3 ms·

That's an interesting position (and I'm not disputing it) but if you were a business that has sovereign territory, you still need utilities (Internet). Penn St

by smoyer 6y ago

That's an interesting position (and I'm not disputing it) but if you were a business that has sovereign territory, you still need utilities (Internet). Penn State has its own data-centers, is part of the Internet2 consortium and has dedicated fiber to both AWS and Azure but you're still ultimately reliant on those you're leasing the connections from - even if you own the fiber, you're probably leasing the right-of-way/poles/etc. So ... how do we build an Internet that's fully inter-connected but can withstand actions by the few relatively large carriers? Thinking about it a bit more, isn't the effect of having a few large network providers analogous to the issue with BitCoin having a few large mining operations in effective control of the ecosystem?

MichaelZuo 6y ago
Satellite fleets?
- smoyer 6y ago
  Which could theoretically be done today but then you'd be a tenant of Musk. I really think the idea of mesh networks is great for neighborhoods but can it be extended to a global-scale network (well ... continent scale until there's a personal communication mechanism that will cross oceans)? EDIT: This was posted on today's "Who is hiring?" thread - not really p2p Internet but ... https://news.ycombinator.com/item?id=25633106 https://news.ycombinator.com/item?id=25633106
  - jki275 6y ago
    You can theoretically run a very low bandwidth set of links over any radio technology, including HF which will give you that long range connectivity. Problem is, people expect streaming video on their smartphones. You can't do that over low bandwidth over the horizon links. Continent scale? Depends on the continent. No way to do that in the Americas, there's just too much open space. You could do it in many cities I do think.
  - MichaelZuo 6y ago
    I imagine there are no upper bounds to the growth of satellite communication capabilities other than physical limits. Extrapolating a few decades at current growth rates out the network be enormous.
toast0 6y ago
Carrier's ability to influence has everything to do with their duopoloy on the last mile to the eyeballs. If you're in the position to lease a connection to an internet exchange, you've got tons of options for transit from there. And leasing fiber is generally a matter of location and cost, not content; I don't know if it's strictly common carrier, but it might be.

by smoyer 6y ago

MichaelZuo 6y ago

Satellite fleets?

smoyer 6y ago
Which could theoretically be done today but then you'd be a tenant of Musk. I really think the idea of mesh networks is great for neighborhoods but can it be extended to a global-scale network (well ... continent scale until there's a personal communication mechanism that will cross oceans)? EDIT: This was posted on today's "Who is hiring?" thread - not really p2p Internet but ... https://news.ycombinator.com/item?id=25633106 https://news.ycombinator.com/item?id=25633106
- jki275 6y ago
  You can theoretically run a very low bandwidth set of links over any radio technology, including HF which will give you that long range connectivity. Problem is, people expect streaming video on their smartphones. You can't do that over low bandwidth over the horizon links. Continent scale? Depends on the continent. No way to do that in the Americas, there's just too much open space. You could do it in many cities I do think.
- MichaelZuo 6y ago
  I imagine there are no upper bounds to the growth of satellite communication capabilities other than physical limits. Extrapolating a few decades at current growth rates out the network be enormous.

toast0 6y ago

Carrier's ability to influence has everything to do with their duopoloy on the last mile to the eyeballs. If you're in the position to lease a connection to an internet exchange, you've got tons of options for transit from there. And leasing fiber is generally a matter of location and cost, not content; I don't know if it's strictly common carrier, but it might be.