Hacker News Reader

Comment by ctdonath | Hacker News Reader

bandwidth caps here are killer

I remain amazed that no ad-hoc network solutions are available. Every wireless device & router should provide at least some off-official-network connectivity. It's not even a hardware problem. Over time, enough overlap should provide a pretty good alternate network service with no caps on bandwidth or content, giving incentive for providing bridges and "unapproved" services. Certainly would have its issues, but would provide more options with almost zero cost.

View on HN

Yes, I'm sure ad-hoc mesh wifi is an excellent solution to limited across-the-ocean bandwidth.

It is a solution to monthly data caps, enabling solutions like Popcorn Time. As has been noted, if infrastructure exists to facilitate questionable solutions, then use-case pressures arise to persuade better solutions to larger problems. If people have enough bandwidth to serve their video-streaming interests, some solution will appear, persuading both local data infrastructure improvements and, say, Netflix warehousing data in NZ for more efficient use of across-the-ocean bandwidth.

And speaking of across-the-ocean bandwidth: don't underestimate the bps of a freighter stuffed to the gills with MicroSD cards. https://xkcd.com/691/

The bandwidth is impressive. The latency... not so...

Everyone sharing everything they have downloaded in a giant cache via mesh networking? Yip, its a great solution. One person endures a slow download, all those who follow have a grand speedy old time.

You don't necessarily need across-the-ocean bandwidth to facilitate NZ file sharing. Just adapt the "college dorm has a massive fileserver with pirated movies, so that movies only need to be downloaded from the internet at large once" model.

I blow part of my monthly 'real across-the-ocean internet' limit downloading Night of the Living Dead from archive.org, then provide access to it for other New Zealanders using the hypothetical local mesh network.

So now you have a single home computer and a crappy 802.11 router serving the first week of House of Cards to everyone in New Zealand? Or, even worse, you have a BT net on the mesh which completely overloads whatever poor sap(s) happen(s) to be in the bridge in the graph?

And we haven't even gotten to any of the other problems, like address assignment, routing, etc.

With this purely hypothetical mesh network, ideally files would be spread out across the network, bittorrent style. If the mesh has a single node bridging two otherwise large and connected graphs, then the mesh network would be in a poor state to begin with. This is assuming a healthy network.

More realistically, ISPs could place far higher bandwidth limits on local (to NZ) traffic. That is, if it is just the undersea cables that are lacking, not the rest of the infrastructure as well...

I'm not saying either of these are realistic or easy. I'm just saying that the sarcastic quip "I'm sure ad-hoc mesh wifi is an excellent solution to limited across-the-ocean bandwidth." is not a real criticism of the proposal.

A lot of these are solved problems. In one neighborhood in NYC the local mesh wifi was able to provide better internet service than Verizon. http://rhicenter.org/redhookwifi/

Bittorrent could be scoped to the country too, it doesn't have to seed or download across the ocean.

I have no idea what kind of network you are suggesting.

Every wireless-network device is capable of connecting directly to other wireless devices rather than (or in addition to) just connecting to a router/hub. Imagine every wireless device recursively within range of your device (connected to yours, or connected to another device connected to yours, or ...), all capable of forming an ad-hoc network operating independently of the routers/cell-towers/etc normally used, a wireless network bridging all censorship/control boundaries, operating for free, no way to shut it all down. Insofar as there are limits to its extent, some people would provide gateway points to Internet backbones.

I think you're missing about 4 layers of the OSI stack.

Pardon my lack of encyclopedic thoroughness in a casual discussion.

And I think you're significantly underplaying the technical and financial challenges associated with scaling up a mesh network.

I've watched networking scale from spending 9 hours downloading a book over 110bps (yes, that's one hundred ten bits per second) dialup as an amazing ability, to streaming multiple live HD video from anywhere to wireless devices at 17Mbps (contrast seventeen million bits per second) as a cultural norm.

I do not doubt the ability to solve technical and financial challenges associated with scaling up an ad-hoc network. Been there, done that, can do it again.

https://en.wikipedia.org/wiki/Mesh_network

Reply on news.ycombinator.com