3 ms·

>The solution is simple. Exchange the client/server model for a peer-to-peer model. By peer-to-peer and I mean literally IP to IP and not something like instant

>The solution is simple. Exchange the client/server model for a peer-to-peer model. By peer-to-peer and I mean literally IP to IP and not something like instant messenger that bounces messages through a central server. Only then can you assure a fix to interference. This is my go to response as to "what to do about social media." And I opine|rant|blather on about it whenever I see folks complaining about or claiming that some rules|regulation|policy will "fix" social media. And such tracts are becoming more and more common (cf. this article[0][1], entitled OK, but What Should We Actually Do About Facebook? I Asked the Experts. from today). But, as in the article linked, no one ever suggests decentralized communications. I'm not sure why, but there certainly are barriers to making decentralized "social media" work. Those same barriers are what make centralization of much of our digital universe necessary and hugely profitable. Those barriers are primarily (there are other issues, but are mostly part of the result rather than the cause) related to the ways in which end-users are provided access to the Internet. I'm specifically referring to asymmetric internet links. Most internet links are $X Mb/sec download and ($X / 25 or so) Mb/sec upload. What's more, there are often policy restrictions on "consumer" internet links that block them from "running servers," whether that be port blocking or throttling. This keeps decentralized communication from being viable. Other than these artificial barriers, there's nothing stopping folks from hosting their own cat/family/meme photos/videos and/or other content. There are plenty of decentralized platforms that can do so with minimal set up and configuration. However, if you can't serve more than a couple of concurrent connections at a reasonable speed, that breaks the whole decentralized/P2P paradigm. Because my upload bandwidth is the limiting factor in being able to host content. If, on the other hand, you have a link with symmetric bandwidth without policy restrictions, serving your own content is a no-brainer. As such, my "solution" for addressing the "censorship" of the centralized "social media" platforms is to make them irrelevant by eliminating the entire reason for their existence -- asymmetric bandwidth favoring download vs. upload on internet links. [0] https://www.nytimes.com/2021/11/04/opinion/facebook-experts.html https://www.nytimes.com/2021/11/04/opinion/facebook-experts.... [1] (same as [0], just from archive.md): https://archive.md/qpNtV https://archive.md/qpNtV Edit: Added title of linked article in the comment.

austincheney 5y ago
YouTube was founded when most people had slow internet and they seemed to have done ok. I have fiber at my house and do about 920mb/s both up and down. It’s only a matter of time before that becomes almost universally common enough to rule out bandwidth from any consideration. As a result I am writing a peer-to-peer application assuming universal fiber, 4K displays, and IPv6 are already common.
- morsch 5y ago
  At the same time YouTube was founded we also had widespread peer to peer audio and video distribution, despite asymmetric Internet connections. Probably the largest peer to peer networks? It worked fine. It was on the verge of being real time for video, when it was replaced by centralised, commercial technology.
- nobody9999 5y ago
  >I have fiber at my house and do about 920mb/s both up and down. It’s only a matter of time before that becomes almost universally common enough to rule out bandwidth from any consideration. I hope you're right. Where I live (the US), most ISPs have cable TV systems, interests in streaming platforms, or both. That provides incentives to maintain the status quo WRT asymmetric links and centralization. ISPs like Verizon, which had previously pushed hard on fiber deployment, are refocusing on wireless, as that's a much bigger profit center for them. I hope you're right about the inevitability of (multi-)GB/sec symmetric connections, but the current incentive structure argues against it, at least in the US. Which is why I believe that last-mile FTTP[0] should largely be local, public/quasi-public[1] utilities that support themselves by selling interconnects to multiple ISPs. Those ISPs can then compete on speed, price, services, etc. [0] https://en.wikipedia.org/wiki/Fiber_to_the_x#FTTP https://en.wikipedia.org/wiki/Fiber_to_the_x#FTTP [1] https://www.investopedia.com/terms/q/quasi_public_corporation.asp https://www.investopedia.com/terms/q/quasi_public_corporatio...

austincheney 5y ago

YouTube was founded when most people had slow internet and they seemed to have done ok. I have fiber at my house and do about 920mb/s both up and down. It’s only a matter of time before that becomes almost universally common enough to rule out bandwidth from any consideration. As a result I am writing a peer-to-peer application assuming universal fiber, 4K displays, and IPv6 are already common.

morsch 5y ago
At the same time YouTube was founded we also had widespread peer to peer audio and video distribution, despite asymmetric Internet connections. Probably the largest peer to peer networks? It worked fine. It was on the verge of being real time for video, when it was replaced by centralised, commercial technology.
nobody9999 5y ago
>I have fiber at my house and do about 920mb/s both up and down. It’s only a matter of time before that becomes almost universally common enough to rule out bandwidth from any consideration. I hope you're right. Where I live (the US), most ISPs have cable TV systems, interests in streaming platforms, or both. That provides incentives to maintain the status quo WRT asymmetric links and centralization. ISPs like Verizon, which had previously pushed hard on fiber deployment, are refocusing on wireless, as that's a much bigger profit center for them. I hope you're right about the inevitability of (multi-)GB/sec symmetric connections, but the current incentive structure argues against it, at least in the US. Which is why I believe that last-mile FTTP[0] should largely be local, public/quasi-public[1] utilities that support themselves by selling interconnects to multiple ISPs. Those ISPs can then compete on speed, price, services, etc. [0] https://en.wikipedia.org/wiki/Fiber_to_the_x#FTTP https://en.wikipedia.org/wiki/Fiber_to_the_x#FTTP [1] https://www.investopedia.com/terms/q/quasi_public_corporation.asp https://www.investopedia.com/terms/q/quasi_public_corporatio...