Commit message (Collapse)AuthorAge
* Fix click behaviour1.0v1.0Libravatar Ryan Dwyer2019-03-11
| | | | | | | | | By the time seatop_allows_events was called, seatop_impl was already NULL, causing the function to always return false. This means a press event was sent to clients without a corresponding release event. This patch moves the call to seatop_finish to after the seatop_allows_events check.
* Don't send button events to surfaces when dragging or resizingLibravatar Ryan Dwyer2019-03-11
| | | | | | | | | | | It turns out sending button events during all seat operations is not desirable. This patch introduces a new property `seatop_impl.allows_events` which allows each operation to define whether button events should be passed to the surface or not. The `down` seat operation is the only one that supports this. As all the other seatops don't support it, the calls to seat_pointer_notify_button prior to starting them have been removed.
* stringop.c: refactor a few functionsLibravatar Ian Fan2019-03-11
* stringop.c: clean up headersLibravatar Ian Fan2019-03-11
* stringop.c: remove unused functionsLibravatar Ian Fan2019-03-11
| | | | The only use of `join_list` in swaybar/tray/icon.c has been rewritten.
* sway_view_child: add listener for view unmapLibravatar Brian Ashworth2019-03-11
| | | | | | | | | | Since not all child views's have an unmap event, it is possible for it to still be mapped (default state) in the destruction handler. When the destruction handler is called, the corresponding view may have already been freed and the memory location reallocated. This adds a listener for the view unmapping and removes the mapped status. This ensures that the child view is damaged due to destruction while the view still exists and not after.
* detect_proprietary: use strncmpLibravatar Brian Ashworth2019-03-11
| | | | | Only the main nvidia module needs to be blocked. Others such as nvidiafb are benign and do not need to be blocked
* fullscreen: init floating on disable without sizeLibravatar Brian Ashworth2019-03-11
| | | | | | | If a container gets mapped as fullscreen and set to floating by criteria, the size and location are never set for the floating container. This adds a check in container_fullscreen_disable for a width or height of 0 and calls container_init_floating
* meson: use pkg-config var for scdoc pathLibravatar emersion2019-03-11
* Fix crash in cmd_workspace when layer surface has focusLibravatar emersion2019-03-11
* Update language in sway.desktop & sway(1)Libravatar Drew DeVault2019-03-11
* arrange: use int not size_t for title offsetsLibravatar Brian Ashworth2019-03-11
| | | | | | | | | This changes `apply_tabbed_layout` and `apply_stacked_layout` to use `int` instead of `size_t`. This is necessary for tabbed and stacked containers to be positioned correctly when the y-location is negative. The reasoning for this is signed plus unsigned is always an unsigned value. This was causing the y-location of the container to be positioned near `INT_MIN` due to an unsigned integer underflow
* damage: remove output_damage_viewLibravatar Brian Ashworth2019-03-11
| | | | | | | This removes `output_damage_view` since it is unnecessary. The logic has been moved into its only caller `output_damage_from_view`. When damaging the whole view, `output_damage_whole_container` should be used instead
* output_damage_whole_container: damage subsurfacesLibravatar Brian Ashworth2019-03-11
| | | | | | | | This adds an iterative call in `output_damage_whole_container` to damage the subsurfaces for all visible views that are inside of the container. This is needed to damage subsurfaces that extend outside the box of the container. Without this, those subsurfaces will create artifacts when moving or resizing.
* fix "directive argument is null" errorsLibravatar Jeff Peeler2019-03-11
* Fix crash when moving window to scratchpadLibravatar minus2019-03-11
* Add missing swaymsg completionsLibravatar Peter Grayson2019-03-11
| | | | | | | | | The `-m/--monitor` option was missing from the bash and fish completions. The `subscribe` IPC message type was missing from the bash, fish, and zsh completions. Signed-off-by: Peter Grayson <pete@jpgrayson.net>
* Add -p/--pretty option to swaymsgLibravatar Peter Grayson2019-03-11
| | | | | | | | | | | This new option forces pretty (non-raw/non-JSON) output. By default, when not using a tty, swaymsg outputs using the "raw" format. This makes it impossible to, for example, pipe the pretty output to a pager such as `less` since piping does not use a tty. The new -p/--pretty option gives the user explicit control over the output format while retaining the default tty-dependent behavior. Signed-off-by: Peter Grayson <pete@jpgrayson.net>
* ipc: describe libinput device configurationLibravatar Brian Ashworth2019-03-11
| | | | | | | | This adds the device configurations to the ipc response for libinput devices. Only supported configuration options for the device will be added. This also moves `libinput_send_events` inside a new `libinput` object that contains the rest of the configuration options. sway-ipc(7) has been updated to reflect the changes and document the new additions.
* ipc: fix fullscreen deco_rectLibravatar Brian Ashworth2019-03-11
| | | | | | This fixes the deco_rect reported by the ipc for fullscreen containers to be all zeroes. Children of the fullscreen container should still have their decorations reported correctly
* ipc: fix rect for stacked childrenLibravatar Brian Ashworth2019-03-11
| | | | | This now takes all titlebars for stacked children into account for the ipc property `rect`
* Fix container_parent_layout for scratchpad windowsLibravatar emersion2019-03-11
* ipc: change {,deco_}rect to match i3Libravatar Brian Ashworth2019-03-11
| | | | | | | | | | | | This fixes the `deco_rect` and `rect` properties in the IPC responses to match i3's behavior. `deco_rect` should be relative to the parent node, not the current node. This also takes tabbed and stacked decorations into account and will calculate `deco_rect` for all containers since tabbed and stacked child containers will have decorations. `rect` should exclude the window decorations.
* Set DISPLAY after initializing XwaylandLibravatar emersion2019-03-11
| | | | This is necessary after https://github.com/swaywm/wlroots/pull/1596
* Make raw keysyms take precedence over translatedLibravatar Ben Challenor2019-03-10
| | | | | Allows both BackSpace and Shift+BackSpace to be bound under the US keyboard layout, per #3705.
* meson: update scdoc requirement to >= Brian Ashworth2019-03-04
| | | | | | | | Since scdoc 1.9.1 is bugged, this updates the meson version check to >= 1.9.2 and drops the version requirement from the README. This should make it more obvious to users who have 1.9.1 that they need to update scdoc to be able to compile man pages and hopefully cut down on the duplicate issues
* Allow concurrent clicksLibravatar Ryan Dwyer2019-03-04
| | | | | | | | | | | | | | | | | | | | | If two cursor buttons are pressed at the same time, the client will now be notified of the second button press. The main reason for not sending the concurrent presses was due to an early return in dispatch_cursor_button if a seatop is in progress. This patch makes it call seat_pointer_notify_button prior to returning. But it also has to make sure there's not a mismatch in events such as a release without a press. Prior to this patch, the down seatop would send press and release events in its begin and finish functions. No other seatops did this. A press event would be sent prior to starting tiling drag, but never an associated release. After this patch, no seatops send their own press or release events. We send them prior to calling the seatop begin functions, then the first part of dispatch_cursor_button handles all presses during seatops and when releasing the seatop.
* sway.1.scd: document environment vars set by swayLibravatar Brian Ashworth2019-03-04
| | | | | This just documents the few environment variables set by sway in sway.1.scd
* render_floating: skip fullscreen floatersLibravatar Brian Ashworth2019-03-04
| | | | | | | | If a floater is fullscreen either on a workspace or globally, it should not be rendered on any output is is not fullscreened on. When rendering it on an output it should not be rendered on, there will be an extraneous border along the adjacent side of the output. This adds a check in render_floating to skip all fullscreened floaters
* Fixes crash in spawn_swaybg (closes #3733)Libravatar Noam Preil2019-03-04
* floating_maximum_size: change default behaviorLibravatar Brian Ashworth2019-03-04
| | | | | | | | | | | | | This changes the way zero (which is the default) is interpreted for both the width and height of `floating_maximum_size`. It now refers to the width and height of the entire output layout, which matches i3's behavior. This also removes duplicated code to calculate the floating constraints in three files. Before this, `container_init_floating` used two-thirds of the workspace width/height as the max and the entire workspace width/height was used everywhere else. Now, all callers use a single function `floating_calculate_constraints`.
* Minor fix of code duplication.Libravatar hugbubby2019-03-04
| | | | Removes 3~ lines of code that didn't need to be restated.
* meson: check scdoc versionLibravatar emersion2019-03-04
* seat: don't send button release when not pressedLibravatar emersion2019-03-04
| | | | | | | | | All seat operations except "down" eat the button pressed event and don't send it to clients. Thus, when ending such seat operations we shouldn't send the button released event. This commit moves the logic used to send pressed/released into the "down" operation.
* tray: fix pixmap colorsLibravatar Ian Fan2019-03-04
| | | | by converting from network byte order to host byte order
* sway-ipc.7: clarify window_rect omits decorationsLibravatar Brian Ashworth2019-03-04
| | | | | | According to the i3 ipc documentation, `window_rect` excludes the window decorations from the calculation. This just clarifies that in `sway-ipc.7.scd`
* Add output dpms to manpageLibravatar emersion2019-03-04
* Print Meson featuresLibravatar emersion2019-03-04
* Set minimum wlroots versionLibravatar emersion2019-03-04
* Add sway-ipc.7.scd to document IPC protocolLibravatar Brian Ashworth2019-03-04
| | | | | | | This add `sway-ipc.7.scd` that documents the IPC protocol. This also increased the minimum scdoc version from 1.8.1 to 1.9.0 to allow for table cells to be continued on the following line
* config: remove double assignement to result in get_output_configLibravatar Rouven Czerwinski2019-03-04
* cursor: intitialize sx and sy to zeroLibravatar Rouven Czerwinski2019-03-04
| | | | | If node_at_coords does an early return without setting these values, they can be used uninitialized later. Initialize both to zero.
* cursor: remove unused node assignementLibravatar Rouven Czerwinski2019-03-04
| | | | | The node variable is not used before its reassigned later in the function, remove the assignement.
* execute_command: do not strip qoutes for cmd_modeLibravatar Brian Ashworth2019-03-04
| | | | | | | | | `cmd_mode` performs its own quote stripping for the mode string to avoid double stripping quotes for `cmd_bindcode` and `cmd_bindsym` in `config_command` and `execute_command`. Stripping quotes in `execute_command` for `cmd_mode` will also result in double stripping, which will cause issues for any mode string with spaces, such as pango markup.
* fix smart_borders description in manual1.0-rc4Libravatar Carlo Abelli2019-02-25
| | | | The wording for smart_borders was opposite the actual behavior.
* Make load_include_configs void. Fix some cases where WD would not be restored.Libravatar Connor E2019-02-25
* Fix crash exiting fullscreened floating containerLibravatar minus2019-02-25
| | | | | | | | container_floating_move_to_center and container_fullscreen_disable were calling recursively when the container spawned as a fullscreen floating container (via for_window). Such a window now doesn't crash sway anymore but is still configured with a wrong, zero size, making it not directly usable.
* ipc: add missing fields to disabled outputsLibravatar Caleb Bassi2019-02-25
| | | | | | i3 requires all outputs to have certain fields, including 'primary', 'current_workspace', and 'rect' which were missing on disabled outputs. https://i3wm.org/docs/ipc.html#_outputs_reply
* Update for swaywm/wlroots#1517Libravatar emersion2019-02-25
* handle_seat_node_destroy: update seat->workspaceLibravatar Brian Ashworth2019-02-25
| | | | | | | If an unmanaged or layer surface is focused when an output gets disabled and an empty workspace on the output was focused by the seat, the seat needs to refocus it's focus inactive to update the value of `seat->workspace`.