[BUG] Memory leak, unresponsive application until DE kills it

[voiceinthedark]

Describe the bug
Since the last update, the application stopped responding, and kept exhausting system memory until the process gets killed by Gnome desktop.
I thought this was a problem because of running the process from inside tmux, however the same behavior occurs outside of tmux. I tested this on my main terminal emulator, Alacritty as well as on the default arch terminal, same behavior.

To Reproduce
Steps to reproduce the behavior:

1. run pyradio
2. press L on any station to start playing
3. process hangs

Expected behavior
I expected that pressing the play button will play the selected station.

System (please complete the following information):

• OS: EndeavorOS (Built on Arch)
• Python version: 3.13.7
• PyRadio Version: 0.9.3.11.19
• Player used: mpv

Additional context
Pyradio is installed through AUR

pyradio.log

[s-n-g]

I have installed the OS on VM, but cannot update it at all

I have updated the mirrorlist and all... but nothing works

Is there something I am not doing right?

I am running the latest pyradio version on manjaro and arch and i have no problem whatsoever

Let's look at it from another angle...
Does mpv plays stations?
Try this:

mpv --no-video --quiet --profile=pyradio https://stream.zeno.fm/qpn8mkt8c4duv

Do you hear sound?
Can you adjust the volume with / and *?

[voiceinthedark]

I attempted it and heard nothing at all. It also eats up a huge chunk of my memory, it jumps from 40% usage to 75% while running the process on an 8GB system.
That's weird, is the problem mpv related?
I have ncmpcpp media player and it's working fine

[s-n-g]

Yes, this is mpv related

Try without any parameters

mpv https://stream.zeno.fm/qpn8mkt8c4duv

Still the same?

Did you install anything related to mpv?
Any lua script or anything?
Try this

mv ~/.config/mpv ~/.config/mpv.old

and try to play the station with mpv

BΤW, ncmpcpp is not related to mpv, it is a front-end for MPD, a totally different thing

[voiceinthedark]

Thank you very much. It worked after removing the config file, it was more than a 35,172,165 millions of records of pyradio volume settings, basically 235MB of config file, wonder what caused it.

[s-n-g]

ok, glad it worked out!

This was probably because of a bug that existed in a previous pyradio version that would append instead of overwrite volume settings in mpv's config file. So, this was my bad, sorry!