this post was submitted on 20 Dec 2024
590 points (96.7% liked)
Technology
60052 readers
3344 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I would want to see some data on costs, because I think you might be overselling the difficulty and cost a bit (I don't actually know, just my good faith belief). Imagine if every content creator ran their own instance. Instead of needing to worry about every user coming to a single group of servers, the Creator only needs to worry about the cost of hosting their own content and the traffic they get.
With the number of YouTubers who have to get sponsorships and Patreon anyway, it doesn't really seem that infeasible or unreasonable to expect content creators to run their own thing or pay to have someone else to do it. Doesn't seem like the YouTube money is that lucrative, anymore, so not like it would be all that different, either.
https://gbtimes.com/how-much-does-youtube/
https://fastercapital.com/content/YouTube-cost-structure--YouTube-Analytics--Unveiling-the-Hidden-Costs-in-the-Cost-Structure.html
Estimated annual server cost: approximately $1 billion
Estimated annual data center cost: approximately $5 billion
Estimated annual bandwidth cost: $3 billion
Good luck running that shit from your closet server.
Sure, but you're assuming all content is on one server. With something like PeerTube, content is federated.
That said, I don't think federation is the solution here because a popular video is going to completely swamp that instance, but something P2P would probably work if you can stream from multiple seeders. Even if you copy like we do w/ Lemmy, you'd still end up with a handful of instances that are way more popular than the rest and those would get hammered if there's a particularly popular video.
If you can spread that $6B (ignoring bandwidth here) over 10M people, you end up with a very reasonable $600/year, and costs would go down as more people join the network. I also assume a lot of that is duplication to handle demand spikes, which is baked in to the P2P system, so a P2P system would probably be way cheaper to scale up.
If you read the links, this includes their server clusters and employees across the entire world all doing complex load balancing and maintenance.
Sure, and none of that is necessary with a proper P2P system. If I'm torrenting something, it'll naturally pull from seeders near me over seeders on the other side of the planet, so load balancing happens by every client being greedy.
The complex load balancing is only necessary because it's a centralized service.
This protocol already exists and so does the system, PeerTube.
Why no significant quantity of people use it is apparent after you try it for a while; the entire server system cannot handle the commensurate volume of content and interactions that YouTube is popular for.
I thought PeerTube's problem was largely federation (need to know which servers to use), which results in making it hard to find content to watch and probably has something to do with how load balancing works (i.e. are you mostly streaming from your instance?). I think Lemmy has a similar problem, but it's at least pretty fast because text and images are a lot easier to manage than video.
Which is, in fact, exactly how PeerTube works: it's got BitTorrent built right into it.
Frankly, it's ridiculous how people keep harping on this "problem" as if it isn't long since solved.
I thought it was largely federated? I don't know how the internals work, so I don't know what group of peers it'll pull from.
Regardless, the problem PeerTube has little to do with its technical foundation IMO, but the network effect. If we get people to start using it, either we'll fix it or we'll develop something better, but getting creators to move is the first step.
According to that first link, it costs $6.1 billion to $11.7 billion annually to run YouTube. Even if you segment that into niche video communities, it'll still cost hundreds of millions of dollars annually to host it, if you get a decent amount of traffic.
This is why YouTube is a monopoly. Because they have the ridiculous amount of money to throw at a "free" video hosting site. Any other video host would crumble under the weight of YouTube's level of traffic. That's also why some others, like Nebula, require a subscription model to function. Or any movie/TV show streaming service. They can't afford to host that stuff for free.
This is also why Google is so obsessed with cracking down on anti-ad software. That's how they make the money that pays for YouTube.
According to this, as of Jan 2024 there were 14 Billion videos on Yt. So effectively a dollar and change to host a video for all YT users.
Obviously it doesn't work like that, but if the above commenter's point was that I, a content creator, host my video and manage my own costs, and that video is linked via whatever federation, I can monetize and limit as needed as a creator, thus popular videos get paid to host, and unpopular videos are hosted for more or less table stakes because they're only getting X hits per Month.
Some kind of WordPress-like container with a built-in safety switch for overages and - hey presto. Well, it's a thought anyway.
I dunno, it seems do-able, even if the Great Unwashed are going to stick with YT and getting ads up the wazoo to see "I Stuffed My Face In A Fusion Reactor - Watch What Happens Next" and the like.
Several tried. Nothing as elaborate as cross dissemination, federation or whatever. But at least 5 to 10 years ago it proved to be almost impossible. Platforms like Rooster teeth, which was 100% subscription based, I think never broke even and still relied on YT ads for the majority of the revenue. Some big and small channels tried to at least just catalog, archive and serve their own videos and the costs still became astronomical really fast. Whenever you see one of those very old channels, most of them don't conserve copies, let alone original source footage of their entire material. Everyone just delete their videos once they've been on YouTube for a month or so now, and they have to download their own videos when they want to reuse old footage.
Storage is cheap today, yes, but video really eats storage at an alarming rate. Specially now that 4k is the standard. So you have to reuse storage over and over. Transcoding is also really fast and optimized with modern algorithms, but it takes specialized graphical cards and data centers charge a premium to use servers with such capacities. Self hosting will never be able to satisfy a moderate demand. Get anything above 100 users simultaneously transcoding videos and a non-specialized server will halt to a grind just on IO calls to hard drives alone.
Once you consider all those factors it is obvious why YouTube is such a miracle.
peertube federates on activitypub
It's such a miracle becuase the world gave it all of their content for free.
At least in part.