I think we've all seen that yt_dlp is a bit in limbo here and will "soon" start using an external JS engine in order to keep up with YT's attempt to break it.
Fortunately regular videos still load even if recent additions to QMPlay2 no longer work at the moment, but playlists (for example https://www.youtube.com/watch?v=1iYc1xP9LLU&list=OLAK5uy_lLg3StcGZwF_pAeOiYzxmZSIpAmB-I2oo) don't load at all.
It does seem that a slow fallback is still possible though:
> yt_dlp -F "https://www.youtube.com/watch?v=1iYc1xP9LLU&list=OLAK5uy_lLg3StcGZwF_pAeOiYzxmZSIpAmB-I2oo" | & fgrep "Extracting URL:" | sed -e "s/.*URL: //g"
https://www.youtube.com/watch?v=1iYc1xP9LLU&list=OLAK5uy_lLg3StcGZwF_pAeOiYzxmZSIpAmB-I2oo
https://www.youtube.com/playlist?list=OLAK5uy_lLg3StcGZwF_pAeOiYzxmZSIpAmB-I2oo
https://www.youtube.com/watch?v=cF2lS0zFCU4
https://www.youtube.com/watch?v=tpExq4JlvYE
https://www.youtube.com/watch?v=OoqRET835o0
https://www.youtube.com/watch?v=D3y9YvHkGb0
https://www.youtube.com/watch?v=0QWDeqK4RfY
https://www.youtube.com/watch?v=o3yKHtvJoiI
https://www.youtube.com/watch?v=1iYc1xP9LLU
https://www.youtube.com/watch?v=qqNwIt3s2sE
https://www.youtube.com/watch?v=pNFWmWmibVo
https://www.youtube.com/watch?v=6o7fBUiTtB0
https://www.youtube.com/watch?v=sXB_kXfN27M
https://www.youtube.com/watch?v=xdSIfcjN3e0
https://www.youtube.com/watch?v=mulGUaPQJKo
https://www.youtube.com/watch?v=2HliaU_EHaY
https://www.youtube.com/watch?v=khn-jDh9x7A
https://www.youtube.com/watch?v=32z_DD_LSgI
https://www.youtube.com/watch?v=cIOP-tJk_68
https://www.youtube.com/watch?v=8t906JiURSI
https://www.youtube.com/watch?v=KWiJ_Cz-2u0
https://www.youtube.com/watch?v=EynWBh0u59s
https://www.youtube.com/watch?v=Wxs-YP8O70w
https://www.youtube.com/watch?v=h2_M-8feKtc
https://www.youtube.com/watch?v=dIg9fTA6fk8
https://www.youtube.com/watch?v=f0AtY6cDcQ0
https://www.youtube.com/watch?v=cJWH10-ZMa8
When I say slow, I mean that just obtaining the watch URLs with the above command takes roughly 1min on my notebook, and fetching all the video titles (after pasting the list into the Playlist panel) takes as many times 20-30 seconds.
Supposing you'd accept to implement a fallback based on this kind of extraction it should at least be possible to cut short the time to populate the new playlist group, by using the URLs as titles initially and only updating them when they are played?
I think we've all seen that yt_dlp is a bit in limbo here and will "soon" start using an external JS engine in order to keep up with YT's attempt to break it.
Fortunately regular videos still load even if recent additions to QMPlay2 no longer work at the moment, but playlists (for example https://www.youtube.com/watch?v=1iYc1xP9LLU&list=OLAK5uy_lLg3StcGZwF_pAeOiYzxmZSIpAmB-I2oo) don't load at all.
It does seem that a slow fallback is still possible though:
When I say slow, I mean that just obtaining the watch URLs with the above command takes roughly 1min on my notebook, and fetching all the video titles (after pasting the list into the
Playlistpanel) takes as many times 20-30 seconds.Supposing you'd accept to implement a fallback based on this kind of extraction it should at least be possible to cut short the time to populate the new playlist group, by using the URLs as titles initially and only updating them when they are played?