Genuine question (because I’m looking too): without YouTube, where would you go to watch all the diverse videos they host? It’s a really difficult business model. Look at how expensive Floatplane is to the user. Luke and Linus have talked about how difficult it is to run on WAN Show, too: https://youtube.com/watch?v=1mZrsunukUA
A fediverse platform would almost definitely be a worse experience in terms of speed and video quality because residential internet (at least in the majority of the US) just doesn’t have the upload to support multiple HD video streams. Therefore, it’s not really possible to host at home; a basic server at Hetzner could probably do a dozen or two direct streams with no conversion, but storage is kind of expensive just because there’s so much content, and then there’s the need for moderation, high uptime, security, “good” UX design…
Then of course on top of all that when you don’t have creators getting paid by ad revenue, fewer will be able to spend the time on production quality because they’ll be doing it after work, so the length and/or quality suffers.
I dunno dude, I really hope someone smarter than me has figured this out, but it’s a tough problem.
You are correct. Fundamentally, it’s the hosting and storage issue that’s the crux of all this.
And the only choices available are another corporation hosting and paying/passing on the cost, or all of us hosting on a peer-to-peer network, which will be slow, but doable.
Having said that, the peer hosting method would work though, and shouldn’t be dismissed out of hand. We just shouldn’t expect the same level of service we do from YouTube or any corporation hosting videos.
I mean torrenting works but i’ll be damned if I’d need to wait for the buffer to fill up every 30sec for the 1440p video.
You’d need multiple versions pre-encoded to reduce network transfer and serverside transcoding.
I mean torrenting works but i’ll be damned if I’d need to wait for the buffer to fill up every 30sec for the 1440p video.
You’d need multiple versions pre-encoded to reduce network transfer and serverside transcoding.
It would definately be a slog to watch, vs a service that can just deliver the video to you in real time on demand.
My only point is that you would be able to watch the video, after the slog, so that avenue should not be discounted as an option. Its not a great option, but still, an option.
Genuine question (because I’m looking too): without YouTube, where would you go to watch all the diverse videos they host? It’s a really difficult business model. Look at how expensive Floatplane is to the user. Luke and Linus have talked about how difficult it is to run on WAN Show, too: https://youtube.com/watch?v=1mZrsunukUA
A fediverse platform would almost definitely be a worse experience in terms of speed and video quality because residential internet (at least in the majority of the US) just doesn’t have the upload to support multiple HD video streams. Therefore, it’s not really possible to host at home; a basic server at Hetzner could probably do a dozen or two direct streams with no conversion, but storage is kind of expensive just because there’s so much content, and then there’s the need for moderation, high uptime, security, “good” UX design…
Then of course on top of all that when you don’t have creators getting paid by ad revenue, fewer will be able to spend the time on production quality because they’ll be doing it after work, so the length and/or quality suffers.
I dunno dude, I really hope someone smarter than me has figured this out, but it’s a tough problem.
You are correct. Fundamentally, it’s the hosting and storage issue that’s the crux of all this.
And the only choices available are another corporation hosting and paying/passing on the cost, or all of us hosting on a peer-to-peer network, which will be slow, but doable.
Having said that, the peer hosting method would work though, and shouldn’t be dismissed out of hand. We just shouldn’t expect the same level of service we do from YouTube or any corporation hosting videos.
I mean torrenting works but i’ll be damned if I’d need to wait for the buffer to fill up every 30sec for the 1440p video.
You’d need multiple versions pre-encoded to reduce network transfer and serverside transcoding.
It would definately be a slog to watch, vs a service that can just deliver the video to you in real time on demand.
My only point is that you would be able to watch the video, after the slog, so that avenue should not be discounted as an option. Its not a great option, but still, an option.
Peertube is already just as good in terms of performance, it does need more content and better discoverability for sure
Does that performance scale once the user base matches even a fraction of YouTube? Probably not.
It uses P2P so I don’t see why not
I don’t think you can judge performance in an ‘apple versus oranges’ (pop usage) scenario though.
They are using known tech, they did not reinvent the wheel
That’s not my point though. My point is how much capacity versus how many people are using the service.
You’re trying to equate the same level of capacity when there’s different levels of usage.
Finally, how you implement the tech is as important as the tech itself, as it can be done either well, or poorly.