2006-02-10 Thomas Thurman <thomas thurman org uk>
Avoid a memory leak when checking which workspace(s) a window is
on. Fixes#322059.
* src/workspace.h (struct MetaWorkspace):
* src/workspace.c (meta_workspace_new, meta_workspace_free):
added list_containing_self member to MetaWorkspace
* src/window.c (meta_window_get_workspaces): use
window->workspace->list_containing_self instead of allocating (and
leaking) such a list on the fly.
2006-01-20 Elijah Newren <newren gmail com>
Avoid flashing when closing a maximized window. Fixes#317254.
* src/window.c (unmaximize_window_before_freeing): new function
that just fixes the net_wm_state and sends a configure_notify,
(meta_window_free): use unmaximize_window_before_freeing() instead
of meta_window_unmaximize() to avoid flicker
2006-01-15 Elijah Newren <newren@gmail.com>
Fix accidental overzealous focus holding by the terminal
introduced by the original patch in bug 326159. Windows launched
from panel icons, the panel menu, or global keybindings should get
focus now. #326159.
* src/display.c (meta_display_open, event_callback):
* src/display.h (struct MetaDisplay):
* src/keybindings.c (process_event):
* src/window.c (meta_window_set_user_time):
Add a new allow_terminal_deactivation field to MetaDisplay and use
it to track whether the user's last action was interaction with
the terminal or some outside action (global keybinding, clicking
on a dock, etc.) likely to launch a new window.
* src/window.c (window_state_on_map):
Allow the focus switch from a terminal to something else if
allow_terminal_deactiviation is true.
* src/keybindings.c (handle_panel_keybinding):
Remove some unneeded code.
Fri Jan 13 14:40:19 2006 Søren Sandmann <sandmann@redhat.com>
* configure.in: Add a dependency on libcm when building with
compositor.
* src/window.c (meta_window_hide): Make this function static.
* src/window.c (implement_showing): Use meta_compositor_minimize()
to do a weird minimize effect.
* src/compositor.[ch]: Beginning of new GL based compositor.
* src/screen.h (struct _MetaScreen): Add void pointer to
compositor data.
* src/screen.c (meta_screen_new): Remove obsolete compositor
stuff; initialize compositor_data. Don't composite manage screen
out of this function.
* src/errors.c (x_error_handler): Check that display is non-NULL
before using it. Add comment about how that can happen.
* src/display.c (meta_display_{begin,end}_grab_op): Remove
explicity damage of windows.
* src/display.c (meta_display_open): Composite manage all the
screens.
2006-01-10 Elijah Newren <newren@gmail.com>
Add a raise on click option, basically only because all the major
distros are patching it in anyway. See #326156.
* src/metacity.schemas.in: add the new gconf key and explanation
* src/prefs.[ch] (#define KEY_RAISE_ON_CLICK, static gboolean
raise_on_click, update_raise_on_click, meta_prefs_init,
change_notify, meta_prefs_get_raise_on_click,
meta_preference_to_string):
Add all the normal preference handling stuff for this new
raise-on-click option.
* src/core.c (meta_core_show_window_menu):
* src/display.c (event_callback, meta_display_begin_grab_op):
* src/window.c (window_activate, meta_window_configure_request, ):
Only raise the window if in raise_on_click mode.
* src/display.c (meta_display_begin_grab_op,
meta_display_end_grab_op, meta_display_check_threshold_reached):
* src/display.h (struct MetaDisplay):
* src/window.c (meta_window_handle_mouse_grab_op_event):
if not in raise-on-click mode only raise on button release if the
click didn't start a move or resize operation; needs a few extra
MetaDisplay fields to handle this
* src/core.c (meta_core_user_lower_and_unfocus):
no need to do the MRU shuffling if not maintaining the stacking
order == MRU order invariant
* src/frames.c (meta_frames_button_press_event):
* src/window.c (meta_window_begin_grab_op):
remove an unneeded window raising that is already handled elsewhere
2006-01-10 Elijah Newren <newren@gmail.com>
Don't "steal" focus from terminal windows for new window mappings
as the difference in usage between terminals and other apps seems
to suggest this difference in treatment. See #326159 for details,
feedback welcome.
* src/window.[ch] (__window_is_terminal): New function, currently
an ugly hack and should be replaced by a new property set by
applications if the behavior works to our liking,
(window_state_on_map): don't transfer focus to new windows from
terminals unless the new window is a transient of the focused
terminal
* src/keybindigns.c (handle_panel_keybinding): panel run dialog
keybinding should be counted as an explicit transfer of focus to
the new window, so override the
don't-transfer-focus-from-terminals in this case
2006-01-09 Elijah Newren <newren@gmail.com>
More thorough handling of source indication. Part of #326041.
* src/window.c (window_activate): new function based off the old
meta_window_activate but which also takes source indication into
account, (meta_window_active): just call window_activate() with
the necessary source indication to get the behavior wanted,
(meta_window_client_message): check source indication too for
_net_active_window messages
* src/window.h (enum MetaClientType): convenience enum for source
indication handling
2006-01-09 Elijah Newren <newren@gmail.com>
Make the taskbar less flash happy and fix up some related stacking
issues. #326035.
* src/window.c (windows_overlap): new function,
(meta_window_show): if a window is denied focus but doesn't
overlap with the focus window there is no need to set the demands
attention hint nor stack that window below the focus window,
(meta_window_get_outer_rect): we're not modifying the window so
declare it to be const
2006-01-02 Elijah Newren <newren@gmail.com>
Patch from Björn Lindqvist to fix a logic error. #322149.
* src/window.c (update_resize): && should have been ||.
2005-12-27 Elijah Newren <newren@gmail.com>
Make the workspace switcher work with dual-head (non-xinerama)
setups. Fixes#319423.
* src/display.c (meta_display_open, event_callback,
meta_display_focus_the_no_focus_window):
* src/display.h (struct MetaDisplay,
meta_display_focus_the_no_focus_window):
* src/keybindings.c (primary_modifier_still_pressed):
* src/screen.c (meta_screen_new):
* src/screen.h (struct MetaScreen):
* src/window.c (meta_window_new_with_attrs, meta_window_show):
* src/workspace.c (meta_workspace_focus_default_window):
Replace display->no_focus_window with a no_focus_window for each
screen.
* src/display.[ch] (meta_display_xwindow_is_a_no_focus_window,
event_callback):
* src/window.c (meta_window_new_with_attrs):
New utility function, meta_display_xwindow_is_a_no_focus_window(),
for checking if the given xwindow is a no_focus_window for one of
the screens.
2005-12-12 Elijah Newren <newren@gmail.com>
* src/window.c (update_net_frame_extents): make the debugging
message actually correspond to the code. Patch from Björn
Lindqvist. Fixes#322051.
2005-11-22 Elijah Newren <newren@gmail.com>
Don't allow removing a window from maximized or fullscreened state
to place the titlebar under the top panel. Fixes#322075.
* src/display.c (handle_net_moveresize_window): fix up previous
comments now that I know a little more, modify the code just
slightly to clarify that this is NOT a manual user move/resize
operation
* src/window.c (meta_window_unmaximize,
meta_window_unmake_fullscreen,
meta_window_shove_titlebar_onscreen):
don't claim that these are manual user move/resize operations
2005-11-18 Elijah Newren <newren@gmail.com>
Merge of all the changes on the constraints_experiments branch.
This is just a summary, to get the full ChangeLog of those
changes (approx. 2000 lines):
cvs -q -z3 update -Pd -r constraints_experiments
cvs -q -z3 diff -pu -r CONSTRAINTS_EXPERIMENTS_BRANCHPOINT ChangeLog
Bugs fixed:
unfiled - constraints.c is overly complicated[1]
unfiled - constraints.c is not robust when all constraints
cannot simultaneously be met (constraints need to be
prioritized)
unfiled - keep-titlebar-onscreen constraint is decoration
unaware (since get_outermost_onscreen_positions()
forgets to include decorations)
unfiled - keyboard snap-moving and snap-resizing snap to hidden
edges
109553 - gravity w/ simultaneous move & resize doesn't work
113601 - maximize vertical and horizontal should toggle and be
constrained
122196 - windows show up under vertical panels
122670 - jerky/random resizing of window via keyboard[2]
124582 - keyboard and mouse snap-resizing and snap-moving
erroneously moves the window multidimensionally
136307 - don't allow apps to resize themselves off the screen
(*cough* filechooser *cough*)
142016, 143784 - windows should not span multiple xineramas
unless placed there by the user
143145 - clamp new windows to screensize and force them
onscreen, if they'll fit
144126 - Handle pathological strut lists sanely[3]
149867 - fixed aspect ratio windows are difficult to resize[4]
152898 - make screen edges consistent; allow easy slamming of
windows into the left, right, and bottom edges of the
screen too.
154706 - bouncing weirdness at screen edge with keyboard moving
or resizing
156699 - avoid struts when placing windows, if possible (nasty
a11y blocker)
302456 - dragging offscreen too restrictive
304857 - wireframe moving off the top of the screen is misleading
308521 - make uni-directional resizing easier with
alt-middle-drag and prevent the occasional super
annoying resize-the-wrong-side(s) behavior
312007 - snap-resize moves windows with a minimum size
constraint
312104 - resizing the top of a window can cause the bottom to
grow
319351 - don't instantly snap on mouse-move-snapping, remove
braindeadedness of having order of releasing shift and
releasing button press matter so much
[1] fixed in my opinion, anyway.
[2] Actually, it's not totally fixed--it's just annoying
instead of almost completely unusable. Matthias had a
suggestion that may fix the remainder of the problems (see
http://tinyurl.com/bwzuu).
[3] This bug was originally about not-quite-so-pathological
cases but was left open for the worse cases. The code from
the branch handles the remainder of the cases mentioned in
this bug.
[4] Actually, although it's far better there's still some minor
issues left: a slight drift that's only noticeable after
lots of resizing, and potential problems with partially
onscreen constraints due to not clearing any
fixed_directions flags (aspect ratio windows get resized in
both directions and thus aren't fixed in one of them)
New feature:
81704 - edge resistance for user move and resize operations;
in particular 3 different kinds of resistance are
implemented:
Pixel-Distance: window movement is resisted when it
aligns with an edge unless the movement is greater than
a threshold number of pixels
Timeout: window movement past an edge is prevented until
a certain amount of time has elapsed during the
operation since the first request to move it past that
edge
Keyboard-Buildup: when moving or resizing with the
keyboard, once a window is aligned with a certain edge
it cannot move past until the correct direction has
been pressed enough times (e.g. 2 or 3 times)
Major changes:
- constraints.c has been rewritten; very few lines of code from
the old version remain. There is a comment near the top of
the function explaining the basics of how the new framework
works. A more detailed explanation can be found in
doc/how-constraints-works.txt
- edge-resistance.[ch] are new files implementing edge-resistance.
- boxes.[ch] are new files containing low-level error-prone
functions used heavily in constraints.c and edge-resistance.c,
among various places throughout the code. testboxes.c
contains a thorough testsuite for the boxes.[ch] functions
compiled into a program, testboxes.
- meta_window_move_resize_internal() *must* be told the gravity
of the associated operation (if it's just a move operation,
the gravity will be ignored, but for resize and move+resize
the correct value is needed)
- the craziness of different values that
meta_window_move_resize_internal() accepts has been documented
in a large comment at the beginning of the function. It may
be possible to clean this up some, but until then things will
remain as they were before--caller beware.
- screen and xinerama usable areas (i.e. places not covered by
e.g. panels) are cached in the workspace now, as are the
screen and xinerama edges. These get updated with the
workarea in src/workspace.c:ensure_work_areas_validated()
2005-10-03 Elijah Newren <newren@gmail.com>
Patch from Ross Cohen to make alt-esc (show windows instantly)
actually show minimized windows too. Fixes#107072.
* src/keybindings.c (process_tab_grab): initialize tab_unminimized
to FALSE for the target window when starting the grab, when
advancing through the list check to find the previous window and
re-minimize it if it was tab-unminimized, unminimize the new
window we're alt-esc'ing to if it's minimized, (do_choose_window):
raise and unminimize the initial window as well in alt-esc'ing
* src/window.h (struct _MetaWindow): add a tab_unminimized field
* src/window.c (meta_window_new_with_attrs): initialize
tab_unminimized to false
2005-10-03 Elijah Newren <newren@gmail.com>
Patch from Björn Lindqvist to have ancestors come along with the
transient when moving the window from one workspace to another.
Fixes#314977.
* src/window.c (meta_window_change_workspace): have all ancestors
change workspaces too
2005-08-03 Ray Strode <rstrode@redhat.com>
Improve the behavior of keyboard move/resize and edge
snapping. Still not perfect, bug 310888.
* src/effects.c (draw_xor_rect): Make the outside of a
wireframe rectangle line up with the outside edge of its
window, instead of centering the wireframe edges on the
window edges.
* src/keybindings.c (process_keyboard_move_grab): allow
edge snapping in wireframe mode. Adjust code to take
into account changed semantics of find_next_*_edge
functions.
(process_keyboard_resize_grab_op_change): new function
to take some orthogonal logic out of
process_keyboard_resize_grab_op. Only allow keyboard
resize cursor to go to flat edges, not corners.
(process_keyboard_resize_grab): allow edge snapping in
wireframe mode. Fix up snapping logic.
* src/place.c (get_{vertical,horizontal}_edges): use
GArray instead of int *, since the number of output
edges isn't known until the middle of the function now.
Use xor rect extents instead of window extends if in
wireframe mode.
(meta_window_find_next_{vertical,horizontal}_edge: add
new source_edge_position parameter to specify which edge
on the active window to start from when looking for next
edge on the screen. Return the coordinate of the edge
found and not the coordinate of where the window should be
moved to snap to where the edge was found.
* src/window.c (update_move): all the user to specify
an edge to resize with mouse in keyboard resize mode.
window
2005-07-24 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_raise): raise the window as well as
its ancestor; fixes a stacking bug with an ancestor that has more
than one child window. Fixes one of the issues in #307875.
2005-07-24 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_free): restore original window size if
the window was maximized, as the FIXME says. ;-) Fixes#137185.
2005-07-23 Elijah Newren <newren@gmail.com>
* src/window.c: (meta_window_activate): revert the patch from
#128380--change _NET_ACTIVE_WINDOW behavior to what it originally
was.
2005-06-02 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_configure_request): If a window
requests to be raised and is denied, set the demands attention
hint. Fixes#305882.
2005-06-02 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_focus): if the window has a modal
transient which is being unmanaged, don't focus it. Fixes the
Metacity issue reported in #305362.
2005-05-26 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_configure_request): Patch from Greg
Hudson to make sure window position is calculated correctly for
reconfigure requests when part of the XWindowChanges structure is
uninitialized. Fixes#305257.
2005-05-26 Elijah Newren <newren@gmail.com>
* src/window.c: (meta_window_new_with_attrs): put all transients
of the new window, if any exist, in the calc_showing queue. Fixes
#303284.
2005-02-23 Elijah Newren <newren@gmail.com>
* src/window.c: (meta_window_new_with_attrs): Fix crash that
occurs when stupid apps claim that a window is its own parent.
#168207
2005-02-21 Elijah Newren <newren@gmail.com>
Handle keynav vs. mousenav in mouse and sloppy focus modes. Fixes
#167545.
* doc/how-to-get-focus-right.txt: Update due to this new method
for handling keynav vs. mousenav, plus various other updates that
I previously forgot.
* src/display.h: (struct _MetaDisplay): add a mouse_mode boolean
* src/display.c: (meta_display_open): initialize mouse_mode to
true, (event_callback): have EnterNotify and LeaveNotify events
set mouse_mode to true when focusing a window
* src/keybindings.c: (process_tab_grab): set mouse_mode to false
when using alt-tab/alt-esc, (do_choose_window): likewise,
(do_handle_move_to_workspace): set mouse_mode to false on
move-window-to-workspace-<n> keybindings
* src/window.c (idle_calc_showing): if we're in keynav mode while
using sloppy or mouse focus, use metacity_sentinel to avoid
EnterNotify events being generated from events other than mouse
movement.
* src/workspace.c (meta_workspace_activate_with_focus): add a
FIXME in a potentially duplicate section of code,
(meta_workspace_focus_default_window): use the same focus choice
as click-to-focus if in keynav mode.
2005-02-20 Elijah Newren <newren@gmail.com>
* src/window.c: (meta_window_activate): when receiving a
_NET_ACTIVE_WINDOW message, switch to the desktop where the window
is located before activating instead of moving the window to the
current desktop. Thanks to Lubos Lunak for catching this issue.
Fixes#128380.
2005-02-20 Elijah Newren <newren@gmail.com>
* src/window.c (meta_window_show): Ignore all focus and
focus-stealing-prevention code in meta_window_show when not
showing the window for the first time. Fixes#167199.
2005-02-20 Elijah Newren <newren@gmail.com>
Big patch to cover about 6 different issues in order to correct
rare problems with timestamps (make sure window selected in
tasklist actually gets focus, sanity check timestamps to avoid
rogue apps hosing the system, correct the updating of
net_wm_user_time, correctly handle timestamps of 0 when comparing
xserver timestamps for those who have had their systems up for
over 25 days or so, add some debugging information to verbose
logs, some code cleanups). Fixes all issues listed in #167358.
* src/display.h: (struct _MetaDisplay): clarify comment on
last_focus_time, introduce a new variable--last_user_time,
(XSERVER_TIME_IS_BEFORE macro): put this functionality into a
separate macro and then introduce a new macro with this name that
uses the old one but adds additional special-case checks for
timestamps that are 0, (comment to
meta_display_set_input_focus_window): add information about how
last_user_time should be used in this function
* src/display.c (santiy_check_timestamps): new function,
(meta_display_open): intialize display->last_user_time,
(meta_display_get_current_time_roundtrip): use the timestamp,
which is known to be good, in order to sanity_check_timestamps,
(event_callback): use the new meta_window_ste_user_time() function
in order to correct problems, use the timestamp of KeyPress and
ButtonPress events, which are known to be good, in order to
sanity_check_timestamps, (timestamp_too_old): new function for
common behavior of meta_display_focus_the_no_focus_window and
meta_display_set_input_focus_window, with added checking for
display->last_user_time in addition to display->last_focus_time,
(meta_display_set_input_focus_window): replace some of the code
with a call to timestamp_too_old(),
(meta_display_focus_the_no_focus_window): replace some of th ecode
with a call to timestamp_too_old()
* src/window.h: (meta_window_set_user_time): new function to
abstract the many things that need to be done when updating the
net_wm_user_time of any window
* src/window.c: (meta_window_activate): add debugging spew, make
sure the comparison is made with last_user_time NOT
last_focus_time, use meta_window_set_user_time() function in order
to correct problems, (meta_window_client_message): add a newline
to a debugging message to make them easier to read,
(meta_window_set_user_time): new function
* src/window-props.c (reload_net_wm_user_time): use the new
meta_window_ste_user_time() function in order to correct problems
2005-02-12 Elijah Newren <newren@gmail.com>
Raise the ancestor of a window instead of the window itself.
Fixes#166894.
* src/window.c: (find_root_ancestor): new function,
(meta_window_raise): get the ancestor of the given window and
raise it if possible instead of the window
2005-02-12 Elijah Newren <newren@gmail.com>
Don't unconditionally place splashscreens (and other
not-to-befocused windows) below the focus window. Fixes#167042.
* src/window.c: (intervening_user_event_occurred): new function
taken from the timestamp comparison portion of the old
window_takes_focus_on_map function, (window_state_on_map): new
function with remainder of old window_takes_focus_on_map function
that determines both whether the window will take focus and
whether it should be placed on top, (meta_window_show): use
place_on_top_on_map to determine window stacking instead of trying
to infer it from takes_focus_on_map
2005-02-11 Elijah Newren <newren@gmail.com>
Avoid new windows being obscured by the focus window (and thus
possibly lost). Fixes#166524.
* src/place.c: new MetaWindowDirection enum,
(find_most_freespace): new function to find where there is the
most space available around the focused window,
(meta_window_place): if a window is denied focus and the window
overlaps the focused window, retry the first-fit algorithm only
paying attention to the focus window position and if that fails
just find the location on the screen with the most space
available.
* src/window.h: (struct MetaWindow): new
denied_focus_and_not_transient bitfield
* src/window.c: (meta_window_new_with_attrs): initialize
denied_focus_and_not_transient, (meta_window_show): set and unset
the denied_focus_and_not_transient field appropriately
2005-02-08 Elijah Newren <newren@gmail.com>
Avoid using CurrentTime when focusing, handle it better in case we
miss any cases. Fixes#166732.
* src/window.c: (meta_window_shade): use
meta_display_get_current_time_roundtrip() to ensure we have a
valid timestamp, (meta_window_unshade): same
* src/display.c: (meta_display_set_input_focus_window): If
CurrentTime was passed, get one from the XServer in addition to
throwing a warning, (meta_display_focus_the_no_focus_window): same
2005-02-08 Elijah Newren <newren@gmail.com>
* src/window.c: (meta_window_activate): If we're not passed a
timestamp, make sure to manually get one. Fixes#166728.
2005-02-06 Elijah Newren <newren@gmail.com>
Ignore xconfigurerequest events for stacking when it should be
safe to do so. Again, thanks to Crispin Flowerday for the test
case. Thanks to KWin for the inspiration (and to Google for
indexing their source code). Fixes the other half of #166395.
* src/window.c: (meta_window_configure_request): if the
active_window is from a separate application than the one getting
the configure request and the net_wm_user_time of the active
window is later than that of the window getting the configure
request, then ignore the request.
2005-02-06 Elijah Newren <newren@gmail.com>
If activation requests are too old, set the demands_attention hint
instead of actually activating. Thanks to Crispin Flowerday for
the test case and for testing the patch. Fixes half of #166395.
* src/window.c: (meta_window_activate): if the request came before
the last focus time, set the demands attention hint instead
2005-02-04 Elijah Newren <newren@gmail.com>
Make sure window->border_only is initialized so we don't get
random windows without decorations. Thanks to Sinisa Segvic and
Owen Taylor for providing test cases. Fixes#145131.
* src/window.c: (update_mwm_hints): Be sure to call
recalc_window_features even if no MWM hints are set
2005-01-31 Elijah Newren <newren@gmail.com>
Correct the stacking when return from fullscreen mode. Fixes
#165718.
* src/window.c: (meta_window_unmake_fullscreen): Update the layer
after resizing the window
2005-01-28 Elijah Newren <newren@gmail.com>
* src/menu.c: Change wording of menu to "Always on Current
Workspace" from "Put on All Workspaces", remove a quick-key
conflict between "On _Top" and "Only on _This Workspace" by
switching the latter to "_Only on This Workspace"
* src/window.c: Remove a comment that is no longer necessary
(since bug 87531 has been fixed)
2005-01-28 Elijah Newren <newren@gmail.com>
Take into account the appropriate list of windows when placing a
new one. Fixes#165381.
* src/place.c: (meta_window_place): use
meta_window_showing_on_its_workspace(w) instead of !w->minimzed,
also take into account sticky windows
* src/window.[ch]: rename window_showing_on_its_workspace to
meta_window_showing_on_its_workspace and export it
