Yark: YouTube Archiver with Offline UI
github.com
github.com
( no post download processing, no streaming, just access the "raw file" and make it work no matter what )
Does anyone who used Yark knows if the subtitles are also downloaded?
The "difficulty" was implementing the server/webapp supporting range-requests in a performant ( whatever that means ) and robust way. For example, most browsers request the first bytes with the range: 0- but then when they think they have enough information about the video because they expect the metadata to be in the beginning of the file ( which normally is but not every time ), they cancel the request and do another one. So the server must be minimally robust, also be 100% correct with the byte alignment it sends, a difference of 1 isn't a big deal and the video plays until the very end where the browser request the last bytes, the server tells it is has but the video players still waits for the last byte so it shows it's buffering.
Of course this was a "weekend project", for practical purposes just use VLC and/or a "real" battle tested webserver like Nginx or even Caddy