That still doesn’t cover a significant portion of youtube’s functionality like reccomendations, comments, allowing creators to edit/adjust videos after the fact.
Seems to me that anything beyond the actual hosting and serving of the video file is unnecessary to include by default in a federated video streaming solution. To drill down a bit, recommendations don’t need to be handled by an algorithm, the content creator can make their own list of videos or playlist - do we really want another reco algo passively controlling what we feed our minds? Comments could be something as simple as a mastodon or lemmy thread with the video as the OP. Content editing after the fact doesn’t seem like its that big a deal aside from computational and bandwidth overhead which would seem small compared to the task of serving multiple thousands of viewers at once.
I think we missed each other. My overall point is that aside from the hosting/serving, other federated networks/services could pick up the slack. The Federated Youtube doesn’t have to mirror Youtube exactly, or even mirror functionality all-inclusively (ie with reccos and comments etc. built-in), but could lean on other federated servers to provide similar functionality.
As I said, comments could be a lemmy/mastodon thread. Recommendations or other discoverability could be other threads or maybe even a completely different service that hasn’t been created yet, I don’t know, but I do know that any reco algo needs to be open and subscribed to, not jammed down our throats and gamed. In the meantime, everyone’s got a search engine, right?
Ultimately I don’t live in this social media/open source/development space too much, I just saw a way for these things to be built/used together to achieve an effect, distributing dev and process overhead and load across all the networks. I don’t have any insight on the bigger, more pertinent, file distribution problem.