| Age | Commit message (Collapse) | Author |
|---|
|
The criteria struct now uses properties for each token type rather than
the list_t list of tokens. The reason for this is that different token
types have different data types: pcre, string and number to name a few.
This solution should be more flexible moving forward. A bonus of this is
that criteria is now easier to understand when looking at the struct
definition.
The criteria parser has been rewritten because the previous one didn't
support valueless pairs (eg. [class="foo" floating]).
Criteria now has types. Types at the moment are CT_COMMAND,
CT_ASSIGN_WORKSPACE and CT_ASSIGN_OUTPUT. i3 uses types as well.
Previously the assign command was creating a criteria with 'move to
workspace <name>' as its command, but this caused the window to appear
briefly on the focused workspace before being moved to the assigned
workspace. It now creates the view directly in the assigned workspace.
Each view will only execute a given criteria once. This is achieved by
storing a list of executed criteria in the view. This is the same
strategy used by i3.
Escaping now works properly. Previously you could do things like
[class="Fire\"fox"] and the stored value would be 'Fire\"fox', but it
should be (and now is) 'Fire"fox'.
The public functions in criteria.c are now all prefixed with criteria_.
Xwayland views now listen to the set_title, set_class and
set_window_type events and criteria will be run when these happen. XDG
shell has none of these events so it continues to update the title in
handle_commit.
Each view type's get_prop function has been split into get_string_prop
and get_int_prop because some properties like the X11 window ID and
window type are numeric.
The following new criteria tokens are now supported:
* id (X11 window ID)
* instance
* tiling
* workspace
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
If the output is rotated, the scissor box needs to be transformed in the
opposite rotation.
|
|
This implements the title_format command, with a new placeholder %shell
which gets substituted with the view type (xwayland, xdg_shell_v6 or
wl_shell).
Example config:
for_window [title=".*"] title_format %title (class=%class instance=%instance shell=%shell)
|
|
This allows the title's texture to always be the full width of the text,
and clipped at render time according to the desired width (eg. tabs...).
As an added bonus, the texture no longer needs to be updated when
containers are arranged.
|
|
|
|
|
|
The alpha component is merged with the container's opacity.
Completes #1882.
|
|
Before freeing sway_output, NULL the wlr_output reference to it. Check for that
NULL in layer_shell handle_destroy. Don't damage null container in unmap.
Additionaly, terminate swaybg if its output is being disabled.
|
|
|
|
|
|
Implements rendering of borders. Title text is still to do.
Implements the following configuration directives:
* client.focused
* client.focused_inactive
* client.unfocused
* client.urgent
* border
* default_border
|
|
|
|
|
|
Replaces arrange_windows() with arrange_root(), arrange_output(),
arrange_workspace() and arrange_children_of().
Also makes fullscreen views save and restore their dimensions, which
allows it to preserve any custom resize and is also a requirement for
floating views once they are implemented.
|
|
|
|
|
|
This makes Chromium and Firefox context menus appear when fullscreen.
|
|
|
|
|
|
|
|
|
|
|
|
It appears client_pending is the state that should be used here.
|
|
* Render background when using fullscreen, because transparency.
* Check that fullscreen surface allows input.
* Don't look for surfaces in top layer if there's a fullscreen view.
|
|
* Render fullscreen views without wlr function, which makes popups and
lockscreen work.
* Don't allow input events to surfaces behind fullscreen views.
* Use correct output dimensions (for rotated outputs).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This fixes dmenu
|
|
Don't send frame done immediately when rendering
|
|
|
|
Seems like it doesn't work this way
|
|
Fine-grained damage tracking
|
|
Implement move [left|right|up|down]
|
|
|
|
|
|
|
|
|
|
|
