Sometimes services can fail. Failed services will generally not be
restarted by systemd. To start previously failed services we can
just reset their failed state before starting our session target.
GNOME and Plasma do the same thing.
See: https://github.com/alebastr/sway-systemd/pull/11
"lib.cartesianProductOfSets is a deprecated alias of lib.cartesianProduct."
Rename happened in nixpkgs commit `228621e42dc43f936b66e0ed042c90c511aa0535`.
To avoid this kind of message:
swayidle[181123]: - [Line 1054] Unable to connect to the compositor. If your compositor is running, check or set the WAYLAND_DISPLAY environment variable.
The 24.05 update for KDE Connect moved the kdeconnectd binary from
`/libexec` to `/bin`, so this fix will check the version of the
package used and set the path accordingly.
`splitString "" "some string"` would throw exception `error: invalid
regular expression` on Darwin (or accurately, on Nix built against
libc++). Refer to https://github.com/NixOS/nix/issues/7208 for
details.
Before Nix handles that issue correctly, we should refrain from using
`splitString ""`, and luckliy `lib.stringAsChars` can do the same thing.
Using the option
wayland.windowManager.hyprland.systemd.enableXdgAutostart
users can now choose to start applications present in
`$XDG_CONFIG_HOME/autostart` when starting their sway session.
See 0144ac418e and
https://github.com/nix-community/home-manager/pull/3747
The PR #5299[0] actually adds the requirement on "tray.target" for the
kdeconnect service, when kdeconnect-indicator is set to true. What it
should do is add the requirement directly to the latter. This commit
fixes it.
[0]: https://github.com/nix-community/home-manager/pull/5299#issue-2252834123
Helps avoid successful build but Sway failing to start.
To meaningfully test this, I had to actually use `pkgs.sway` in the
test rather than the stub, but left all other tests using the stub and
changed them to skipping the test.
We need to pass `--unsupported-gpu` as Sway checks for `nvidia` in
`/proc/modules`, and the Nix sandbox has `/proc/modules` available.
Adds a module to enable managing Remmina, an RDP client, with a Home
Manager module, providing a systemd service and mimetype integration
that can be disabled if so desired.
The systemd.time documentation defines the shorthands `daily` and
`minutely` which are currently not included in the nix-gc module.
This commit adds the `daily` option, but omits `minutely` since it's not
a timescale that would make sense to run a gc for.
https://www.freedesktop.org/software/systemd/man/latest/systemd.time.html
The `exec` command does not do any shell parsing and does not
understand the `&&` which is how the extraCommands are added after
dbus activation. There doesn't seem to be a reason for `exec` here
anyway so just remove it and allow shell parsing.
* imapnotify: enable STARTTLS if enabled in email account config
Since version 2.3.10 goimapnotify supports starttls. In version 2.3.11 a
typo in the settings was fixed, using tlsOptions.starttls to enable it.
This commit enables starttls in the goimapnotify config file if it is
enabled in the email account's imap settings.
* imapnotify: test generated config file
Compare the generated service config file with a manually verified
sample.
workspace 10 is part of upstream's default config, but was missing in home-manager.
The initial "sway: add module" PR (02d6040003) went through multiple iterations and had workspace 10 included for a brief moment.
Until the author removed it in a force-push commenting
> Have removed the last change which added bound ${modifer}+0 to workspace number 10 as this messed up workspace numbering in sway.
The reason might have been, that sway used to sort the workspaces in the order they appeared in the config.
Attribute sets in nix are sorted, but not "naturally sorted", meaning `bindsym Mod1+0 workspace number 10` comes before `bindsym Mod1+0 workspace number 1`.
It's unclear if that's what really happened. A workaround would have been to use `lib.lists.naturalSort` in `keybindingsStr`.
But I cannot reproduce this anymore in any way.
I assume this has been fixed many years ago by now.
upstream config: 020a572ed6/config.in (L113-L134)
This removes the services.keepassx module and all options related to it.
This is because KeePassX has been unmaintaned for a long time, and the
migration path to KeePassXC is not clear.
Note, this package probably hasn't worked for nearly a year since the
package was removed from Nixpkgs:
https://github.com/NixOS/nixpkgs/pull/216748Closes#4949
Signed-off-by: Sumner Evans <me@sumnerevans.com>
The previous syntax (`plugin:name { ...settings}`) is not working with
more than one plugin. There is no documentation for this, just the
source code [1].
With this update the plugins paths aren't generated together with the
full config (so the "plugin" field is not "important" anymore) and the
plugins settings are generated like the other fields.
[1] 4d403dac32/src/config/ConfigManager.cpp (L1574)
The plugin setting in the Hyprland config is used both for defining
plugin paths and configuring the plugins. This fix removes the
silent override of the plugins settings converting them to the
`plugin:<name> { ...settings }` syntax.
DBus activated services such as mako use the XCURSOR_THEME and
XCURSOR_SIZE environment variables to decide how to show the cursor,
so without these, the cursor may not match the rest of the desktop
when hovering over (in this example) mako notification surfaces.
Osmscout-server includes a setting in its UI to create a systemd user
service and socket to run the server on demand. This does not function
correctly on NixOS, for two reasons:
1. It assumes that the binary path is stable (e.g.
/usr/bin/osmscout-server), which is not the case on NixOS.
2. It auto-detects the unwrapped binary path, which doesn't work.
This module allows the user to access the same functionality on NixOS.
The service was never marked with a failed state with the previous
approach, which could lead broken synchronisation pair states to go
undetected.
The module now uses a timer instead of unlimited restarts, which does
not have this issue.
When using the previous approach I've always gotten errors that I can't
reload config on the .lock file that exists in /tmp when you run a
standard configured hyprland.
This commit improves this by using hyprctl to find instances to reload
instead.
We can remove the HYPRLAND_INSTANCE_SIGNATURE bogus assignment once
https://github.com/hyprwm/Hyprland/issues/4088 is resolved.
Co-authored-by: Carl Hjerpe <git@hjerpe.xyz>
Add the option sourceFirst to the hyprland module. When this option is
enabled source entries will be put near the top of the file, so that
the variables declared in other files can be used by the other
configuration entries.
Add "source" to the list of important prefixes when the former option
is enabled.
Resolves#4729
Occasionally, swayidle crashes with a failure to connect to the
Wayland session. Ideally, swayidle should automatically restart
instead of leaving the system in a vulnerable state.
