2003-06-04 Rob Adams <robadams@ucla.edu>
Precompute groups to guarantee that meta_group_list_windows always
returns the correct list of windows. See Bug #96973
* src/window.h: change cached_group variable to group
* src/window.c (meta_window_new): change cached_group to group and
call meta_window_compute_group
* src/groups.c (meta_window_get_group): simply return
window->group rather than computing it and returning
window->cached_group
(meta_window_compute_group): new function computes window->group.
Designed to be called once from meta_window_new
(remove_window_from_group): change cached_group to group
CVS2003-05-29 Rob Adams <robadams@ucla.edu>
Use a new property _METACITY_SENTINEL to eliminate a race
condition that causes focus to behave badly with sloppy/mouse
focus when lots of windows are mapped/unmapped, such as with a
workspace switch. The EnterNotify events on a display are ignored
until the PropertyNotify sent after all the window maps is
received. This is a fix for #110970.
* src/display.[ch]: New _METACITY_SENTINEL atom.
(event_callback): ignore EnterNotify if the sentinel isn't clear,
and decrement the sentinel counter when the PropertyNotify is
received.
(meta_display_increment_focus_sentinel): new function. Increments
the sentinel counter and updates the property on a root window on
this display.
(meta_display_decrement_focus_sentinel): Decrement the sentinel
counter.
(meta_display_focus_sentinel_clear): returns whether the sentinel
counter is zero.
* src/window.c (idle_calc_showing): after showing windows, call
meta_display_increment_focus_sentinel on each display for windows
to be shown.
* src/workspace.[ch] (meta_workspace_activate_with_focus): new
function activates a workspace and focuses a particular window
after the workspace is activated.
(meta_workspace_activate): now just a wrapper for
meta_workspace_activate_with_focus
* src/keybindings.c: use new meta_workspace_activate_with_focus
function to ensure that focus will follow the focused window
through the workspace switch.
: ----------------------------------------------------------------------
2003-05-18 Havoc Pennington <hp@pobox.com>
* src/window.c (unminimize_window_and_all_transient_parents):
revert broken change that assumed foreach_ancestor iterated
over the window itself. Andrew Sobala, Rob Adams,
#113232
`
2003-05-16 Havoc Pennington <hp@redhat.com>
* src/window.c (meta_window_new): fill in window->desc sooner
since we use it sooner now.
* src/display.c (meta_display_open): init
display->grab_update_alarm
* src/window.c (meta_window_new): initialize the always_sticky
field
(meta_window_new): initialize the update_icon_queued field
Patch from Julien Olivier bug #92335 for converting "show desktop
mode" to "all windows are minimized" when you open a new window,
instead of just mapping all the windows again.
* src/window.c (meta_window_activate): minimize all windows before
coming out of show desktop mode.
(meta_window_unminimize): don't toggle show desktop mode here
* src/screen.c (meta_screen_minimize_all_except): new function
2003-05-03 Havoc Pennington <hp@pobox.com>
* src/keybindings.c (handle_move_to_workspace): when moving
window to another workspace, don't switch to that workspace.
* src/window.c (menu_callback): when moving window to another
workspace, don't switch to that workspace.
2003-03-30 Rob Adams <robadams@ucla.edu>
* src/window.c (meta_window_move_resize_internal): When passing
frame geometry to meta_window_constrain, send null if no frame.
Possible fix for #109039.
Fri Mar 28 14:13:37 2003 Soeren Sandmann <sandmann@daimi.au.dk>
* src/window.c (update_resize): Only cap refresh rate when not
using SYNC. Remove bogus update-if-we-moved-more-than-a-delta.
* src/window.c (update_move): Don't cap refresh rate during
moves. Remove bogus update-if-we-moved-more-than-a-delta.
2003-03-11 Havoc Pennington <hp@pobox.com>
Should fix#108108, #106217, tracked down by Owen Taylor and
Frederic Crozat
* src/window.c (meta_window_foreach_transient): change
MetaWindowForeachFunc to return a boolean for whether to continue
(meta_window_foreach_ancestor): new function
(window_should_be_showing): use meta_window_foreach_ancestor
(unminimize_window_and_all_transient_parents): ditto
(update_sm_hints): ditto
(meta_window_is_ancestor_of_transient): ditto
* src/stack.c (get_maximum_layer_of_ancestor): use
meta_window_foreach_ancestor
2003-03-16 Rob Adams <robadams@ucla.edu>
* window.c (meta_window_show_menu): Free old window menu if it
already exists so we don't end up with more than one. Fix for
#108392.
2003-02-27 Havoc Pennington <hp@redhat.com>
Switch over to new constraints code, unquestionably introduces
some bugs, but should get us on the right path.
* src/window.c (meta_window_get_work_area_all_xineramas): create
this function again as it turned out to be legitimate for window
position constraint
(adjust_for_gravity): use the width/height from the configure
request to compute the requested move
(meta_window_move_resize_internal): use meta_window_constrain
(update_size_hints): clamp max size to MAXSHORT to avoid worrying
about overflow stuff
* src/constraints.c (meta_window_constrain): don't base placement
on uninitialized variables, general hacking
* src/Makefile.am (metacity_SOURCES): add constraints.c,
constraints.h
* src/constraints.c (meta_window_constrain): update the
cut-and-paste aspect ratio code to have latest bugfixes
2003-02-27 Havoc Pennington <hp@pobox.com>
* src/main.c (main): add more debug spew about conditional
build stuff
(main): panic to "Simple" theme
* src/window.c, src/window-props.c: move WM_NORMAL_HINTS and
WM_PROTOCOLS to new property system; don't queue move resize on
updating WM_PROTOCOLS; move WM_HINTS to new property system;
reload icon in an idle handler.
Changes made on plane from FOSDEM, syncing from laptop.
2003-02-28 Mark McLoughlin <mark@skynet.ie>
Give me back my keys.
* src/keybindings.c: (meta_window_grab_keys): don't
grab keys on DOCK windows.
* src/window.c: (recalc_window_type): re-grab the
keys.
2003-02-23 Havoc Pennington <hp@pobox.com>
Patch from Rob Adams addresses #95014 (placement issues),
makes first fit algorithm "center tile", adds most code
for per-xinerama workspaces (#86682) but disables it for now.
* src/workspace.c (meta_workspace_get_work_area_for_xinerama)
(meta_workspace_get_work_area_all_xineramas): new xinerama
functions, maintain workspace->work_areas with a different
work area for each xinerama. However for now all the work
areas are the same, because haven't quite figured out how
_NET_WM_STRUT is supposed to work
* src/window.c: adapt to new meta_window_* xinerama APIs
(meta_window_get_work_area_current_xinerama): new xinerama
API
(meta_window_get_work_area_for_xinerama): new xinerama API
(constrain_position): be a bit more clever about which xinerama's
work area we choose to use.
* src/stack.c: adapt to new Xinerama API
* src/screen.c (reload_xinerama_infos): invalidate all work areas
(meta_screen_get_xinerama_for_rect): new function
(meta_screen_window_intersects_xinerama): new function
* src/place.c (find_first_fit): change to use
"center tiling" (center a screen full of tiled windows,
rather than aligning them top left). Adapt to new
xinerama functions.
2003-02-22 Rob Adams <robadams@ucla.edu>
* window.c (set_wm_state): modify comment to explain why the icon
window element is set to None. Fix for #97357 thanks to Gregory
Merchan.
2003-01-25 Havoc Pennington <hp@pobox.com>
* src/stack.c (window_is_fullscreen_size): When checking if a
window is fullscreen size, only require it to be at the origin
of the work area, not at the origin of the screen/xinerama.
Still require it to be full screen in width x height.
May fix xine in the case where the user has a top panel.
* src/window.c (constrain_position): restore the ability for
undecorated windows to position themselves overlapping the top
panel, but don't let decorated windows do so. Oh the hacks...
2003-01-22 Havoc Pennington <hp@pobox.com>
* src/window.c (update_struts): be robust against the panel's
lame "set a negative number for struts" thing, even though
we'll also fix the panel.
2003-01-21 Havoc Pennington <hp@pobox.com>
Fix for the "mangles focus window when switching workspaces
and using mouse focus" bug
* src/stack.c (meta_stack_get_default_focus_window_at_point): new
function
* src/screen.c (meta_screen_focus_mouse_window): new function
(meta_screen_focus_default_window): new function
* src/workspace.c (meta_workspace_activate): use the
new meta_screen_focus_default_window()
2003-01-17 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_handle_mouse_grab_op_event): fix event
compression code to use GDK algorithm suggested by Owen, should be
more efficient.
2003-01-14 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_maximize, meta_window_unmaximize)
(meta_window_make_fullscreen, meta_window_unmake_fullscreen):
recalc_window_features() after making these changes, should fix
#103317
2003-01-11 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_handle_mouse_grab_op_event): implement
compression of motion events (drop all but the most recently
received), guessing at fixes for #103009
2003-01-09 Havoc Pennington <hp@redhat.com>
* src/window.c (recalc_window_features): argh, we were making all
dialogs skip taskbar; when did that get added. Fix to match
libwnck, only skip taskbar when the dialog is transient for some
other app window.
2003-01-06 Arvind Samptur <arvind.samptur@wipro.com>
* src/window.c: (constrain_position) don't apply
offscreen height difference. This would get the
window under the panel on a resize or a move.
Fixes#102418
2003-01-05 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_show): focus new windows even in
mouse focus mode, #89981, patch from readams@hmc.edu
2003-01-05 Havoc Pennington <hp@pobox.com>
* src/frames.c (meta_frames_apply_shapes): handle
the client having a shape mask, fixes#101806
* src/core.c (meta_core_get_client_xwindow): new function
* src/frame.c, src/frame.h: keep a flag for whether we need to
update the frame shape
* src/window.c (meta_window_new): select for ShapeNotify
* src/display.h, src/display.c: actually query the shape
extension, instead of just using it all over the place.
* src/prefs.c (update_application_based): don't let people turn on
application_based, as it just causes funky bugs. We can reenable
the pref when/if it ever does something useful.
2003-01-03 Havoc Pennington <hp@redhat.com>
* src/display.c: include the Xrandr header file
* src/window.c (meta_window_fill_horizontal)
(meta_window_fill_vertical): maximize to work area, not entire
screen. doh.
2002-12-18 Havoc Pennington <hp@redhat.com>
* src/window.c (meta_window_new): select ColormapChangeMask
on toplevel windows, maybe a partial fix for #101478
2002-12-17 Havoc Pennington <hp@pobox.com>
* src/display.c (meta_display_begin_grab_op): don't use "(null)"
for null pointers, use "none", so I can distinguish
glibc-generated (null) which is a bug.
(key_event_description): ditto
(meta_display_begin_grab_op): ditto
* src/window.c (update_sm_hints): ditto
* src/keybindings.c (reload_modmap): ditto
(meta_display_process_key_event): ditto
2002-12-15 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_new): get window name before anything
else.
* src/xprops.c (validate_or_free_results): instead of suggesting
how to get window title etc. with xprop, just print out the
window title. much better.
2002-12-09 Havoc Pennington <hp@pobox.com>
* AUTHORS: add myself here, bug #100789
* src/display.c (meta_display_set_grab_op_cursor): drop
PointerMotionHintMask
* src/window.c (meta_window_handle_mouse_grab_op_event): don't use
XQueryPointer, as we aren't using PointerMotionHint now
* src/display.c (event_callback): rearrange a bit of code
for slight speedup and clarity
* src/window.c (update_resize)
(meta_window_handle_mouse_grab_op_event): implement
usage of the _METACITY_UPDATE_COUNTER
(meta_window_handle_mouse_grab_op_event): fix code that
used event->xbutton with a motion event
* src/display.c (meta_display_open): add new atoms, and
initialize Xsync if we have it
(grab_op_is_resizing): new function
(meta_display_begin_grab_op): create an alarm monitoring
window's _METACITY_UPDATE_COUNTER
(meta_spew_event): conditionalize this on WITH_VERBOSE_MODE
and print alarm events.
* src/window.c (meta_window_new): fetch _METACITY_UPDATE_COUNTER
* configure.in (HAVE_XSYNC): check for Xsync extension
2002-12-08 Havoc Pennington <hp@pobox.com>
* README: updates
* src/window.c (MAX_RESIZES_PER_SECOND): change to 20 instead of
30, just as an experiment.
(MOVE_THRESHOLD): change 15 to 20
(RESIZE_THRESHOLD): change 15 to 20
* src/util.c (ensure_logfile): kill this function when verbose
mode is disabled.
2002-12-08 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_fill_vertical)
(meta_window_fill_horizontal): new functions to resize to
fill screen
* src/keybindings.c: add vert, horiz maximize
* src/prefs.c: had vert, horiz maximize
* src/metacity.schemas.in: shorten some overlong short
descriptions that make the keybindings capplet look ugly.
Add maximize_vertically, maximize_horizontally keys.
2002-12-08 Havoc Pennington <hp@pobox.com>
* src/prefs.c (meta_prefs_get_application_based): make this always
return FALSE for now, to avoid bug reports.
* src/util.c (ensure_logfile): put "opened log file" message on
stderr so it will normally land in ~/.xsession-errors
* configure.in: remove extra AC_ARG_PROGRAM
* src/display.c (event_callback): handle the toggle-verbose message
* src/tools/metacity-message.c: add a toggle-verbose message, been
meaning to do this for a while.
* src/util.c (meta_set_verbose): if verbose mode is enabled and we
don't support it, then exit.
* src/prefs.c: allow building without gconf (currently means some
prefs are no-ops)
* src/util.c, src/util.h: support defining macros to
kill all verbose output entirely. (Removes the code and strings
associated with it)
* configure.in: don't get METACITY_PROPS_LIBS if not building the
config dialog.
(HAVE_GCONF): allow building sans gconf, if you are size-sensitive
and not using gnome.
(WITH_VERBOSE_MODE): add ability to disable all the verbose debug
spew strings, to shrink the binary.
(--disable-sm): allow SM support to be forced on or off
(--disable-startup-notification): allow forcing this on or off
2002-11-30 Havoc Pennington <hp@pobox.com>
* src/screen.c (STARTUP_TIMEOUT): lengthen to 15 seconds
* src/util.c (utf8_fputs): hmm, return a value
* src/screen.c (meta_screen_apply_startup_properties): new
function to apply initial workspace based on startup sequence.
* src/window.c (meta_window_new): load _NET_STARTUP_ID
(meta_window_get_startup_id): new function
* src/window-props.c (meta_display_init_window_prop_hooks): add
hooks for _NET_STARTUP_ID
* src/display.c (event_callback): send property events to
groups.
* src/xprops.c (meta_prop_get_values): make a type of INVALID
mean to ignore that property (don't fetch its value).
* src/group.c (meta_group_property_notify): new function
* src/screen.c (set_supported_hint): support _NET_STARTUP_ID
* src/display.c (meta_display_open): add _NET_STARTUP_ID to atoms
we initialize
* src/group-private.h: private header shared between
group-props.c, group.c
* src/group-props.h, src/group-props.c: new files to contain
functions for retrieving group properties
* src/window.c (meta_window_same_application): change this a bit
to work with new definition of group
* src/group.c (meta_window_get_group): always create a group for
every window, using the window's own ID as group leader if
required.
* src/window.c (update_wm_hints): handle changes to group leader
* src/group.c (meta_window_group_leader_changed): new function
* src/display.h (struct _MetaDisplay): _NET_WM_WINDOW_TYPE_SPLASH,
not SPLASHSCREEN. Reported by Gregory Merchan and Matthias Clasen.
* src/screen.c (startup_sequence_timeout): when timing out a
startup sequence, send a remove message, don't just time it out
locally.
2002-11-22 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_change_workspace): patch from
Hidetoshi Tajima to move a window's transients when moving
the window between workspaces. #98900
2002-11-21 Havoc Pennington <hp@pobox.com>
* src/display.c (meta_display_open): init ret_to to
RevertToPointerRoot out of sheer paranoia; don't want no
RevertToNone in my code!
2002-11-19 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_client_message): do a
recalc_window_features after setting new wm_state in order
to update skip_pager in addition to wm_state_skip_pager
(set_net_wm_state): base _NET_WM_STATE on skip_pager not
wm_state_skip_pager, ditto for skip_taskbar
2002-11-03 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_new): use window-props.h stuff for a
couple of properties
(implement_showing): fix printf string
* src/xprops.c (meta_prop_free_values): new function
* src/window-props.h, src/window-props.c: start moving code that
handles loading window properties into this file.
2002-11-03 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_calc_showing): split into "see if we
should be showing" and "actually show/hide" functions
(idle_calc_showing): rework to first unmap all newly-hidden
windows from bottom to top then map all newly-showing windows from
top to bottom resulting in fewer exposes, #95220
2002-10-28 Havoc Pennington <hp@pobox.com>
* src/window.c, src/display.c: store the window menu on the
display and blow it away when a window closes, so we don't
get funny stuck menus. Patch from Martin Garton #87514
2002-10-27 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_free): move
meta_window_shutdown_group() much earlier in the destroy process.
May fix#96928 tracked down by Kjartan Maraas and Martin Garton.
* src/group.c (meta_window_get_group): never add window to a group
after we've started unmanaging the window
2002-10-26 Havoc Pennington <hp@pobox.com>
* src/xprops.c (meta_prop_get_text_property): new function
(meta_prop_get_wm_hints): new function
(meta_prop_get_class_hint): new function
2002-10-26 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_new): use multi-value-get on a couple
of properties for testing
* src/xprops.c (meta_prop_get_values): implement multi-value-get
* src/window.c (update_mwm_hints): XFree motif hints as we changed
it to use Xmalloc
* src/xprops.c: massively rework this to set up a
get-multiple-properties-at-once API.
* src/async-getprop.c (ag_Xmalloc): new function
2002-10-24 Havoc Pennington <hp@pobox.com>
* src/window.c (meta_window_get_icon_geometry): make public
* src/screen.c (meta_screen_ensure_tab_popup): put the alt+tab
highlight-window indicator on the icon, not the window itself,
if the window is minimized.
2002-10-21 Havoc Pennington <hp@redhat.com>
Optimizations for managing new windows (do not all take effect if
METACITY_DEBUG=1). Bug #96404
* src/keybindings.c (meta_change_keygrab): use error trap nesting
and conditionalize on meta_is_verbose() to avoid a ton of XSync
* src/display.c (meta_change_button_grab): ditto
Throughout: move to new error trap setup to save on XSync calls,
new setup is:
* src/errors.c (meta_error_trap_push_with_return): new function,
an error trap that needs to care about return value and thus
sync even if an outer trap still exists
(meta_error_trap_pop_with_return): new function
(meta_error_trap_pop): add "last_request_was_roundtrip"
argument allowing us to avoid XSync() if we just did
a GetProperty or whatever.
* src/util.c (meta_warning): flush the warning file descriptor
* src/Makefile.am (INCLUDES): define G_LOG_DOMAIN
2002-10-18 Havoc Pennington <hp@redhat.com>
* src/window.c (meta_window_refresh_resize_popup): only create the
resize popup if width_inc or height_inc are > 1
* src/resizepopup.c: Clear out all the weird tickmark cruft,
saves us about 2.5K of binary size, whee
(meta_ui_resize_popup_new): take display/screen arguments and make
multihead-safe #94349
2002-10-16 Havoc Pennington <hp@redhat.com>
* src/workspace.c: workspaces are all per-screen now, fix
accordingly
* src/core.c: fix multihead workspace stuff
* src/keybindings.c: multihead-rama
* src/screen.c (meta_screen_show_desktop): new functions to
replace display equivalents
* src/display.c (meta_display_get_workspace_by_screen_index): get
rid of this
(meta_display_get_workspace_by_index): get rid of this
(event_callback): handle _NET_SHOWING_DESKTOP message per-screen
* src/screen.c (meta_screen_get_workspace_by_index): new function
* src/screen.h (struct _MetaScreen): move workspace list, and
showing_desktop flag, to be per-screen
* src/window.c (window_query_root_pointer): return whether pointer
is on window's screen
(meta_window_handle_mouse_grab_op_event): don't use coordinates
from other screens when updating a window operation on the current
screen. I can't believe no one has reported this...
