.cache error

[nelwa]

Hi,

Firstly, thank you for this amazing script!

When running the program it gives these error messages:

Couldn't read cache at: .cache                               cache_handler.py:87
Couldn't write token to cache at: .cache                     cache_handler.py:97

The script still works, but it means that if a large playlist download is interrupted, and started again, it does not resume at the next song in the playlist, and tries to download everything again.

I could not find any information anywhere on how to configure the .cache directory, but I've tried setting permissions of my ~/.cache folder which had no effect.

Any thoughts?

[nelwa]

Further, is there a way to permanently set the client_id & client_secret variables? I have to re-enter them on any new terminal opened.

[SathyaBhat]

Hey @nelwa

the "couldn't write to cache" message is about the authentication tokens, not the downloads. This has been a long standing bug that comes up on specific machines but I haven't been able to figure out which ones, especially since it comes from the underlying library.

The playlist resume behaviour certainly can be improved - I haven't had the time to look into rewriting it, to be honest. But I understand the pain.

For client_id and client secret to be set permanently, update your shell's profile files to set it. See this question for example: https://unix.stackexchange.com/questions/117467/how-to-permanently-set-environmental-variables

[Maritsu]

[...] This has been a long standing bug that comes up on specific machines but I haven't been able to figure out which ones, [...].

If it's of any help, I'm getting the same caching error on a Lenovo Legion 5-15IMH05, running EndeavourOS x86_64; Python v3.11.6.

[Maritsu]

[...] This has been a long standing bug that comes up on specific machines but I haven't been able to figure out which ones, [...].

If it's of any help, I'm getting the same caching error on a Lenovo Legion 5-15IMH05, running EndeavourOS x86_64; Python v3.11.6.

Update: the issue has been resolved for me when running the command again, this time in the output directory.