Version Information:

• Distribution Information ( run uname -a ): Linux FrameWork 5.16.11-arch1-1 #1 SMP PREEMPT Thu, 24 Feb 2022 02:18:20 +0000 x86_64 GNU/Linux
• swhkd version ( swhkd -V ) swhkd 1.0.0

Describe the bug: I installed from the AUR (swhkd-git), when I try to run start the daemon I get the following error msg:

$ swhks &
$ pkexec swhkd -c ~/.config/swhkd/swhkdrc
[2022-03-02T00:49:56Z WARN  swhkd] Running swhkd as root!
Error: Os { code: 38, kind: Unsupported, message: "Function not implemented" }

Expected behavior: The daemon to start succefully.

Actual behavior: Daemon doesn't start and crashes.

To Reproduce:

$ paru -S swhkd
$ swhks &
$ pkexec swhkd -c ~/.config/swhkd/swhkdrc

I also tried installing straight from the github:

$ git clone https://github.com/waycrate/swhkd
$ cd swhkd
$ git rev-parse HEAD
3986dd01a31c168e91a3de038911745293dfa574
$ make setup
$ make glibc
$ sudo make install

But the same crash happens.

Ignored events (i.e. events other than key presses for now) are re-emitted to the uinput devices. This allows grabbing mouses and other pointer devices while still allowing the cursors to work.

In some cases, events are not re-emitted at all when the uinput device handles keys, relative axes and switches. By sending switch events to a separate uinput device just for switches and another one for keys and relative axes, we can re-emit keys, pointer events, and switches events properly.

Version Information:

• Distribution Information ( run uname -a )

Debian 5.18.16

• swhkd version ( swhkd -V )

swhkd 1.2.1

Describe the bug: pkexec swhkd turn off the wifi on start. I have to enable it manually after.

Note

• It turns off the wifi (not disconnect)
• Reload config is fine.

Expected behavior: Don't turn off wifi

Actual behavior:

❯ pkexec swhkd -d -c ~/.config/swhkd/swhkdrc
[2022-11-28T13:34:22Z TRACE swhkd] Logger initialized.
[2022-11-28T13:34:22Z TRACE swhkd] Invoking UID: 85819
[2022-11-28T13:34:22Z TRACE swhkd] Setting process umask.
[2022-11-28T13:34:22Z ERROR swhkd] XDG_RUNTIME_DIR has not been set.
[2022-11-28T13:34:22Z TRACE swhkd] Reading /run/swhkd/swhkd_85819.pid file and checking for running instances.
[2022-11-28T13:34:22Z DEBUG swhkd] Previous PID: 1559337
[2022-11-28T13:34:23Z WARN  swhkd] Running swhkd as root!
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting initgroups...
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting EGID...
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting EUID...
[2022-11-28T13:34:23Z DEBUG swhkd] Using config file path: "/home/timgreen/.config/swhkd/swhkdrc"
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting EGID...
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting EUID...
[2022-11-28T13:34:23Z DEBUG swhkd::perms] Setting initgroups...
[2022-11-28T13:34:23Z TRACE swhkd] Attempting to find all keyboard file descriptors.
[2022-11-28T13:34:23Z TRACE swhkd] Other: sof-hda-dsp HDMI/DP,pcm=5
[2022-11-28T13:34:23Z TRACE swhkd] Other: sof-hda-dsp HDMI/DP,pcm=4
[2022-11-28T13:34:23Z TRACE swhkd] Other: sof-hda-dsp HDMI/DP,pcm=3
[2022-11-28T13:34:23Z TRACE swhkd] Other: sof-hda-dsp Headphone
[2022-11-28T13:34:23Z TRACE swhkd] Other: sof-hda-dsp Mic
[2022-11-28T13:34:23Z TRACE swhkd] Other: Integrated RGB Camera: Integrat
[2022-11-28T13:34:23Z TRACE swhkd] Other: ELAN067A:00 04F3:3197 Touchpad
[2022-11-28T13:34:23Z TRACE swhkd] Other: ELAN067A:00 04F3:3197 Mouse
[2022-11-28T13:34:23Z TRACE swhkd] Other: Integrated RGB Camera: Integrat
[2022-11-28T13:34:23Z TRACE swhkd] Other: ThinkPad Extra Buttons
[2022-11-28T13:34:23Z TRACE swhkd] Other: Wacom Pen and multitouch sensor Finger
[2022-11-28T13:34:23Z TRACE swhkd] Other: Wacom Pen and multitouch sensor Pen
[2022-11-28T13:34:23Z TRACE swhkd] Other: TPPS/2 Elan TrackPoint
[2022-11-28T13:34:23Z TRACE swhkd] Other: Video Bus
[2022-11-28T13:34:23Z TRACE swhkd] Other: Power Button
[2022-11-28T13:34:23Z TRACE swhkd] Other: Lid Switch
[2022-11-28T13:34:24Z TRACE swhkd] Other: PC Speaker
[2022-11-28T13:34:24Z TRACE swhkd] Other: Sleep Button
[2022-11-28T13:34:24Z DEBUG swhkd] Keyboard: AT Translated Set 2 keyboard
[2022-11-28T13:34:24Z DEBUG swhkd] 1 Keyboard device(s) detected.
[2022-11-28T13:34:24Z ERROR swhkd] XDG_RUNTIME_DIR has not been set.
[2022-11-28T13:34:37Z DEBUG swhkd] state_modifiers: {

To Reproduce: Just run pkexec swhkd

Additional information: Anything else you'd like us to know ?

Happy to grab other logs to happy debugging.

Example:

# /home/waycrate/.config/swhkd/swhkdrc
super + enter
    alarcritty
super + shift + w
    firefox

# /home/waycrate/.dwm/swhkdrc
use /home/waycrate/.config/swhkd/swhkdrc
super + q
    dwmc quit

As you can see in the code, we can anyone of use import include source as the token

Todo:

• [x] error checks for each config file imported and report error with the file path
• [ ] do some clean up
• [ ] write test functions for testing a config file (read the file and compare the result)
• [ ] support relative path and ~/ in path

Describe the bug:

Can see here a cargo.toml file in the root folder of this repo / the main project repo. However cannot seem to just cargo install ... the software like other rust packages. So is this something planned for the future?

Have tried the following on the cmdline:

⏚ θ127° [id:~/.bin] $ cargo install Swhkd
    Updating crates.io index
error: could not find `Swhkd` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhkd
    Updating crates.io index
error: could not find `swhkd` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhkd-git
    Updating crates.io index
error: could not find `swhkd-git` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhkd-musl-git
    Updating crates.io index
error: could not find `swhkd-musl-git` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhkd-bin
    Updating crates.io index
error: could not find `swhkd-bin` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhkd-musl-bin
    Updating crates.io index
error: could not find `swhkd-musl-bin` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ cargo install swhks
    Updating crates.io index
error: could not find `swhks` in registry `crates-io` with version `*`
⏚ θ127° [id:~/.bin] 101 $ 

Additional information: Anything else you'd like us to know ?

Version Information:

• linux 5.16.11-arch1-2
• swhkd 1.1.0

Describe the bug: sxhkd has the modal feature, e.g. (example from https://github.com/baskerville/sxhkd/blob/master/README.md )

super + o ; {e,w,m}
	{gvim,firefox,thunderbird}

swhkd doesn't seem to recognize the semicolor ;. Is this implemented yet?

Expected behavior: Drop in replacement for sxhkd.

Actual behavior: semicolon ; causes error.

To Reproduce: same as above

super + o ; {e,w,m}
	{gvim,firefox,thunderbird}

Additional information: Great project! Thank you!

From what I've seen, std::fs::File::open() doesn't take into account the UID when trying to detect files. To prevent detection of root-owned files in root directories, I've resorted to using the system /bin/cat command for now to open files. It's hacky, but it seems to work.

Version Information:

• Distribution Information ( run uname -a ) Linux xircon-w6567sz 5.16.16-zen1-1-zen #1 ZEN SMP PREEMPT Mon, 21 Mar 2022 22:59:42 +0000 x86_64 GNU/Linux
• swhkd version ( swhkd -V ) swhkd 1.1.7

Describe the bug: Chords not working

super + n:  {n,p,d,x}
    {variety -n, variety -p, variety -t, variety --pause && notify-send Paused} 

[2022-03-28T13:43:13Z ERROR swhkd] Config Error: Error parsing config file "/etc/swhkd/swhkdrc". Unknown symbol at line 11.

Expected behavior: Should work as it does in sxhkd.

Actual behavior: See error swhkd does not load.

We're currently missing a changelog file. Having a changelog file directly inside the repository would help track changes. This would also make it easier to create new releases as we could just copy and paste what is already available in the changelog file.

I would suggest using something like Keep a Changelog for this. We can add this file with the latest changes only for now, and try to add previous releases to this file gradually later on.

We could also enforce updating the changelog on PRs to ensure we do have everything in there.

This issue is mainly to discuss this and to think about possible solutions.

what i meant is that with sxhkd the keybindings conflict and sxhkd is given the preference (i think that's how you spell it) so if i press alt + a when i'm focused on pcmanfm the keybinding will work on sxhkd but not pcmanfm

Discord message

Somehow we should provide an option to swallow keys so that they don't go to a focused application / the tty.

As a fix for CVE-2022-27814, root privileges are dropped to the calling user when (re)loading the config file. Privileges were sometimes dropped but never raised again, which caused crashes when sending SIGHUP to swhkd multiple times in a row.

This now always raises privileges after successfully reading the config file. Fixes #155.

To help plan features and work on long-term goals for the project, we could use GitHub milestones.

This would be particularly helpful for deprecation or for big features that could then be split into multiple smaller issues to distribute work more easily.

For now, we could create a 1.3.0 milestone and add some features we would like to implement. Fixes could just be added as needed and as implemented.

This would help in bringing features to stable releases in a timely manner (see #194).

There is currently no formal process for releasing a stable version of swhkd.

For 1.2.2, I created a separate branch, split from a specific older commit, on which I updated the Cargo.toml and Cargo.lock files. As the 1.2.2 version is a patch release (semver), this made sense to not include larger changes, such as new features.

This way of doing releases could be extended to all new swhkd releases.

Process proposal:

1. Branch off from a commit -- either from the main branch, or from an already existing release branch
2. (Optional) Cherry-pick commits from other branches, or develop fixes/features as needed
3. Update changelogs in the main branch and in the new branch
4. Update the Cargo.toml and Cargo.lock files in the new branch
5. Make sure tests pass
6. Tag the Cargo update commit with the release version
7. Create a release on GitHub

Ideally, 2, 3 and 4 would be bundled in a PR, which would ensure 5 gets verified.

Pros:

• Easy to cherry-pick fixes, security patches, and other updates onto the stable release branches
• Less version bumps commits on the main branch
• Clear history showing where releases diverge from the main branch
• Probably simpler to package, which may increase adoption (especially for non-rolling distros such as Debian, Ubuntu, Suse, etc)

Cons:

• This implicitly assumes that we support older releases, which may increase maintenance in the long run
  • This would in turn means we need to discuss how long we should support these older releases
  • And also, what would "support" actually mean for us?
• (too) many branches for each stable release, which will clutter the GitHub UI (personally, I think it is fine development-wise to have many branches)
• Divergence from the main branch

If we go forward with this process (or another), it could be documented either directly inside the repository (PROCESSES.md?) or the GitHub wiki.

As always, this proposal is debatable, feel free to chip in!

Currently, git commits tend to follow the [subject] title style. Unfortunately, there is no strict guide for this format as the link given in the CONTRIBUTING.md file is not up to date (https://github.com/facebook/draft-js/blob/a9316a723f9e918afde44dea68b5f9f39b7d9b00/CONTRIBUTING.md links to a 404). In addition, this format is not followed closely by contributors or maintainers: see 04e5125, e4a490 and 321c58c for three different commit styles.

This was mentioned briefly in #181 (comment https://github.com/waycrate/swhkd/issues/181#issuecomment-1357503117) where we discussed adopting Conventional Commits with automation help from the cocogitto project.

This issue is for discussing adopting Conventional Commits or alternatives.

We should also discuss if/how to enforce it. Should we add git hooks? GitHub actions checks? Manually, as part of the PR review process?

This is a continuation of a discussion from #71.

First, a small summary of how device grabbing was implemented in swhkd as I understand it:

To avoid grabbing all devices and swallowing their events (mouse cursor movements, for instance), a keyboard detection algorithm was first implemented in 8bbd34c. It checks if a given device is a keyboard by checking for the presence of the Enter key. This was not completely correct, as some users still had their non-keyboard devices grabbed by swhkd (see #131) and this was causing other issues. For some others, some hardware keys were not recognized as keyboards (see #71).

In answer to these issues, a --device argument was added to swhkd in b53b8c. This allows users to explicitely tell what devices to use. With some configuration from users, this thus solves the issues.

Ideally, users should not have to worry about what device swhkd should grab - it should do it correctly out of the box.

In #163, we added support for re-emitting most device events from the grabbed device to the virtual uinput device of swhkd.

With this improvement, it should not be necessary anymore to avoid grabbing all devices, as swhkd can just re-emit events as needed. Additionally, the --device argument is probably unnecessary and could be deprecated.

Cargo.lock was updated using make release.

Please note that this PR is to be merged into the 1.2.2 branch. There will be a bit of divergence compared to the current main branch, but this allows us to make a proper fix release.

1.2.2 contains a number of fixes. Below is a tentative release text for this version based in part on keep a changelog as discussed in #181:

1.2.2

Added

• swhks debug command line option to get verbose logging
• swhks log command line option to log to a file
• swhkd-vim installs alongside swhkd in the AUR package

Changed

• swhks no longer defaults to a verbose output (see debug flag)
• swhks logs to a file by default
• swhks now logs command output by default
• man pages are now gzipped as part of the build process

Fixed

• Incorrect short version flag documentation in swhkd man page.
• Release key events are no longer shadowed by press events (#159)
• Sending multiple SIGHUP to swhkd to reload the config file no longer crashes (#155)

I am not 100% sure that the changelog above is accurate, feel free to discuss it. Notably, I am not sure what 2832de8 is about. @EdenQwQ would you mind explaining it a bit for the changelog? Additionally, a2cdcee is quite large, and I am not sure if I understood all changes.