YouTube: many related channels being searched, leading to recursion error

[reddragonguy]

Checklist

• I'm reporting a broken site support issue
• [x ] I've verified that I'm running youtube-dl version 2020.11.29
• [ x] I've checked that all provided URLs are alive and playable in a browser
• [x ] I've checked that all URLs and arguments with special characters are properly quoted or escaped
• [x ] I've searched the bugtracker for similar bug reports including closed ones
• I've read bugs section in FAQ

Verbose log

PASTE VERBOSE LOG HERE

Description

Recently i noticed that additional channels were being downloaded that were not specified in my configuration file. the time to processed my file also increased 5x. I added the --playlist-end 30 to my file to try to help with this, and now i get a recursion error. Has something changed where yt-dl now crawls related channels too? Is there a setting to change this?
youtube recursion.txt

[dstftw]

In new Youtube UI channel links point to Home tab that is downloaded. If you need to download another tab - pass the URL of that tab.

[reddragonguy]

shouldn't the recursion error still be looked at and addressed? If i wanted to keep the links i have and the new behavior, then there is still a problem with the looping between friends/channels and the eventual recursion error.

[ZizzyDizzyMC]

I was just coming here to check if this type of issue was reported.

I have a script that uses the latest version of YouTube-DL and noticed earlier today that one of my channel directories were filled with garbage and realized it was downloading things other than just the channel URL provided.

Unless there's something here that I'm missing. I'm using the latest version, but it's the one installed with pip3

Edit: It took me a bit to try and understand what dstftw is saying, does this mean that things were changed in such a way that breaks compatibility between most older download scripts?

If I'm reading this right, instead of downloading a channel by default - youtube-dl now downloads every link in a channels / page? Including playlists and other channels linked on the Home tab of the page.

This seems like functionality that'd have been mitigated and locked behind a command flag to enable, this is dangerous to those who are used to just passing in a channel link. Lots of automation out there that just update the script and pass in channel URL's like my automation tools.

It appears I can mitigate this myself by adding /videos to every channel URL passed into the bot, but I really don't think that should be on the user to know - considering how long the old functionality was kept in a stable state.

[reddragonguy]

@ZizzyDizzyMC yes, i looked at the code after the comment and it seems you now have to add a suffix of /videos, or /playlists, or even /featured to limit the urls searched on the page.

IMHO, it should, by default, limit itself to mimic the old behavior and add /videos and then add a new option to be more greedy if desired. But at least now i can get back to the old way.

[reddragonguy]

@dstftw thank you for the quick fix.

on a tangent, but related matter about time? i've poured over the output file settings many times and have tried various options. How does one use the playlist-index successfully without yt-dl searching all the videos of a channel. I use playlist-reverse, dateafter now-30days, and it still has to look at every single video on the channel and decide to download or not. I tried using playlist-end 30, but then the ids start at 30 (or whatever number used).

Is there a parameter i've missed to use? I'm using this with Plex and always trying to guarantee a unique episode number in the filename: --output '/workdir/%(uploader)s [%(channel_id)s]/%(playlist_index)s - %(title)s [%(id)s].%(ext)s'