003 Owen Taylor <otaylor@redhat.com>
* gdk/gdk.c (gdk_parse_args): g_set_prgname("<unknown>")
if argc is 0 as well, instead of leaving it unset.
(#116023, Michael Meeks)
2003-07-29 Tor Lillqvist <tml@iki.fi>
Fix for #108007, #112402, #117042: There was confusion in
gdk/win32 at various places whether a window position refers to
the decoration position or the client area position. Also whether
window size includes decorations or not.
The correct interpretation apparently is that in GDK (like in
X11), a top-level window position means the decoration's position,
but size means the window's inner size (client area size). In the
Win32 API, the window size usually includes the decorations,
though.
* gdk/win32/gdkevents-win32.c (decode_key_lparam): Move inside
#ifdef G_ENABLE_DEBUG.
(handle_configure_event): New function, generates GDK_CONFIGURE
events from WM_SIZE and WM_MOVE messages. Even if no event is
generated because of the event mask, still set the private
position and size fields. Calculate position and size correctly.
(gdk_event_translate): Call handle_configure_event().
* gdk/win32/gdkgeometry-win32.c: Cosmetics.
* gdk/win32/gdkwindow-win32.c: Use GDI_CALL() and API_CALL()
macros. Cosmetic debugging output changes.
(SafeAdjustWindowRectEx): Remove. If an application wants to
locate a window outside of the screen, it's not GDK's business to
prevent it. And anyway, with multiple monitors, negative
coordinates are perfectly normal.
(gdk_window_new): Adjust the window size for decorations after
_gdk_window_init_position() has done its job. (But the big window
code currently is presumably broken on Win32 anyway.)
(gdk_window_move): The position passed in is supposed to be that
of the window border, so don't need to adjust for decorations.
(gdk_window_resize, gdk_window_move_resize): Simplify somewhat.
2003-07-27 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c (build_key_event_state): On Win9x,
the keyboard state returned by GetKeyboardState() doesn't
distinguish between the left and right Control and Alt keys. Thus
we cannot detect AltGr (which is supposed to be left Control +
right Alt) the same way as on NT-based systems, but have to accept
either Control + either Alt as AltGr.
2003-07-25 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkprivate-win32.h
* gdk/win32/gdkglobals-win32.c: New flags _gdk_input_locale_is_ime
and _gdk_keyboard_has_altgr.
* gdk/win32/gdkevents-win32.c: Lots of changes. Most important
ones detailled here.
Code that has been ifdeffed out for a long time removed. Remove
some really old doc comments that were left behind for some public
functions, the official ones are in the X11 backend anyway. Change
GDK_WINDOW_OBJECT() calls to GdkWindowObject casts. Reformat
multi-line boolean expressions to have the operators at ends of
lines.
As mouse capture with SetCapture() indeed seems to work OK, no
need to have the correspoinding macro USE_SETCAPTURE and ifdefs.
Ifdef out the gdk-ping-msg stuff. I don't remember why it was
needed at some time, and things seem to work fine now without
(knock on wood).
Ifdef out the search for some Latin locale keyboard layout being
loaded. Not used currently, but might be needed after all, if we
decide that we want to be able to generate ASCII control character
events with a non-Latin keyboard.
(assign_object): New helper function, handles the g_object_ref()
and unref() calls when assigning GObject pointers.
(generate_crossing_events): Also generate the GDK_NOTIFY_INTERIOR
enter event when the pointer has moved to an ancestor window. Was
left out by mistake.
(gdk_window_is_ancestor): Renamed from gdk_window_is_child().
(gdk_pointer_grab, gdk_pointer_ungrab): Implement the confine_to
functionality, using ClipCursor().
(find_window_for_mouse_event): Splice part of code into new
function find_real_window_for_grabbed_mouse_event().
(fixup_event, append_event, apply_filters): New functions, code
refactored out from elsewhere.
(synthesize_enter_or_leave_event, synthesize_leave_event,
synthesize_enter_event,
synthesize_leave_events,synthesize_enter_events): Also take a
GdkCrossingMode parameter, in preparation to generating
GDK_CROSSING_GRAB and GDK_CROSSING_UNGRAB events.
(fixup_event, append_event, fill_key_event_string): New functions,
code refactoring.
(vk_from_char, build_keypress_event, build_keyrelease_event):
Removed as part of dropping WM_CHAR handling.
(build_key_event_state,gdk_event_translate): Call
GetKeyboardState(), once, for each keyboard message, instead of
several calls to GetKeyState() here and there.
(gdk_event_translate): Fix bugs #104516, #104662, #115902. While
at it, do some major refactoring, and some fixes for potential
problems noticed while going through the code.
Don't handle WM_CHAR at all. Only handle WM_KEYDOWN and
WM_KEYUP. Don't need the state variables related to whether to
wait for WM_CHAR or not, and whether the current key is
AltGr. Remove lots of complexity. Thus don't need the
use_ime_composition flag.
Not handling WM_CHAR means dead key handling will have to be taken
care of by GTK, but that seems to work fine, so no worry.
Another side-effect is that Alt+keypad digits don't work any
longer, but it's better to learn to use GTK's ISO14755 support is
anyway.
Be more careful in checking whether AltGr is involved. Only
attempt to handle it if the keyboard actually has it. And
explicitly check for *left* Control plus *right* Alt being
pressed. Still, allow (left) Alt and/or (right) Control with AltGr
chars.
Handle keys using similar code as in the X11 backend. As we have
built a keymap in gdkkeys-win32.c anyway, use it by calling
gdk_keymap_translate_keyboard_state() to look up the keysym from
the virtual key code and keyboard state. Build the key event
string in exactly the same way as the X11 backend.
If an IME is being used, don't generate GDK events for keys
between receiving WM_IME_STARTCOMPOSITION and
WM_IME_ENDCOMPOSITION, as those keys are for the IME.
For WM_IME_COMPOSITION, handle all the Unicode chars returned from
the IME, not just the first one.
gdk_event_translate() is still quite complex, could split the
message handler cases out into separate functions.
On mouse events, when the mouse is grabbed, use
find_real_window_for_grabbed_mouse_event() in order to be able to
generate correct crossing events.
No longer take a pre-allocated GdkEvent as parameter. Instead,
allocate events as needed and append them to the queue. (This is
different from how gdk_event_translate() in the X11 backend
works.) This change made the code much clearer, especially in the
cases where we have to generate several GDK events for one Windows
message. Return FALSE if DefWindowProc() should be called, TRUE
if not. If DefWindowProc() should not be called, also return the
value to be returned from the window procedure.
(Previously, the interaction with gdk_event_translate()'s caller
was much more complex, when we had to indicate whether the
already-queued event should be left in the queue or removed, and
in addition also had to indicate whether to call DefWindowProc()
or not, and what value to return from the window procedure if
not.)
Don't use a separate "private" variable required to be pointing to
the GdkWindowObject of the "window" variable at all times. Just
use casts, even if looks a bit uglier.
Notice destroyed windows as early as possible, and break out of
the messsage switch.
Use _gdk_pointer_root as current_window when the pointer is
outside GDK's top-level windows.
On WM_INPUTLANGCHANGE, set _gdk_input_locale_is_ime as
appropriate, based on ImmIsIME().
(gdk_event_translate, gdk_event_send_client_message_for_display,
gdk_screen_broadcast_client_message): Implement client messages.
Use a registered Windows message to pass GDK client messages. Note
that the amount of user data is restricted to four bytes, as it is
carried in the LPARAM. (The WPARAM is used for the message type
"atom".)
(real_window_procedure): Adapt for new gdk_event_translate()
interface.
* gdk/win32/gdkmain-win32.c (_gdk_windowing_init): Set
_gdk_input_locale_is_ime initially.
* gdk/win32/gdkwindow-win32.c: Use g_object_ref()/unref() instead
of g_colormap_ref()/unref().
(gdk_window_new): Made code a bit more like the X11 one, pretend
to handle screens (although we just have one for now).
* gdk/x11/gdkevents-x11.c
(gdk_event_send_client_message_for_display,
gdk_screen_broadcast_client_message): Document the user data
limitation on Win32.
* gdk/win32/gdkevents-win32.c (print_event): More complete enter
and leave notify detail output.
* gdk/win32/gdkkeys-win32.c (update_keymap): Make dead keys
visible to GDK and GTK. Store the corresponding GDK_dead_* keysym
for those, so that the GtkIMContextCimple compose tables will
work. Deduce if the keyboard layout has the AltGr key, and set the
above flag accordingly.
2003-07-20 Hans Breuer <hans@breuer.org>
* makefile.msc : new file to build it all
* Makefile.am : ... added to EXTRA_DIST
* gdk/makefile.msc gtk/stock-icons/makefile.msc
gtk/makefile.msc.in tests/makefile.msc : updated
* gdk/gdk.def gtk/gtk.def : export a bunch of new
functions
* gtk/gtkbbox.c : use g_return_val_if_fail() if
there is a value to return
* gtk/gtkfontsel.c gtk/gtkiconfactory.c :
static correctness
* gtk/gtkicontheme.c : ported to use GTimeVal instead of
struct timeval and GTK_DATA_PREFIX instead of GTK_DATADIR
* gtk/gtkicontheme.c : use HAVE_LC_MESSAGES
* gtk/gtkimmulticontext.c : use gtkprivate.h
* gtk/stock-icons/stock_color_picker_25.png
gtk/stock-icons/stock_jump_to_rtl_16.png
gtk/stock-icons/stock_jump_to_rtl_24.png
gtk/stock-icons/stock_redo_rtl_16.png
gtk/stock-icons/stock_redo_rtl_24.png
gtk/stock-icons/stock_undelete_rtl_16.png
gtk/stock-icons/stock_undelete_rtl_24.png
gtk/stock-icons/stock_undo_rtl_16.png :
gtk/stock-icons/stock_undo_rtl_24.png :
readded as binary
Tue Jul 8 20:11:04 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkevents-x11.c (get_real_window)
gdk/x11/gdkinput-x11.c (_gdk_input_common_init)
gdk/x11/gdkimage-x11.c (_gdk_windowing_image_init)
gdk/x11/gdkprivate-x11.h (_gdk_windowing_image_init)
Don't assume that all events start with XEventAny - Xkb events
don't! (#105745). So, only do that for core events, and for
non-core events, add a system for registering event types
that start with XEventAny.
* gdk/x11/gdkevents-x11.c (gdk_event_translate):
Check to see if the result of gdk_window_lookup_for_display()
is actually a window.
Wed Jul 9 19:48:26 2003 Kristian Rietveld <kris@gtk.org>
Merge from stable.
* gdk/gdkdisplay.c (gdk_display_get_window_at_pointer): don't
directly return from ->window_at_pointer, but set win_x and win_y
first if needed. (Fixes#110166, reported by Arno Charlet).
* gdk/gdkwindow.c (gdk_window_get_composite_drawable): change
function to accept GdkDrawables and not just GdkWindows. This was
already done in some other functions a while back, but this patch
got lost for some reason.
Tue Jul 8 17:36:21 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkasync.[ch] gdk/gdkdnd-x11.c:
Change _gdk_x11_send_xevent_async() to
_gdk_x11_send_client_message_async() avoid using Xlib
internals that are different on Solaris.
(#116917, Morten Welinder)
Sat Jul 5 09:55:38 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdnd-x11.c (precache_target_list): Lookup
all the atoms in the target list at once.
* gdk/x11/gdkdnd-x11.c (gdk_drag_get_protocol_for_display):
Local drag short-circuit.
* gdk/x11/gdkdnd-x11.c (gdk_drag_motion): For the
local case, poke the actions in directly instead of
going through xdnd_set_actions.
* gdk/x11/gdkdnd-x11.c (xdnd_read_actions): Short-circuit
the local case.
* gdk/x11/gdkdnd-x11.c (xdnd_manage_source_filter): Don't
XSelectInput()/add a filter if the drag is local.
* gdk/x11/gdkdnd-x11.c (gdk_drag_get_selection):
Simplify handling of "XdndSelection".
* gdk/x11/gdkevents-x11.c (gdk_event_send_client_message_to_all_recurse): Somehow,
a WM_STATE => _NET_WM_STATE change hand been made here.
* gdk/x11/gdkproperty-x11.c (_gdk_x11_precache_atoms): Free
xatom_names, not atom_names.
* tests/testdnd.c (target_drag_motion): Make the trashcan
open again. (Got lost in deprecation cleanup.)
Sat Jul 5 00:49:32 2003 Owen Taylor <otaylor@redhat.com>
* gtk/gtkdnd.c (gtk_drag_begin): Move the grabs up
earlier, so we don't get a big pile of uncompressed
motion events before the grab takes effect.
* gtk/gtkdnd.c (gtk_drag_update_icon): Call
gdk_window_process_all_updates() so that the area
we are dragging over gets a change to redraw.
* gtk/gtkdnd.c (_gtk_drag_dest_handle_event): Use
gdk_window_get_position(), not gdk_window_get_origin().
* gdk/x11/gdkwindow-x11.[ch] (gdk_window_have_shape_ext)
gdkdisplay-x11.h: use XShapeQueryExtension to
avoid duplicate XQueryExtension when we actually
use it. Remove unnecessary caching in GdkDisplayX11.
Fri Jul 4 23:49:22 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkasync.[ch] (_gdk_x11_get_window_child_info):
Function to get a range of information about all the
children of a window in a single pass.
* gdk/x11/gdkdnd-x11.c (get_client_window_at_coords_recurse)
gdk/x11/gdkdnd-x11.c (gdk_window_cache_new): Use
_gdk_x11_get_window_child_info() to greatly reduce
the number of roundtrips.
Fri Jul 4 22:57:18 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkasync.[ch] (_gdk_send_xevent_async): Add
a function to XSendEvent() and call a calback on
failure/success.
* gdk/x11/gdkdnd-x11.c (xdnd_send_xevent): Short-circuit
messages to the same process, use _gdk_send_xevent_async().
Fri Jul 4 22:26:27 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkwindow-x11.[ch] gdkevents-x11.c: Split
toplevel-specific pieces of GdkWindowImplX11 into
a separate GdkToplevelX11 structure.
Fri Jul 4 15:57:52 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkevents-x11.c (gdk_wm_protocols_filter):
Use asynchronously _gdk_x11_set_input_focus_safe
to avoid having to trap errors and XSync().
* gdk/x11/gdkwindow-x11.c (gdk_window_focus): Use
_gdk_x11_set_input_focus_safe() here as well.
* gdk/x11/gdkevents-x11.c (gdk_check_wm_state_changed):
Rework handling of property notifies on _NET_WM_STATE
so that we ignore _NET_WM_DESKTOP notifies unless we
really care.
* gdk/x11/gdkimage-x11.c (gdk_image_check_xshm): Use
XShmQueryExtension() rather than XQueryExtension() to
avoid extra rountrip.
* gdk/x11/gdkwindow-x11.c (_gdk_windowing_window_init):
Remove unused call to XGetWindowAttributes()
* gdk/x11/gdkdisplay-x11.c (gdk_display_open): Remove
unused call to XGetKeyboardControl().
* gdk/x11/gdkdisplay-x11.c gdk/gdk.def (gdk_display_flush):
Add (#99571)
* gdk/win32/gdkevents-win32.c gdk/linux-fb/gdkevents-fb.c
No-op implementations of gdk_display_flush().
* gdk/gdkwindow.c (gdk_window_process_all_updates): Use
gdk_display_flush() rather than gdk_flush() to avoid
XSync().
* gdk/x11/gdkwindow-x11.c (update_wm_hints)
gdk/x11/gdkwindow-x11.h: Centralize all handling of WM_HINTS here
so that we don't have to get the property back from the server.
* gdk/x11/gdkwindow-x11.c (show_window_internal): Store
the serial of when we map a toplevel to allow optimizing
out notifies on _NET_WM_STATE/_NET_WM_DESKTOP.
* gdk/x11/gdkevents-x11.c (gdk_event_translate): Don't
XTranslateCoordinates() for override-redirect windows.
Fri Jul 4 15:59:27 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkwindow-x11.c (gdk_window_set_group): Remove comment
about setting window group after the window is mapped from docs
- nothing the ICCCM forbids that.
* gdk/x11/gdkcursor-x11.c (gdk_display_get_maximal_cursor_size):
Fix g_return_val_if_fail() in void return function.
* configure.in: Fix misplaced comma that was resulting
in XShm always being disabled.
2003-07-04 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdrawable-win32.c (gdk_win32_draw_polygon,
gdk_win32_draw_segments, gdk_win32_draw_lines): We can calculate
the width and height of the bounding rectangle only after the
minumum x and y have been found, and need a separate loop for
it. Thanks to Bruce Hochstetler for providing a sample program
exhibiting the bug.
Wed Jul 2 18:00:56 2003 Owen Taylor <otaylor@redhat.com>
* gtk/gtkicontheme.[ch]: Implement a loader for
named themed icon based on from gnome-desktop library
by Alex Larsson.
* gtk/gtkiconthemeparser.[ch]: .ini file parsing code
from gnome-desktop.
* gtk/gtkiconfactory.[ch]: Add
gtk_icon_source_set/get_icon_name() to allow stock icons
to be based off of named theme icons.
* gtk/gtkiconfactory.c: Rework sources so that the source
is *either* a pixbuf, or a filename, or an icon name,
instead of the pixbuf/filename mix it was before. Put a
workaround for get_pixbuf() so that it can return the
filename pixbuf, e.g, for render_icon().
* gtk/gtkiconfactory.c: Make the default setup use
themed icons, and add builtin icons to the default
icon theme for all the standard pixbufs, so we
don't rely on actually having an icon theme on disk.
* gtk/gtkrc.c: Add support for @"icon-name" to specify
a themed icon for a stock icon source.
* tests/Makefile.am test/testicontheme.c: Add a test
program from gnome-desktop.
* gdk/x11/gdkevents-x11.c gtk/gtksettings.c: Add
Net/IconThemeName / gtk-icon-theme-name setting.
* gtk/gtkiconfactory.c (ensure_cache_up_to_date): Actually
update the icon cache serial so we don't continually
* gtk/gtkwidget.c: Fix a couple of references in doc comments
to ::direction_set that should have been to ::direction-changed
2003-07-01 Matthias Clasen <maclas@gmx.de>
* configure.in: Check for Xcursor.
* gdk/x11/gdkcursor-x11.c:
* gdk/gdkdisplay.h:
* gdk/gdkcursor.h: RGBA cursor support based on Xcursor. New functions:
gdk_cursor_new_from_pixbuf(), gdk_display_supports_cursor_alpha(),
gdk_display_supports_cursor_color(), gdk_display_get_default_cursor_size() and
gdk_display_get_maximal_cursor_size(). (#69436)
2003-06-30 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdnd-win32.c (find_window_enum_proc): New function,
callback proc for EnumWindows().
(gdk_drag_find_window_for_screen): Actually take the drag_window
argument into account: Instead of using WindowFromPoint(), use
EnumWindows(), to be able to skip the drag_window. (#116320, Tony
M Brown, Herman Bloggs)
* gdk/win32/*.c: Replace gdk_drawable_ref()/unref() and
gdk_window_ref()/unref() calls with g_object_ref()/unref().
Consistently use %p format in debugging output of pointers and
HANDLEs.
2003-06-28 Tor Lillqvist <tml@iki.fi>
Fix for #111028, thanks to J. Ali Harlow, who writes:
I found that the GdkPixmap->GdkImage reference really isn't
important. It's only really there to have somewhere convenient to
store the location of the pixel data in the pixmap and as an easy
way of accessing the dimensions of that data. I have therefore put
together a fix which removes this reference entirely which seems
to solve the problem.
* gdk/win32/gdkpixmap-win32.h (struct _GdkPixmapImplWin32):
Instead of a pointer to a GdkImage, keep a pointer to the pixels
directly.
* gdk/win32/gdkimage-win32.c (_gdk_win32_setup_pixmap_image): Remove.
(_gdk_win32_new_image): New function, replacing the above. Creates
a GdkImage without any associated GdkPixmap.
(gdk_image_new_bitmap, _gdk_image_new_for_depth): Use it instead.
* gdk/win32/gdkprivate-win32.h: Remove from here, too.
* gdk/win32/gdkcursor-win32.c (gdk_cursor_new_from_pixmap)
* gdk/win32/gdkdrawable-win32.c (blit_from_pixmap)
* gdk/win32/gdkgc-win32.c (_gdk_win32_bitmap_to_hrgn)
* gdk/win32/gdkmain-win32.c (_gdk_win32_drawable_description):
* gdk/win32/gdkpixmap-win32.c (gdk_pixmap_impl_win32_finalize,
gdk_pixmap_new, gdk_bitmap_create_from_data, gdk_pixmap_foreign_new)
Corresponding changes.
2003-06-21 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c (gdk_event_translate): When
emulating X11's automatic grab on button down, pass owner_events
as FALSE. According to the XLib spec, automatic grabs use True for
owner_events when OwnerGrabButtonMask is selected, and I don't see
the X11 backend doing that. (#110271)
(find_window_for_pointer_event): Improve debugging output.
Sun Jun 8 22:03:09 2003 Owen Taylor <otaylor@redhat.com>
* tests/testdnd.c: Use application/x-rootwindow-drop for
root window drops. (#108670, Alex Larsson)
* gdk/x11/gdkdnd-x11.c (gdk_drag_motion) gtk/gtkdnd.c (gtk_drag_drop):
Accept either application/x-rootwin-drop (what GTK+ has always used)
or application/x-rootwindow-drop (what the XDND standard specifies).
Fri Jun 6 11:07:33 2003 Owen Taylor <otaylor@redhat.com>
* gtk/gtkwidget.c (event_window_still_viewable):
Before delivering an event to a widget, check that
(if relevant), the event's window is still viewable.
(#105642, Dennis Björklund)
* gdk/gdkwindow.c (_gdk_window_destroy_hierarchy):
NULL out private->parent, since after destruction
it might not be valid any more.
* gdk/gdkwindow.c (gdk_window_is_viewable): Fix some
accesses before g_return_val_if_fail(). Treat
DESTROYED windows as unmapped.
2003-06-06 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkgc-win32.c (_gdk_win32_gc_new): Set
graphics_exposures and subwindow_mode, too, even if they aren't
currently used.
Wed Jun 4 19:42:17 2003 Owen Taylor <otaylor@redhat.com>
* gdk/gdkpango.c (gdk_draw_layout_line_with_colors):
When drawing underlines, join up adjacent runs
where possible, so we don't get changes in
shaper/font/etc. breaking underlines. (#103662,
Kang Jeong-Hee)
Wed Jun 4 19:24:28 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkgc-x11.c gdk/x11/gdkprivate-x11.h: Keep
track of when we have a clip mask set for the GC,
and when we unset it, or switch to a clip region,
immediately call XSetClipMask (..., None) to avoid
Xlib caching stale data. (#111806)
* gtk/gtktextdisplay.c: Don't set a clip mask
when drawing alpha pixmaps; it isn't necessary any more.
(#111806)
Tue Jun 3 18:32:30 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkevents-x11.c (gdk_wm_protocols_filter): Don't
respond to pings on a root window; prevents infinite
loops if we are selecting for SubstructureNotify on the
root window. (#111945, Sergey V. Udaltsov)
Tue Jun 3 17:39:16 2003 Owen Taylor <otaylor@redhat.com>
#71597, reported by Morten Welinder
* gdk/gdkpixbuf-drawable.c (rgb888amsb): Fix and simplify
(Patch from Christian Petig)
* gdk/gdkpixbuf-drawable.c (rgb{555,565}{a,}{msb,lsb}):
Major rewrite of 555 and 565 conversion routines:
- Move all the bit shifting into a small block of macros,
eliminating much duplication of complicated arithmetic.
- Get rid of 2-pixels at a time code, which was buggy,
hard to maintain, caused unaligned accesses, and
probably didn't actually perform any better.
- Simplify cases where different data types were
used for the little and big endian cases, use
GUINT16_SWAP_LE_BE() where appropriate.
Tue Jun 3 15:05:47 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdrawable-x11.c (convert_to_format): Remove
bad optimization for src_rowstride == dest_stride.
(Didn't consider the case where we were copying only
partial widths of a wider source buffer)
(#113034, reported by Hans Petter Jansson)
2003-05-27 Tor Lillqvist <tml@iki.fi>
* gtk-zip.sh.in: Only include the Default and Emacs themes' gtkrc
files. Include share/gtk-doc/{gdk-pixbuf,gdk,gtk}.
* gdk/Makefile.am: libgdk-win32-2.0.la depends on
win32/libgdk-win32.la.
* gtk/gtk.def: Add gtk_tree_view_column_cell_get_position.
Tue May 20 15:09:29 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdnd-x11.c gdk/x11/gdkwindow-x11.c: Comprehensively
zero all unused parts of client messages (Mainly Motif DND
and EWMH). Also, fix a couple places where we were putting
contents into Motif DND messages that didn't belong.
(#113361, EWMH problem reported by Lubos Lunak)
* gtk/gtkselection.c (_gtk_selection_request): Handle the
case where the property type for the multiple atoms is
ATOM_PAIR by doing the conversions ourselve. This is
needed for Xt inter-operation.
Wed May 14 17:45:32 2003 Owen Taylor <otaylor@redhat.com>
* gdk/gdkkeysyms.h gdk/gdkkeynames.c gdk/gdkkeyuni.c:
Patch from Vasilis Vasaitis to fix Greek_IOTAdi[a]eresis
inconsistency and leave the old name as an aliass (#104873)
Tue May 6 16:50:52 2003 Owen Taylor <otaylor@redhat.com>
Patch from James Henstridge to update to automake-1.7
(#109542)
* autogen.sh: update to call newer tools.
* configure.in: various updates, to use M4 macros to put
variables that change each release at the top.
Use AC_HELP_STRING to format help strings.
Use AC_CONFIG_COMMANDS to generate gdkconfig.h.
* Makefile.am: require Automake 1.7. Remove gdk-2.0.pc and
gtk+-2.0.pc on uninstall. Pass --enable-gtk-doc to configure
during distcheck.
* docs/reference/*/Makefile.am: simplify to use the gtk-doc.make
makefile fragment.
* */Makefile.am: don't use STRIP_BEGIN/STRIP_END.
Use BUILT_SOURCES where appropriate.
Build generated files in builddir rather than srcdir.
Fix uninstall and distclean targets to satisfy distcheck.
===
* gdk/*/Makefile.am: Remove the hacks to get gdkenumtypes.h
built first, since we are now using BUILT_SOURCES.
* gdk/Makefile.am: Remove an outdated comment about gdk_headers.
* demos/gtk-demo/Makefile.am: Fix srcdir != buildd problem
with geninclude.pl.
* configure.in: Update versions to 2.3.0.
2003-05-06 Tor Lillqvist <tml@iki.fi>
Fix for #110165 (thanks to Arnaud Charlet):
* gdk/win32/gdkevents-win32.c (build_keypress_event,
build_keyrelease_event): For unshifted control char, use
lowercase ASCII keyval.
(gdk_event_translate): Similarily, when handling WM_SYSKEYDOWN and
UP (i.e. Alt-something), if it's an unshifted ASCII letter, use
lowercase keyval. Use build_key_event_state() here, too, instead
of minor code duplication.
Mon Apr 21 18:13:46 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdnd-x11.c (gdk_window_cache_filter): Patch
from Alex Larsson fixing problem with restacking during
DND. (#108671)
Mon Apr 21 15:07:29 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdisplay-x11.c (gdk_display_keyboard/pointer_ungrab):
Calling XFlush() after ungrabbing, to avoid problems if
the app subsequently blocks on a long-running operation.
(#106520, reported by Rajkumar Siva)
Fri Apr 18 17:42:45 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkwindow-x11.c (set_initial_hints): Add
SKIP_TASKBAR/SKIP_PAGER to set of properties we
might set on map. (#110019, problem reported
by Loban Rahman)
Fri Apr 18 17:14:33 2003 Owen Taylor <otaylor@redhat.com>
* gdk/gdkevents.c (gdk_event_copy): Fix for the
case when gdk_event_copy() is called on a non-allocated
event. (#109716, reported by Rich Burridge)
Fri Apr 18 16:31:49 2003 Owen Taylor <otaylor@redhat.com>
* configure.in gdk/x11/gdkdisplay-x11.c modules/input/Makefile.am:
Check for XAddConnectionWatch()/X11R6, and if found, disable XIM and
use of XAddConnectionWatch. Remove --enable-xim config option.
Should fix building on X11R6 (#110523, Albert Chin
Fri Apr 18 15:56:46 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdisplay-x11.[ch] gdk/x11/gdkdrawable-x11.[ch]:
Patch from Morten Welinder to catch Sun servers with a
broken implementation of the RENDER extension. (#108309)
Fri Apr 18 15:30:38 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkwindow-x11.c (set_text_property): Use
gdk_free_compound_text(), not g_free(). (#107643,
Michael Zucchi)
Fri Apr 18 14:15:09 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdisplay-x11.c (gdk_display_open):
* gdk/x11/gdkevents-x11.c (gdk_event_translate): You
don't get a XkbMapNotify event if you get a XkbNewKeyboardNotify
event, so select for both. (#108406, Egmont Koblinger)
Mon Apr 7 19:43:15 2003 Owen Taylor <otaylor@redhat.com>
* gdk/x11/gdkdisplay-x11.c (gdk_x11_display_ungrab): Flush
after ungrabbing the server.
* gdk/x11/gdkimage-x11.c: Remove some no-longer-needed
calls to XFlush().
* gdk/x11/gdkwindow-x11.c (_gdk_windowing_window_at_pointer):
Use gdk_x11_display_grab/ungrab.
2003-03-14 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c: Use the signed GET_X_LPARAM() and
GET_Y_LPARAM() to extract x and y coordinates from an LPARAM or
DWORD, and not the unsigned HIWORD() and LOWORD(). Systems with
multiple monitors can have negative coordinates on some of the
monitors. (partial fix for #99496, Arnaud Charlet)
2003-03-13 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c (gdk_event_translate): On
WM_MOUSEMOVE, the test whether we have actually moved (and thus
shouldn't generate an GDK event) moved after the call to
propagate() and translate_mouse_coords(). Otherwise we were
testing wrong values. (#108115, Allin Cottrell)
(gdk_event_translate): On WM_?BUTTONUP, set current_{x,y} the same
way as in WM_?BUTTONDOWN and WM_MOUSEMOVE.
2003-03-13 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkevents-win32.c (gdk_pointer_grab): if a grab cursor
is used, must copy the HCURSOR with CopyCursor(), as it is OK to
destroy the GdkCursor after calling gdk_pointer_grab(). Set the
cursor right away with SetCursor(), as we won't get any
WM_SETCURSOR messages while the mouse is captured.
(gdk_display_pointer_ungrab): Correspondingly, destroy the copy
with DestroyCursor() when no longer used. (#108114, Allin Cottrell)
* gdk/win32/gdkevents-win32.c (gdk_event_translate): On
WM_SYSCHAR, return FALSE from the window procedure to prevent the
DefWindowProc from being called. Otherwise Windows would beep,
thinking you are tring to access a (nonexistent) menu when you
press Alt-something. Don't do this for Alt-Space,
though. (#107454, Martyn Russell)
* gdk/win32/gdkcursor-win32.c: Minor spacing and debug print fixes.
2003-03-09 Tor Lillqvist <tml@iki.fi>
* gdk/win32/gdkdrawable-win32.c (draw_segments): Don't try to
compensate for LineTo() not drawing the end pixel. It causes more
harm than benefits, see bug #81895.
