I am new to podman, so please forgive me that I might be asking about the basics.

Today, I decided to migrate forgejo into a container, to harden my "server". For this, I backed up forgejo using forgejo dump, which gave me a ZIP file. Now I am lost on putting it into the podman container..

Basically, I want to know:

1. How do I access files inside podman container, especially one created through quadlet? Does the quadlet part make a difference?
2. How can I restore forgejo backup created by forgejo dump? Despite this being seemingly common task, there seems to be no documentation about how to restore the backup.

I have been struggling a lot with setting it up. Until moments ago, I had great difficulty debugging networking issue from firewall, which was quite exhausting. So I would like to ask help first for this problem. If anyone could give some pointers or help, it would be greatly appreciated!

I've been struggling with getting a wezterm window running cmus on a specific workspace upon start up for a while now. I can't use assign because the only eligible criteria differentiating it from a generic wezterm window is the pid, and my attempts to get the pid from get_tree and use that have been unsuccessful. I thought I had figured it out, when I put these lines in a another file:

#! /bin/bash
sway workspace 10 && sway 'exec wezterm -e cmus'

then in my config file I have this: exec ./start_cmus.sh

But it doesn't work. If I run start_cmus from the shell, the expected behavior ensues (a wezterm window running cmus is opened on workspace 10).

Any tips?

I'm going to move away from Windows into Linux and have some strategical questions, among others how do you handle backups/restore and how do you switch distros.

There are two main questions (in case of catastrophic failure or transition):

1. How do you reinstall your apps? I'm considering ansible for that - have all my installs (as much as possible) done through ansible playbooks.
2. How do you recover your backup data? Just copy from backup to home directory probably won't do when it comes to a different distro, or? Here I'm considering doing ZFS snapshosts (ZFS pool/datasets would be used for my home directory) to my backup server but not sure how to recover it in case of switching to a different distro. It should be copy in the case of system restoration I guess.

Any other recommendations?

Both don't ship with their own Wayland compositor, but there are enough to choose from.

Xfce comes with a wayland session using labwc out of the box, but was also tested with Wayfire. The devs state you shouldn't hold your breath waiting for the native window manager xfwm to be ported into a Wayland compositor, since they don't know if/when it will be done. Almost all other Xfce components support Wayland now, while retaining X11 compatibility.

LXQt's newest stable release has full Wayland support, with 7 different Wayland compositors to choose from within a GUI settings menu: Labwc, KWin, Wayfire, Hyprland, Sway, River and Niri

https://xfce.org/about/news/?post=1734220800
https://lxqt-project.org/release/2024/11/05/release-lxqt-2-1-0/

PipeWire 1.4.0 (2025-03-06)

This is the 1.4 release that is API and ABI compatible with previous 1.2.x and 1.0.x releases.

This release contains some of the bigger changes that happened since the 1.2 release last year, including:

• client-rt.conf was removed, all clients now use client.conf and are given RT priority in the data threads.
• UMP (aka MIDI2) support was added and is now the default format to carry MIDI1 and MIDI2 around in PipeWire. There are helper functions to convert between legacy MIDI and UMP.
• The resampler can now precompute (at compile time) some common conversion filters. Delay reporting in the resampler was fixed and improved.
• Bluetooth support for BAP broadcast links and support for hearing aids using ASHA was added. A new G722 codec was also added. Delay reporting and configuration in Bluetooth was improved.
• The ALSA plugin now supports DSD playback when explicitly allowed with the alsa.formats property.
• A PipeWire JACK control API was added.
• A system service was added for pipewire-pulse.
• Many documentation and translation updates.
• Many of the SPA macros are converted to inline functions. All SPA inline functions are now also compiled into a libspa.so library to make it easier to access them from bindings.
• The module-filter-chain graph code was moved to a separate filter-graph SPA plugin so that it becomes usable in more places. EBUR128, param_eq and dcblock plugins were added to filter-graph. The filter graph can now also use fftw for doing convolutions. The audioconvert plugin was optimized and support was added to audioconvert to insert extra filter-graphs in the processing pipeline.
• New helper functions were added to parse JSON format descriptions.
• The profiler now also includes the clock of the followers.
• RISCV CPU support and assembler optimisations were added.
• The clock used for logging timestamps can be configured now.
• The JSON parser was split into core functions and helper.
• Support for UCM split PCMs was added. Instead of alsa-lib splitting up PCMs, PipeWire can mark the PCMs with the correct metadata so that the session manager can use native PipeWire features to do this.
• Support for webrtc2 was added to echo-cancel.
• IEC958 codecs are now detected from the HDMI ELD data.
• Conversion between floating point and 32 bits now preserve 25 bits of precision instead of 24 bits.
• A new Telephony D-BUS API compatible with ofono was added.
• The invoke queues are now more efficient and can be called from multiple threads concurrently.
• Clock information in v4l2 was improved.
• An ffmpeg based videoconvert plugin was added that can be used with the videoadapter.
• The GStreamer elements have improved buffer pool handling and rate matching.
• The combine-stream module can now also mix streams.
• link-factory now checks that the port and node belong together.
• The netjack-manager module has support for autoconnecting streams.
• The native-protocol has support for abstract sockets.
• The pulse server has support for blocking playback and capture in pulse.rules.
• The corked state of stream is now reported correctly in pulse-server.
• Fix backwards jumps in pulse-server.
• Latency configuration support was added in loopback and raop-sink.
• The ROC module has more configuration options.
• The SAP module now only send updated SDP when something changed.
• RTP source now has a standby mode where it idles when there is no data received.
• Support for PTP clocking was added the RTP streams.
• The VBAN receiver can now dynamically create streams when they are detected.
• Error reporting when making links was improved.
• Support for returning (canceling) a dequeued buffer in pw-stream.
• Support for emiting events in pw-stream was added.
• pw-cat now support stdin and stdout.

Highlights (since the previous 1.3.83 release)

• Small fixes and improvements.

PipeWire

• Fix some missing includes in metadata.h
• Pass the current error in errno when a stream is in error (#4574)

modules

• Evaluate node rules before loading adapter follower to ensure properties are set correctly. (#4562)

SPA

• Avoid a use after free when building PODs. (#4445)
• Take headroom into account when calculating resync.

Bluetooth

• Fix +CLCC parsing.

GStreamer

• Notify about default device changes in deviceprovider.
• Copy frames between pools and avoid splitting video buffers.

JACK

• Add an option to disable the MIDI2 port flags. (#4584)