AOT compiled yt-dlp

[asuka-mio]

Is there any possibility to use a AOT compiler to build a yt-dlp native binary image(like graalvm) to optimize away interpreter warmup cost(see https://news.ycombinator.com/item?id=31500505).

[asuka-mio]

Additionally the python interpreter startup delay may be resolved through a "yt-dlp daemon"
Seal start a python interpreter and run a "yt-dlp daemon"(a code refactor), the daemon listen for some download arguments, and Seal send a IPC(via binder or shmem?) to "yt-dlp daemon" when user confirm download?