Is your enhancement request related to a problem? Please describe.
As per documentation, public instances typically restart every hour. When viewing a video during this, if the user has proxying enabled, it prevents loading until the user changes qualities or refreshes.
Describe the solution you'd like
The client should retry buffering after ~10-20 seconds of receiving an error. An example error that happens during this restart period is:
VIDEOJS: WARN: Problem encountered with playlist 5-placeholder-uri-5. HLS request errored at URL: https://instance.url/videoplayback... Switching to playlist 5-placeholder-uri-5.
VIDEOJS: WARN: Problem encountered with playlist 5-placeholder-uri-5. Switching to playlist 5-placeholder-uri-5.
Describe alternatives you've considered
The hoster can implement http3-ytproxy as per the public instance improvement guide, and this should prevent this issue. However, not only is this deprecated software and only mentioned in a guide meant for public instance hosters, but this proxy had a lot of stability issues after the recent-ish potoken / visdata / sighelper patch. (Maybe it's just a me issue though since it's working for some ppl.) This also doesn't help if that needs to restart too.
Is your enhancement request related to a problem? Please describe. As per documentation, public instances typically restart every hour. When viewing a video during this, if the user has proxying enabled, it prevents loading until the user changes qualities or refreshes.
Describe the solution you'd like The client should retry buffering after ~10-20 seconds of receiving an error. An example error that happens during this restart period is:
Describe alternatives you've considered The hoster can implement http3-ytproxy as per the public instance improvement guide, and this should prevent this issue. However, not only is this deprecated software and only mentioned in a guide meant for public instance hosters, but this proxy had a lot of stability issues after the recent-ish potoken / visdata / sighelper patch. (Maybe it's just a me issue though since it's working for some ppl.) This also doesn't help if that needs to restart too.
Additional context N/A