vanilla-os / vanilla-system-operator Goto Github PK

task.go line 134 assigns a new local CurrentQueue and shadows the package level one.

Issue Description

The above task was created to automatically start syncthing, installed via apx, on boot. However, it never gets started.

Steps to Reproduce

$ apx install syncthing
$ vso create-task -n "Syncthing" -d "start syncthing on boot" --on-boot -c "apx run syncthing"
$ reboot

after boot:

$ pgrep syncthing

On what version of Vanilla OS this happens?

22.10

Additional Information

No response

Steps to replicate

• Disable SmartUpdate.
• Trigger an Update using vso.

Error

When the vso update is triggered with smart update disabled nothing happens upon confirmation of the message.

cc @mirkobrombin

updates.smart: %!s(bool=false)

using string format %s instead of %v.

Log.

kbdharun@apx-vso-pico:~$ vso android info
Downloading index for repository FDroid
Downloading index for repository izzyondroid
panic: runtime error: index out of range [0] with length 0

goroutine 1 [running]:
github.com/vanilla-os/vanilla-system-operator/cmd.wayInfo(0xc000133d00?, {0xe49740, 0x0, 0x0})
	/__w/vanilla-system-operator/vanilla-system-operator/cmd/waydroid.go:222 +0x1bf
github.com/spf13/cobra.(*Command).execute(0xc000376000, {0xe49740, 0x0, 0x0})
	/__w/vanilla-system-operator/vanilla-system-operator/vendor/github.com/spf13/cobra/command.go:983 +0xabc
github.com/spf13/cobra.(*Command).ExecuteC(0xc000004300)
	/__w/vanilla-system-operator/vanilla-system-operator/vendor/github.com/spf13/cobra/command.go:1115 +0x3ff
github.com/spf13/cobra.(*Command).Execute(...)
	/__w/vanilla-system-operator/vanilla-system-operator/vendor/github.com/spf13/cobra/command.go:1039
github.com/vanilla-os/orchid/cmdr.(*App).Run(0xc000004300?)
	/__w/vanilla-system-operator/vanilla-system-operator/vendor/github.com/vanilla-os/orchid/cmdr/app.go:108 +0x6d
main.main()
	/__w/vanilla-system-operator/vanilla-system-operator/main.go:62 +0x753

Programs that require full device access like adb are unable to work correctly because of missing permissions. It would make running some applications unreasonably difficult.

See here for a possible solution:
89luca89/distrobox#762 (comment)

Another (easier) solution would be using a rootful distrobox container, though that would have other downsides.

Currently the upgrade process for the VSO container only involves our APT repositories, if major changes are made to the image, the container will never get them.

• support for ABRoot v2 new command syntax
• use the new private pkg manager to check/list for updates

Replaces #23

Many users reboot while an update is currently happening, we should make an indicator which shows them that an update is going on.

This will preload the shell (not init) at boot.

Requires: Vanilla-OS/apx#314

Tasks:

• skip on a metered or poor connection
• implement a start command which starts the container if it exist
• create a systemd unit (or any other solution which runs on login) to execute that command

ref: https://docs.podman.io/en/latest/markdown/podman-generate-systemd.1.html

This issue is here as a placeholder. We are discussing this in other development channels.

VSO won't run commands with additional command arguments, i.e. vso run htop -h. It instead prompts you with the "correct" way to use VSO.

In many scenarios in particular workflows it will be better to keep containers independent which was the case in vso v1.
In my testing vso 2.0 seems to break some of this workflow let me explain with an example

• You are using tmux which gets automatically invoked in your default vso container.
• Now you create more apx containers
• tmux gets called in those containers as well as you are auto invoking a tmux session
• So what happens is you invoke a tmux session inside a tmux session which ends up being a loop

This was not the case in in the alpha releases. As all the apx containers were invoked through the host shell.. so containers ran independent of each other despite having same config. Which in my opinion should be the better way otherwise this might conflict with many workflows.

A good thing that can be done is that what apx command inside a vso container does is -> call the host-shell -> which will then invoke the apx container you are trying to enter and exits out of the vso container.

Example -

if you are in a vso-container -> apx enter devbox -> calls the host-shell to invoke the apx enter devbox command and exits the vso container -> devbox container is called from the host-shell which then runs as intended.

I can always do this myself in my configs to check if its an apx container then dont invoke tmux and other things, however this can be considered.

I hope i was able to explain the niche !

If this feature is turned On, the update will be postponed when the resource usage is too high.

Checks:

• connection usage
• memory usage
• cpu usage
• disk usage
• battery level

my Surface Laptop 3 reports two batteries, one of which is probably a bug, but the other is valid and fully charged.

vso's acpi check doesn't account for multiple batteries, which are common in surface devices, and other laptops.

Battery 0: Discharging, 0%, rate information unavailable
Battery 1: Full, 100%
root@vanilla:/etc/vso# 

The go.mod defines the project as github.com/vanilla-os/vso, but the repo is under github.com/vanilla-os/vanilla-system-operator.

Trying to go get this repository as vanilla-os/vso fails because that repository doesn't exist, and using vanilla-os/vanilla-system-operator returns an error since the user "wanted" vanilla-system-operator but go.mod returned vso.

This happens both when you run a binary, i.e. vso run git, and when you exit vso shell. The exact error is:

Error: exit status 1
Usage:
  vso shell [flags]
  
Flags:
  -h, --help   help for shell
  
  ERROR exit status 1

run go mod vendor and make appropriate changes to the debian folder.

With VSO being the main shell in Vanilla OS 2 Orchid, the utility must be able to upgrade the vso container alongside with the host, perhaps following the smart update logic.

Log

kbdharun@apx-vso-pico:~$ vso android init
Config loaded from /etc/vso/config.json
Trying to pull ghcr.io/vanilla-os/waydroid:main...
Getting image source signatures
Copying blob 4f4fb700ef54 done  
Copying blob 12139c1b9ec5 done  
Copying blob 4f4fb700ef54 done  
Copying blob 48c19c37ed9e done  
Copying blob 1da2c33c2e6c done  
Copying blob 5c7c5f8610fb done  
Copying blob fe198750a2bb done  
Copying blob 5d32925ff897 done  
Copying blob 7aa922ba585e done  
Copying blob 3e3c9348720c done  
Copying config 16f14858f8 done  
Writing manifest to image destination
Storing signatures
16f14858f8e5f21bdc9cba08fca6c5627a23891378ca8e63f4234ae152a336b3
Creating 'apx-vso-waydroid' using image ghcr.io/vanilla-os/waydroid:main	 [ OK ]
Distrobox 'apx-vso-waydroid' successfully created.
To enter, run:

distrobox enter --root apx-vso-waydroid

Starting container...                   	 [ OK ]
Installing basic packages...            	 [ OK ]
Setting up devpts mounts...             	 [ OK ]
Setting up read-only mounts...          	 [ OK ]
Setting up read-write mounts...         	 [ OK ]
Setting up host's sockets integration...	 [ OK ]
Integrating host's themes, icons, fonts...	 [ OK ]
Setting up distrobox profile...         	 [ OK ]
Setting up sudo...                      	 [ OK ]
Setting up groups...                    	 [ OK ]
Setting up users...                     	 [ OK ]
Setting up skel...                      	 [ OK ]
Setting up init system...               	 [ OK ]
Firing up init system...                	 [ OK ]
Checking if Waydroid was initialized...
1.4.2
 SUCCESS  The waydroid subsystem has been initialized.


kbdharun@apx-vso-pico:~$ vso android launcher
Config loaded from /etc/vso/config.json
Checking if Waydroid was initialized...
ERROR: WayDroid is not initialized, run "waydroid init"

I propose the support for checking system updates from the gnome software center for making the system more centered and easy to use like fedora

When performing an Update via sudo vso trigger-update --now it runs the apt update, apt upgrade routine and asks for the reboot to make the transaction.

After rebooting the abroot command is missing. VSO becomes version 1.2.0 (if i remember correctly)
Console gives and error: Unknown command: abroot

Triggering an update again via VSO leads to:

Triggering update now...
Smart update is disabled, skipping check.
An error occurred while trying to update the system.
    Please try again later and report the issue if it persists.

    ERR_CODE: %s CMD_TR_UP_NOW
Error: exec: "abroot": executable file not found in $PATH

VSO -v: 1.1.4
ABroot -v: 1.2.2

• populate en.yml with the apx.Trans(..) strings in the cmd package
• add to weblate

Sometimes, after exiting the vso shell, it complains that it is not reachable and suggest deleting the container.

This happens whenever the shell exits with a non zero error code. This could happen for example when pressing Ctrl+D after a failed command.

To replicate

• vso shell
• exit 1

Issue Description

I can't use Waydroid at all, setup fails complaining /dev/binderfs/binder-control doesn't exist.

Steps to Reproduce

As far as I can tell, it is caused by using Orchid with IBT enabled (11th gen intel or newer)

On what version of Vanilla OS this happens?

Unreleased

Additional Information

Attempting to mount /dev/binderfs mount crashes with a segfault, dmesg shows a missing endbr instruction. This is a security feature being triggered caused by the binder module not being compiled with the right options, see https://lwn.net/Articles/889475/

At a glance, it would appear the module should be compiled with -fcf-protection=branch, but i'm not sure how that needs to happen, nor if it's anbox's issue or an issue with dkms.

While I was working on updating the strings, I noticed that there are strings to display errors when exporting or unexporting binaries and applications. But, it doesn't seem to be implemented in pico.go.

Reference (from Pico in en.yml)

error:
    exportingApp: "An error occurred while exporting the application: %s"
    exportingBin: "An error occurred while exporting the binary: %s"
    unexportingApp: "An error occurred while unexporting the application: %s"
    unexportingBin: "An error occurred while unexporting the binary: %s"

in vanilla 2 orchid last update
when i try to install an apk it tells me Installation complete but i cant find the app

I suggest to code a Kernel upgrade utility like Linux Mint or Manjaro Linux. To make the hardware config more easiest.

When I try to install anything it gives error message:
E: Problem executing scripts DPkg::Post-Invoke '/usr/share/vso/hooks/apt-post'
E: Sub process returned an error code
Error: exit status 100
Usage: vso install [flags]
Flags: -h --help help for install

When the user choose the other root partition on boot, propose the rollback via GUI.
VSO has not a GUI but this is the project which best suits the feature.

To do so, use abroot rollback --check-only, it will return 0 if it is possible to rollback, then 1 if not.

with smart updates turned off, vso trigger-update --now still refuses to update because checks failed.

root@vanilla:/etc/vso# cat config.json 
{
  "updates": {
    "schedule": "weekly",
    "smart": false
  }
}root@vanilla:/etc/vso# acpi
Battery 0: Discharging, 0%, rate information unavailable
Battery 1: Full, 100%
root@vanilla:/etc/vso# vso trigger-update --now
Triggering update now...
Starting smart update check...
Low battery detected, skipping update.
Smart update detected device is being used, skipping update.
root@vanilla:/etc/vso# 

We have new events in the CommonChecks which can be used as task events for memory, connection and cpu tasks.

If you use vso export something, it will output --app-name or --bin must be specified..

This is not correct, since the flag should be --app, not --app-name

So that you can install/remove distro-specific software via GNOME Software or some similar app store.

When VSO starts (container boot), systemd logs appears in the user shell, instead of in the background.