This is a technical overview of how everything works, if you are not interested in the backstages please return to the main page
This website was entirely written by hand and without any frontent framework or library, only pure HTML and CSS.
At it's core, it is powered by a free and open source library named yt-dlp, used for retrieving video data and downloading the videos on the server.
The backend is written in Python, and using celery as task delivery with redis for queuing.
The maximum size allowed to download is 2Gb per video, they are then stored on the server, and deleted after 10 to 15 minutes.
When you click on download, you send the url and format to the backend, it then gets the video information and checks its size, and throws and error if the file size is over 2Gb.
It then sends a message to celery, that executes the code for the downloading part on a separate worker, thus allowing several videos to be downloaded simultaneously.
Finally, using yt-dlp, the server downloads the video from youtube and automatically converts it to the requested format with ffmpeg. Then redirects the browser to the downloading page.