This commit introduces the komorebi-client library crate for other Rust
applications to interact with a running instance of komorebi over Unix
Domain Sockets.
Currently the crate re-exports everything one might find in the
komorebi::State struct and everything that is publicly exposed in
komorebi-core.
Public types and methods are still lacking documentation, and this crate
should not be published on crates.io until this is no longer the case.
This commit adds some contribution guidelines and updates the "Window
Manager Event Subscriptions" section with information on using
subscribe-socket and komorebi-client.
This commit includes backwards-compatible renames of
active_window_border_offset and active_window_border_width to
border_offset and border_width respectively.
Since the invisible window border adjustments were removed in #678, the
invisible window borders set by the OS can now also be adjusted via
border_offset and border_width, which is the primary reason for the
rename.
invisible_borders has been marked as deprecated and the previously
deprecated alt_focus_hack option has been removed.
This trades one issue for another, in order of importance:
- Pop-up windows such as a file upload dialog box for Firefox no longer
have a window border drawn over the top - better.
- Opaquely bordered windows without DWM decorations, combined with a -1
offset / single pixel border end up as invisible borders (e.g. EPIC
Games Launcher).
The default window corner rounding is still slightly visible at offset
-1 until this corner radius that completely closes up the transparent
region, without needing to invasively draw over the target window.
topmost has a special meaning, in that it is a sticky raise, which is
not what we want - we don't want to be permanently above all other
windows, as this leads to bugs where for example the windows can end up
stuck above non-topmost windows in a different security context, such as
programs running with administrator priviliges.
In the new border painting strategy, the 20 pixel border is huge. The
border is now offset -1 by default, so as to draw over the standard DWM
1px border (avoiding a "1 pixel see through gap"), and the default width
is 8px.
The SetWindowPos API will inset the provided dimensions by the amount of
space required for window decorations (shadows, etc). Compute the size
of the decorations and add them as padding to the provided size,
resulting in windows being set precisely to the target dimensions.
An active_window_border_width=1, active_window_border_offset=-1 will now
paint over the 1px window decoration consistently.
The border window is always made topmost on resize, so that it paints
over custom borders (e.g. EPIC Games Launcher) as those borders are
opaque and as such a 1px border configuration as above becomes invisible
in that condition.
The layout should leave the space configured for the border, so that the
border always stays within the workspace bounds.
Border offset is cleaned up, as it is no longer a rect, but instead just
a fixed value.
The rect function for adjusting padding now takes a concrete value, as
the optional has no local meaning to the operation, being equivalent to
the default value.
A margin function is added to centralize the notion of increasing the
size of a rect by the given margin, the opposite of the padding
operation.
This fixes a regression from an earlier commit that dropped the DWM
style borders without fixing corner rounding on Windows 11. This is now
fixed for Windows 11 again, while avoiding the extra system border
decorations.
This makes the borders pixel-perfect, and border_overflow can be
disabled on all applications.
Unfortunately this also means we lose the corner rounding, so that may
need to be done differently or the offsets refined in some way to
address that.
Switch to using the DWM API to get Window bounds so as to exclude the
outside of window decorations from the computation.
This is getting close to a precise window size, you can now set an
active border width of 1 and an offset of 1 and get a 1 pixel line
around most windows, except that there's some extra top padding I have
yet to find the cause of.
This implementation needs to be DPI aware, but I haven't yet tested if
the DPI scaling approach is entirely valid - we may instead need to get
the per-monitor DPI scale, identify the monitor the window is on, and
scale to that, rather than using the system wide scale.
Maybe fixes#574
Maybe updates #622
This commit adds the "komorebic whkdrc" command to go along with the
"komorebic configuration" command introduced in
608ec03047, aimed it making it easier to
edit this file using a command line editor.
The "config" command has been renamed in the code to Configuration, with
an alias of "config". The Whkdrc command similarly comes with the "whkd"
alias.
This commit introduces the ability for users to specify colours in the
static config file in either RGB or Hex format.
This is done by introducing a new enum, Colour, which provides variants
wrapping the internal Rgb type, and the HexColour type from the
hex_colour crate. HexColour itself is wrapped in a new struct, Hex, in
order to allow the implementation of the JsonSchema trait.
A number of From<T> implementations have been done in order to make
working with the Colour enum more ergonomic.
Ultimately, the core representation for colours in komorebi remains the
Rgb struct, any and Hex-specified colours will be converted to this Rgb
type before being converted to u32 to pass them on to various Win32 API
calls.
This commit adds the "komorebic config" command as a helper to print the
path to komorebi.json, while respecting the KOMOREBI_CONFIG_HOME
environment variable if it exists.
This is particularly useful for users who wish to edit this file on the
command line, as they can now run commands like:
"lvim $(komorebic config)"
This commit addresses and number of bugs and improves the experience of
working with floating workspaces (ie. Workspace.tile = false).
- When the user moves or resizes a window on a floating workspace,
WindowManagerEvent::MoveResizeStart will no longer trigger, which
prevents the mouse focus from going to the middle of the window rect
after the resize or move action (if mouse_follows_focus = true)
- If active_window_border = true, it will no longer be shown on focused
windows in floating workspaces
- When windows are moved using a komorebic command from a floating
workspace to a tiling workspace and active_window_border = true, the
active window border will be shown again
This commit remove the socket connection retry loop in send_message
which is no longer required after @raggi's changes in
c8f6502b02.
@azinsharaf noticed that when trying to run komorebic commands while
komorebi was not accepting connections, multiple hanging komorebic
instances could be spawned, particularly if commands were retried.
@eythaann proposed an additive fix for this in PR684 but ultimately as
the previous race condition with the query/response commands has been
handed by @raggi we can remove the socket connection retry loop
completely.
This commit addresses a few bugs with regards to maximized window
handling.
- Correctly restoring and focusing when switching to a workspace
containing a window previously maximized with the toggle-maximize
command
- Unmaximizing a window during the initial window scan when the wm
initializes so that windows that are maximized before running komorebi
will no longer have the ugly white bar at the top
- When updating workspace layouts, windows that have been maximized
without using the komorebic toggle-maximize command will be
unmaximized to prevent the wm state drifting out of sync with what is
happening on the screen
This commit ensures that the focus changes to the appropriate window
when a container stack is being cycled through by a user who has
disabled the mouse_follows_focus feature.
fix#680
- Use a single thread to bind the hook, and then start dispatching.
- Use a blocking loop for message dispatching.
- Remove the locks around crossbeam channel, as it's already Send + Sync
This commit splits the komorebi crate into a mixed binary and library
crate.
All types and logic related to window management have been moved under
lib.rs, and are imported from there for use in main.rs, which is now
only responsible for starting and running the window manager process.
In preparation for exposing a new komorebi-client crate in the future,
serde::Deserialize has been derived on Notification and any struct that
may appear in a notification receievd by a process that has subscribed
to event updates.
re #659
This commit adds support for subscriptions via Unix Domain Sockets which
are better suited for IPC between Rust processes compared to Named Pipes
which have issues that I don't want to spend time resolving.
The main motivation for this change is to provide an easy way for the
new zebar project to consume information about komorebi's state in the
Rust backend so that a bar module can be created for komorebi users.
The next step in this process will be to finally refactor the komorebi
crate into a mixed bin/lib crate, and expose all notification-related
structs and maybe some connection helper methods in a new
komorebi-client crate.
The previous "subscribe" and "unsubscribe" komorebic commands have had
the "-pipe" suffix added to them, with aliases in place for the previous
names in order to ensure backwards compat.
Replace the client socket with replies sent on the other side of the
querying stream, for both UDS and TCP clients. This has two results:
unix socket clients such as komorebic no longer race on the socket bind,
fixing the out of order bind error, and the response mixup conditions
that could occur. Queries over TCP now receive replies over TCP, rather
than replies being sent to a future or in-flight command line client.
This commit ensures that the $Env:LOCALAPPDATA/komorebi dir is created
by the quickstart command, as it cannot be assumed that this will always
exist, especially on new machines with recent versions of Windows 11.
fix#671
This commit is a fix that handles a subtle breaking change in
sysinfo::Process:root() which no longer can be used to see if a process
is a scoop shim.
Instead we can stringify the executable path and see if the absolute exe
path contains the substring "shims".
With this fix duplicate process detection is once again working
correctly.
