I’m making a file organizer and I was thinking of how to integrate torrents. The objective is to find which files are owned by more than one person but have low health and automatically generate a torrent containing those files, share it with all other users with a distributed database and seed it.
But if someone indexes a game for example, that have many small files, there would be many torrents nobody would care about. If I generated torrents automatically when a user creates a collection, then there would be many different collections but I’d prefer to have only a collection for all related files like all books of a given author.
I think the best thing is to allow people to vote on the collections through the distributed database and then whichever client owns more files from the collection creates the torrent and links it to the collection in the database. But even if the collections have votes, that doesn’t group the collections. So I would have a ranking of best to worst collection but no way of knowing which is the most complete collection of books for an author.
I could allow users to vote the files that belong in a public collection. And when they create a local collection if it has many similar files with a public collection I could assign a positive vote for each of those files for that user in the public collection. Then I find the user that has more files in a public collection, so that he creates the torrent and shares it.
I would like to hear your ideas on this.
A general programming discussion community.