>For me it looks like the hosting provider gets a huge savings in terms of bandwidth.
The constraint on wide adoption is literally the quantity of "peers" in both Peertube and peer-2-peer.
The underlying human incentives are not there for most people to host a peer node. This limitation applies to all types of digital domains including videos, or files (IPFS or bittorrent), or crypto (Bitcoin/Ethereum peer node).
Let's follow the trail of incentives for one Peertube node:
- in that instance's "About" page, it says: How we will pay for keeping our instance running -- personal funds. If you feel like donating, put up your own instance instead and host some creators you find interesting.
Companies can't build a Youtube/TikTok competitor based on examples like that. Same forces of economic incentives that limits quantity of IPFS peers which means businesses can't use it to replace Cloudflare CDN or AWS S3 buckets.
It should be understandable why most people aren't willing to spend personal funds on hosting home nodes so businesses can freeload off of p2p and save money on bandwidth.
That list of instances is randomly ordered, it isn't like biggest first or anything.
I don't really see how the fact that there is someone who hosts videos using personal funds proves that it isn't possible for a professional company to host videos using the same company. It proves Peertube does work at small scale, and says "not much" about how it would handle large scale with a real budget.
The human incentives are certainly there, it’s the copyright and IP industry that does not want this to come to pass. There was a time when you could find obscure things on torrents with seeders, but that was shut down by the IP rent seeking industry.
I think you’re overestimating how strong that inventive is. There’s a lot that I’d change about the current copyright model but it’s not just rent-seeking keeping P2P from mainstream use: most people do not have high uplink capacity and congestion is a real issue. Companies like Netflix have built a consistent user experience where video plays quickly and smoothly, and the price is low enough that most people aren’t jumping for alternatives — I would bet that if you surveyed the average person the biggest copyright reform they’d want is blocking exclusivity so they could use a single service for everything.
From a business perspective, a big problem is trust and safety. Torrents were also notorious for having mislabeled content or malware, and users are not going to use a system which serves arbitrary content from their IPs, so you have challenges keeping the experience good while building a large enough peer network to provide a competitive experience.
I would bet that if you surveyed the average person the biggest copyright reform they’d want is blocking exclusivity so they could use a single service for everything.
This. While there's definitely piracy before internet, most don't bother because it plays on VHS/DVD. As Gabe Newell stated, the problem of piracy is a service problem.
Isn’t that a result of the push to cable-tvize the internet? Symmetric high bandwidth internet became commonplace during the height of the P2P downloading era where I live. Because the market was there.
I presume it is just a consequence of the enormous costs to build out a fiber network that would yield uncompetitive prices compared to the internet people are willing to accept from their coaxial cable internet providers.
A combination of people do not demand it enough to pay sufficiently to get neighborhoods wired with fiber, and coaxial cable companies having a disincentive to allocate more of their wire capacity to internet bandwidth since they want to avoid becoming a dumb pipe.
The only way out is taxpayer subsidized build out of fiber operated like a utility, such as the system in Chattanooga Tennessee.
Comments
>For me it looks like the hosting provider gets a huge savings in terms of bandwidth.
The constraint on wide adoption is literally the quantity of "peers" in both Peertube and peer-2-peer.
The underlying human incentives are not there for most people to host a peer node. This limitation applies to all types of digital domains including videos, or files (IPFS or bittorrent), or crypto (Bitcoin/Ethereum peer node).
Let's follow the trail of incentives for one Peertube node:
- see list of Peertube instances: https://joinpeertube.org/instances#instances-list
- I pick the 2nd one on the list: https://the.jokertv.eu/about/instance
- in that instance's "About" page, it says: How we will pay for keeping our instance running -- personal funds. If you feel like donating, put up your own instance instead and host some creators you find interesting.
Companies can't build a Youtube/TikTok competitor based on examples like that. Same forces of economic incentives that limits quantity of IPFS peers which means businesses can't use it to replace Cloudflare CDN or AWS S3 buckets.
It should be understandable why most people aren't willing to spend personal funds on hosting home nodes so businesses can freeload off of p2p and save money on bandwidth.
That list of instances is randomly ordered, it isn't like biggest first or anything.
I don't really see how the fact that there is someone who hosts videos using personal funds proves that it isn't possible for a professional company to host videos using the same company. It proves Peertube does work at small scale, and says "not much" about how it would handle large scale with a real budget.
The human incentives are certainly there, it’s the copyright and IP industry that does not want this to come to pass. There was a time when you could find obscure things on torrents with seeders, but that was shut down by the IP rent seeking industry.
I think you’re overestimating how strong that inventive is. There’s a lot that I’d change about the current copyright model but it’s not just rent-seeking keeping P2P from mainstream use: most people do not have high uplink capacity and congestion is a real issue. Companies like Netflix have built a consistent user experience where video plays quickly and smoothly, and the price is low enough that most people aren’t jumping for alternatives — I would bet that if you surveyed the average person the biggest copyright reform they’d want is blocking exclusivity so they could use a single service for everything.
From a business perspective, a big problem is trust and safety. Torrents were also notorious for having mislabeled content or malware, and users are not going to use a system which serves arbitrary content from their IPs, so you have challenges keeping the experience good while building a large enough peer network to provide a competitive experience.
This. While there's definitely piracy before internet, most don't bother because it plays on VHS/DVD. As Gabe Newell stated, the problem of piracy is a service problem.
And also lack of upload capacity for 90% of people due to the lack of symmetric fiber availability.
Isn’t that a result of the push to cable-tvize the internet? Symmetric high bandwidth internet became commonplace during the height of the P2P downloading era where I live. Because the market was there.
I presume it is just a consequence of the enormous costs to build out a fiber network that would yield uncompetitive prices compared to the internet people are willing to accept from their coaxial cable internet providers.
A combination of people do not demand it enough to pay sufficiently to get neighborhoods wired with fiber, and coaxial cable companies having a disincentive to allocate more of their wire capacity to internet bandwidth since they want to avoid becoming a dumb pipe.
The only way out is taxpayer subsidized build out of fiber operated like a utility, such as the system in Chattanooga Tennessee.