c4d87e2f81
handle destroynotify events
2010-04-17 17:27:53 +02:00
d973f30fc2
push X11 changes after a window is mapped
2010-04-17 16:43:34 +02:00
68542f3c22
When assigning children to containers, reset their x window state
2010-04-17 16:41:20 +02:00
53dcebfd8a
put container in fullscreen mode if the fullscreen state is set when mapping
2010-04-17 14:21:34 +02:00
c56867792a
handle client messages (fullscreen window state)
2010-04-17 13:54:45 +02:00
22f38ebde4
clear current_match when dropping state
2010-04-17 12:57:06 +02:00
77ec4219c9
make floating an enum (we need three states, not only two)
2010-04-16 22:57:21 +02:00
8d05039b04
move match_* to match.c
2010-04-16 22:51:25 +02:00
bb220b27d7
check for empty matches
2010-04-16 22:50:20 +02:00
e0b7ae872e
move con_focus to con.c
2010-04-16 21:04:36 +02:00
4796798074
better debug messages
2010-04-16 21:02:34 +02:00
138a790cd0
clean up old workspace when switching (makes test 2 of t/16-*.t pass)
2010-04-16 20:59:21 +02:00
2534f21940
ignore sequence of unmapnotify events (generates enternotify events)
2010-04-16 20:56:49 +02:00
93600ce0fd
implement con_id for matching containers, extend testcase
2010-04-16 15:30:07 +02:00
769501420d
add first version of a new flex/bison based command parser
2010-04-14 20:26:56 +02:00
24725cd94a
re-add fullscreen mode
2010-04-13 20:51:43 +02:00
8959c5005f
cleanups
2010-04-13 19:33:40 +02:00
8e5a831e27
re-add focus follows mouse handling
2010-04-13 18:43:37 +02:00
eec762ea8f
more reformatting/cleanups
2010-04-13 17:52:23 +02:00
dd7acf73e9
re-add support for legacy window titles (WM_NAME)
2010-04-13 17:46:54 +02:00
fd8735a6fd
correctly update/display window title/class
2010-04-13 17:22:34 +02:00
bcfb0d2505
more reformatting
2010-04-13 16:48:42 +02:00
4206db2839
workspace.c: update header, reformat
2010-04-13 16:43:08 +02:00
6aa6fa0af0
correctly focus workspaces
2010-04-13 16:41:23 +02:00
c145f7e529
first step of the big refactoring ("tree" branch).
...
From here on, we can track changes. It made no sense to put the
development up to this point into git.
2010-04-13 13:17:39 +02:00
41b6631f68
Bugfix: null-terminate buffer
2010-03-26 03:04:54 +01:00
a542515f9e
Fix memory leaks
2010-03-26 01:52:39 +01:00
7bd4ea3699
randr: Don’t enable outputs with mode 0x0
...
Just a sanity check for some possibly broken drivers.
2010-03-26 01:52:08 +01:00
a151fd95a8
remove newlines
2010-03-26 00:28:30 +01:00
6fd56757f0
Bugfix: Assign all workspace to new outputs as new outputs get available (Thanks badboy)
2010-03-26 00:13:28 +01:00
1b8299002e
Bugfix: Translate keysyms to keycodes before entering mode (Thanks fallen)
2010-03-25 21:18:46 +01:00
6699d54640
Fix rendering of workspace names after "reload" (Thanks fallen)
2010-03-25 19:08:37 +01:00
469f22caeb
Bugfix: Correctly switch workspace when using the "jump" command (Thanks fallen)
2010-03-25 18:07:40 +01:00
2c42c0c760
Bugfix: Correctly check bitmask for floating_modifier
2010-03-25 02:53:53 +01:00
46e7cf5fe1
Handle destroy notify events like unmap notify events
...
This helps for windows which are immediately destroyed instead of
unmapped, like when starting i3status | ./foobar | dzen2 -dock
and foobar does not exist (i3status and dzen2 will get a SIGPIPE).
2010-03-24 16:52:16 +01:00
e90e80c87d
Bugfix: fix state of keypresses in sighandler (like in i3-input)
2010-03-24 16:10:47 +01:00
7eea1067f8
Bugfix: correctly translate coordinates for floating windows when outputs change
2010-03-24 16:09:43 +01:00
0f7ac09c7b
Bugfix: Don’t unmap windows when current workspace gets reassigned
...
Steps to reproduce were:
1) xrandr --output VGA1 --auto --left-of LVDS1
2) open a terminal on VGA1
3) xrandr --output VGA1 --off
2010-03-24 16:06:21 +01:00
56139f3656
Bugfix: only restore focus if the workspace is focused, not if it is visible
2010-03-23 14:43:35 +01:00
42b638eac0
Bugfix: Don’t enter BIND_A2WS_COND state too early (Thanks fallen)
2010-03-22 15:12:37 +01:00
77efb29d9f
ipc: send a workspace event when the urgency flag changes
2010-03-20 03:09:42 +01:00
35a791f7e6
ipc: also send workspace event when initializing a workspace for an output
2010-03-20 02:56:23 +01:00
46bd9ac58a
Bugfix: correctly re-assign dock clients when output goes inactive
2010-03-20 02:52:06 +01:00
aec40126b4
ipc: implement output event
2010-03-19 22:40:43 +01:00
4ce0d6f014
ipc: implement GET_OUTPUTS
2010-03-19 22:24:52 +01:00
a607eae53a
ipc: include the urgent flag in the workspaces reply
2010-03-19 22:01:21 +01:00
fdcbec248a
Start dock clients on the output they request to be started on according to their geometry
...
Use dzen2’s -xs option to use this, or specify the coordinates using
-x manually.
2010-03-19 21:44:36 +01:00
3c8c426011
Disable RandR if no outputs are found (Thanks fallen)
2010-03-19 18:51:19 +01:00
ee76b2ebf6
Bugfix: When disabling RandR/Xinerama, give a name to the pseudo-output (Thanks fallen)
2010-03-19 18:51:01 +01:00
af00df9321
Use DLOG for debug messages instead of printf (Thanks kruM)
2010-03-19 18:48:36 +01:00
0bb1b718d1
Fix compilation with the old xcb keysyms api
2010-03-19 16:02:33 +01:00
91b6c69eae
little style fix for the last commit
2010-03-19 01:43:11 +01:00
62c4532329
Bugfix: Take window out of fullscreen before entering floating mode.
2010-03-19 01:24:33 +01:00
1680071555
randr: use effective CRTC width/height, not its mode (Thanks moemoe)
...
This fixes setups which use panning and cloning.
2010-03-17 15:56:26 +01:00
a604af6340
make pointer follow the focus when moving to a different screen also for floating windows
2010-03-17 03:18:13 +01:00
b47d0a8932
Bugfix: configure floating windows above tiling windows when moving them to another workspaces (Thanks Sasha)
2010-03-17 00:36:08 +01:00
f7a1a9fb20
ipc: correctly shutdown IPC sockets when exiting/restarting
2010-03-16 02:44:47 +01:00
f9e6f8ba4b
Bugfix: Correctly stack windows when new windows are opened while in fullscreen mode
...
This fixes ticket #195 .
2010-03-15 22:17:00 +01:00
c738b2e454
Don’t use SYNC key bindings for Mode_switch but re-grab keys
...
Before this commit, i3 used key bindings in SYNC mode for bindings
like Mode_switch + <a> and replayed the key if the current state
did not include Mode_switch. This had some problems:
1) The WM needed to acknowledge much more key presses than you
actually had bindings for, thus making the system a bit laggy
sometimes.
2) Users of layouts who constantly type in the third level (like
russian layouts) did not get their cyrillic symbols correctly
(they were not replayed right), neither did the keybindings
work in both modes.
So, the current implementation uses the following approach: XKB
provides an event which contains the current state (including
the current level). i3 signs up for this event and upon receival,
it re-maps the bindings using Mode_switch (enables them when the
level goes to the third level and disables them as soon as the
level goes back to normal). This fixes both problems.
2010-03-14 22:40:58 +01:00
2df1fb8ac8
Implement disabling the internal workspace bar
2010-03-14 12:59:45 +01:00
2df374ca4c
Add configuration option to turn off workspace bar
2010-03-13 19:15:28 +01:00
b6088b803e
ipc: correctly set "focused" for workspaces
...
Didn’t work on empty workspaces before
2010-03-13 16:27:03 +01:00
9cb35383a8
Implement proportional floating-resize.
2010-03-13 13:22:39 +01:00
3db4890683
ipc: implement events, cleanup the code a bit
2010-03-12 21:05:05 +01:00
d86531b958
ipc: return logical workspace numbers, not internal ones
2010-03-12 15:59:38 +01:00
d6f726283c
ipc: also send a reply for COMMAND messages
2010-03-12 15:29:44 +01:00
5a3d1b38e8
ipc: Correctly deal with SIGPIPE/failing write()s
...
If a client disconnects while i3 still wants to write the reply, this
could lead to exits of i3 before.
2010-03-12 15:02:00 +01:00
0f5256dc72
Floating resize uses arbitrary corners
...
This closes ticket #121
2010-03-12 03:11:20 +01:00
9ed5e00107
ipc: change active to visible, introduce focused
...
visible == currently visible on the output it is on
(multiple workspaces can be visible at the same time)
focused == has the focus (only one workspace can be focused)
2010-03-12 03:01:34 +01:00
7c1be83692
Update comment (Thanks Merovius)
2010-03-12 02:59:16 +01:00
5f370b494c
bugfix: don’t remap stack windows errnously when changing workspaces
...
This fixes ticket #193 (long-standing rendering bug).
2010-03-12 00:41:40 +01:00
93a9f3c244
Bugfix: Don’t mess up x/y coordinates in configurerequests for floating windows
...
This was the cause for ticket #93 , which actually has a false
conclusion for the reason of this bug.
This code needs to be refactored.
2010-03-11 23:34:29 +01:00
e11ca75407
ipc: add active flag
2010-03-11 16:48:48 +01:00
9a9ba1b859
ipc: implement GET_WORKSPACES message type
...
This is the foundation to use dzen2 or similar as a complete
replacement for the internal workspaces bar.
A testcase is included, more documentation about the IPC interface
will follow.
2010-03-11 15:58:39 +01:00
f7c8e76782
Select containers above or near the whole snapped width/height
...
This fixes ticket #100 , and is best explained using a little example.
Consider the following layout:
+---+---+
| | X |
+---+---+
| X |
+---+---+
Where X marks a window, so you have an empty container in the upper
left, the container on the bottom is snapped to the right. Before
this commit, nothing would happen when focusing "above". After
this commit, the upper window gets focused.
2010-03-11 00:15:34 +01:00
a70f4353b7
Fix predict_text_width by using xcb_query_text_extents
...
This fixes ticket #173 , at least for the rendering errors.
I don’t really know why I implemented predict_text_width like it
was before (querying the whole table and pulling out information
one by one). Maybe I have overlooked xcb_query_text_extents. In
any case, it works better now.
2010-03-10 23:27:24 +01:00
bd76e994b8
Re-add old Xinerama code for the poor nvidia users
...
Add --force-xinerama when starting i3 to use Xinerama instead of RandR.
This should *ONLY* be done if you have no other choice (nvidia’s
binary driver uses twinview and does not expose the monitor information
through RandR).
2010-03-09 21:25:54 +01:00
4dfe61c2d4
Fixes for the last commit: check outputs instead of workspaces, correctly reset state when switching workspaces, little style fixes
2010-03-08 11:15:31 +01:00
9c77b0f9a1
Implement screen-spanning fullscreen-mode (command: 'fg')
...
This closes ticket #188
2010-03-08 11:15:31 +01:00
85308715ea
Turn nested functions into real functions or macros
...
This enables compilation with llvm-clang and thus closes ticket #101 .
While it makes the code more ugly, I don’t see a beautiful solution
which would enable us to stay with the more elegant solution of
nested functions and still allow compilation with any other compiler
than gcc.
2010-03-07 19:00:34 +01:00
46642d0e45
For num_clients == 0, the last fix is also valid
2010-03-06 16:58:20 +01:00
afc922ef57
Avoid more errors by not trying to re-create a stack win with height == 0
2010-03-06 16:51:17 +01:00
e9ef672c36
Don’t render if no stack_window has been opened
...
This could happen when you have only one client in a stacking/tabbed
container. While not being a critical bug, this avoids X errors.
2010-03-06 16:34:07 +01:00
6be76e4c56
Remove trailing whitespace
2010-03-05 23:26:45 +01:00
4e9c6515f3
Changing to stacking/tabbing toggles mode to default if already in stacking/tiling.
2010-03-05 23:25:34 +01:00
f1a011c014
Clear urgency hints set by i3 (see HEAD~2)
2010-03-05 20:03:38 +01:00
c367eaa369
put windows with WM_CLIENT_LEADER on the workspace of their leader
...
This fixes ticket #163
2010-03-05 18:01:32 +01:00
40475b250f
Allow unsetting WM_CLIENT_LEADER (XCB_NONE is not an invalid value)
2010-03-05 17:42:50 +01:00
8d648b4e37
Update function names, variable names and documentation for the RandR changes
2010-03-05 16:18:41 +01:00
8b192ac7ed
Bugfix: Correctly hide/show workspaces when enabling new outputs, correctly handle focus (Thanks Merovius)
2010-03-05 15:22:12 +01:00
718d62a3cd
Bugfix: Fix clone mode with new RandR code (Thanks Merovius)
2010-03-05 14:32:48 +01:00
d73f1748db
Remove superfluous definitions (Thanks Merovius)
2010-03-05 01:59:52 +01:00
85437877e8
fix comment (Thanks Merovius)
2010-03-05 00:09:52 +01:00
3cfe1b35e0
Update fullscreen client position/size when an output changes (Thanks Merovius)
...
This fixes #187 .
2010-03-03 09:32:31 +01:00
0ec3481c69
draw consistent borders for each frame in a tabbed/stacked container
...
The condition was a relict from older rendering code, I think.
This fixes #172 .
2010-03-03 08:49:07 +01:00
187c72fa96
sighandler: only display on active outputs
2010-03-03 08:48:57 +01:00
e5f222e03e
Save a resize_client() when handling the size hints
2010-03-03 01:23:38 +01:00
7839b7a4fe
Bugfix: checked for wrong flag in size hints
2010-03-03 01:23:38 +01:00
5dbcb0158f
When in fullscreen mode, focus whole screens instead of denying to focus (Thanks dothebart)
...
This fixes ticket #169 .
2010-03-02 15:45:48 +01:00
e3e7ebe23a
Bugfix: only return active screens in get_screen_most
2010-03-02 15:45:03 +01:00
4e69bd65c0
Move autostart after creating the IPC socket in start process (Thanks Sasha)
...
This fixes ticket #179 .
2010-03-02 15:30:23 +01:00
b53c5861a2
Restore geometry of all windows before exiting/restarting (Thanks Sasha)
...
This fixes ticket #185
2010-03-02 15:25:08 +01:00
d08ec00329
use scalloc instead of some places where calloc was still used
2010-03-02 14:42:24 +01:00
aae824b1f3
Change workspace assignments to use the RandR output name instead of <screen>
2010-03-02 13:35:43 +01:00
818e02ef35
huge change: implement RandR instead of Xinerama
...
Thanks to Merovius for doing a proof of concept on this one and
being a driving force behind the idea.
Using RandR instead of Xinerama means that we are now able to use
the full potential of the modern way of configuring screens. That
means, i3 now has an idea of the outputs your graphic driver
provides, which allowed us to get rid of the ugly way of detecting
changes in the screen configuration which we used before. Now, your
workspaces should not be confused when changing output modes anymore.
Also, instead of having ugly heuristics to assign your workspaces
to (the screen at position X or the second screen in the list of
screens) you will be able to just specify an output name.
As this change basically touches everything, you should be prepared
for bugs. Please test and report them!
2010-03-02 12:47:21 +01:00
6f72970ece
add xcb_set_window_rect which configures a window according to a Rect
2010-02-28 20:35:30 +01:00
7caf98dd18
Move update_if_necessary to util.c, will be necessary later
2010-02-28 18:39:11 +01:00
e209fd7d3a
docs: the comparison is size and position, not only size (Thanks Merovius)
2010-02-22 07:09:17 +01:00
4fcbb7980e
work around clients setting 0xFFFF as resize increments
...
I am not sure if behaviour is specified for this case, but as the
workaround is not really a big deal, why bother…
2010-02-20 20:32:43 +01:00
432f06a21e
Add missing carriage return (Thanks fallen)
2010-02-16 19:55:04 +01:00
e225742415
Make the warning about $terminal being deprecated an error (Thanks fallen)
2010-02-16 19:25:07 +01:00
e78bb02082
parser: ignore errors
2010-02-14 16:59:22 +01:00
2af2bc36c9
Document the new options
2010-02-13 19:58:43 +01:00
64cf88403d
lexer/parser: proper error messages
...
Error messages now look like this:
13.02.2010 19:42:30 - ERROR:
13.02.2010 19:42:30 - ERROR: CONFIG: syntax error, unexpected <word>,
expecting default/stacking/tabbed or stack-limit
13.02.2010 19:42:30 - ERROR: CONFIG: in file "inv", line 15:
13.02.2010 19:42:30 - ERROR: CONFIG: new_container foobar
13.02.2010 19:42:30 - ERROR: CONFIG: ^^^^^^
13.02.2010 19:42:30 - ERROR:
2010-02-13 19:42:54 +01:00
79a4e30488
parser: enable verbose error messages, add aliases to tokens
2010-02-13 15:27:43 +01:00
e446747812
Also warp the pointer when moving a window to a another visible workspace (Thanks Thomas)
2010-02-12 16:19:58 +01:00
87f8c501da
Bugfix: Correctly do boundary checking/moving to other workspaces when moving floating clients via keyboard (Thanks sasha)
2010-02-12 13:06:59 +01:00
0b6b8e8380
Bugfix: Use both parts of WM_CLASS (it contains instance and class) (Thanks fallen)
...
Actually, WM_CLASS contains two null-terminated strings, so we cannot
use asprintf() to get its value but rather use strdup() to get both
of them. Both values are compared when a client is matched against
a wm_class/title combination (for assignments for example).
2010-02-12 12:12:25 +01:00
88b9700cdb
Invert logic for the last commit
...
This makes it more clear that the option is meant to be a special
case (it *disables* part of the focus handling). Also, when
initializing the config data structure with zeros, it will get
initialized with the right value.
Furthermore, the config file parser now also accepts various values
which represent "true", not only numbers.
2010-01-29 21:58:50 +01:00
7f10970fc7
Added focus_follows_mouse config option
2010-01-29 21:58:50 +01:00
36909c6e3b
Fixed cursor code in create_window(), allowing the cursor to be changed
2010-01-26 11:28:38 +01:00
148547c765
Fixed cursor orientation when resizing
2010-01-26 11:28:37 +01:00
7270b74732
Implement resize command for floating clients
2010-01-23 18:57:29 +01:00
fde11f9a5c
Bugfix: Don’t leak IPC socket to launched processes
2010-01-23 00:34:29 +01:00
c606d93630
Feature: Cycle through workspaces
...
On command pw/nw, cycle through all workspaces (starting from
previous/next one) until we reach the current one again.
2010-01-19 12:10:45 +01:00
d949a335e6
bugfix: lexer: return to INITIAL state (Thanks dirkson)
...
'floating_modifier' after 'workspace' did not work because of wrong
state
2010-01-08 18:56:01 +01:00
af0b33bcc3
Bugfix: if a font provides no per-char info for width, fall back to the default (Thanks Ciprian)
2010-01-07 14:38:00 +01:00
b6de565445
Bugfix: Use ev_loop_new instead of ev_default_loop because the latter one blocks SIGCHLD (Thanks Ciprian)
...
SIGCHLD was inherited to child processes started by i3 and not
all of them unblocked it (shells generally did, though), leading to
zombie processes.
2010-01-07 13:36:52 +01:00
44191d5ffc
sighandler: last little style fixes
2010-01-03 22:07:39 +01:00
2b1a132c39
grab the pointer inside the signal handler popup
2010-01-03 22:02:07 +01:00
be33c8e599
Check return code of sigaction()
2010-01-03 21:55:22 +01:00
a69dfb3f06
Remove expose event mask, expose events will not be generated in any case
2010-01-03 21:55:02 +01:00
7b2363776a
Use DLOG instead of LOG, remove unnecessary debug statement
2010-01-03 21:54:47 +01:00
0c35573e51
some little style changes
2010-01-03 21:54:26 +01:00
8afdf54078
Update copyright to 2010
2010-01-03 21:53:30 +01:00
614b360bd4
added popup for handling SIGSEGV or SIGFPE
...
the popup is placed on each of the virtual screens
the user can decide to restart or quit i3
in case of an exit a core-dump is generated
2010-01-03 21:32:50 +01:00
715983024d
ewmh: Don’t push workspaces with width=0 and height=0 (Thanks Atsutane, badboy)
...
Even though i3 cannot know the width/height of some workspaces as
long as they are not initialized (say you used workspace 1 and 3,
but not workspace 2), some applications require this information.
In this case, it was Firefox which intersects the available workareas
(see mozilla/gfx/src/gtk/nsScreenGtk.cpp) and did not position some
windows correctly when being confronted with zero-width/height
workspaces.
2010-01-03 18:36:50 +01:00
7c130fb540
Obey the XDG Base Directory Specification for config file paths
...
This means you can now put your i3 config into
$XDG_CONFIG_HOME/i3/config, which probably is ~/.config/i3/config
if not set otherwise.
2010-01-03 16:58:41 +01:00
ba2dd3a3eb
Bugfix: Containers could lose their snap state (Thanks Atsutane)
...
When being on a different workspace than the one where the snapped
container is, the function to cleanup cols/rows would clean up too
much.
2010-01-01 22:40:50 +01:00
092f3139d8
Remove superfluous #include <assert.h> (Thanks badboy)
2010-01-01 17:30:27 +01:00
65cae2cad1
Bugfix: Don’t put dock clients into floating mode (Thanks xeen)
2010-01-01 16:19:42 +01:00
75aac5bc02
ewmh: implement support for _NET_WORKAREA (rdesktop can use that)
...
Please note that rdesktop’s -g workarea option will not work on
64-bit systems at the moment because of a bug in rdesktop (see the
rdesktop-devel mailing list).
2009-12-31 17:48:41 +01:00
e7e9e8e49d
ewmh: correctly set _NET_ACTIVE_WINDOW
2009-12-25 15:19:39 +01:00
0641e6a1a3
ewmh: correctly set _NET_CURRENT_DESKTOP to the number of the active workspace
2009-12-25 15:05:17 +01:00
ba82a3e63b
Bugfix: Fix NULL pointer dereference in workspaces which have preferred screens but were not used yet (Thanks badboy)
2009-12-23 00:39:03 +01:00
3e53ecf69e
retab! src/container.c
2009-12-22 23:43:05 +01:00
848d9c1b01
Make containers containing exactly one window behave like default containers
...
Starting from this commit, a borderless window will always be
borderless if it is the only window in a container. For example,
you can have Firefox borderless in a tabbed container and as soon
as the download manager or a viewer gets opened, the container
will be rendered like a normal tabbed container.
This solves the user-interface dilemma of borderless/1-px-border
windows inside stacked/tabbed containers, at least for this special
case. Thanks to Merovius for this suggestion.
2009-12-22 23:40:06 +01:00
9df64a8d0a
Correctly exit when another window manager is already running
...
This is implemented by checking if setting the redirect mask returned
an error or not.
2009-12-22 11:29:24 +01:00
e900a8d23d
xinerama: correctly put windows which are assigned to a specific screen on that screen when it becomes available (Thanks badboy)
2009-12-21 22:30:08 +01:00
0b5554c762
xinerama: change some log messages to errors
2009-12-21 20:41:02 +01:00
Michael Stapelberg