Tibet's Mesh

siriuskase writes "Volunteers are building a low-cost wireless mesh network to provide cheap, reliable data and telephony to community. I would love to see a free wireless mesh that's not dependent on any government or corporation take over the world."

Comments

It's not exactly 'open' or 'free'

[TubeSteak]

Dharamsala's growing mesh is not open to laptop-toting visitors. The bandwidth its operators have to share is limited, costly and much of it comes from BSNL, the government-controlled telecom provider. So for now, access is limited mostly to schools, government offices and nonprofits, which pay a nominal fee and host equipment to further the network's reach.

Admins reluctantly installed a content filter at one site because so many adults were visiting porn sites

Contrast with: "I would love to see a free wireless mesh that's not dependent on any government or corporation take over the world."

Wouldn't we all?

Ultimately, those mesh networks are going to be tied into Gov't or corporate owned backbones. Mesh networks are not going to be the solution for developing countries and I don't understand how anyone expects an independant mesh network to magically appear.

Like anything else, it will require (expensive) investments in infrastructure. Or am I missing something?

Re:It's not exactly 'open' or 'free'

[also-rr]

Like anything else, it will require (expensive) investments in infrastructure. Or am I missing something?

Yes, I did used to do this for a living (and would do it again if someone wants to pay me for it).

In theory there is nothing to stop a world spanning global mesh, except of course for:

1. limitations of the speed of light
2. the size of the routing table
3. lack of trust in intermediate nodes
4. node induced latency
5. oceans

Let's say that your networking technology has a max range of 100m, and the town is a very densly inhabited 10km square, which is pretty much a best case. Getting accross town requires 100 hops and if each of them adds 10ms latency (a pretty low estimate) thats 1000ms. How do you fancy playing quake with 1000ms of lag?

The *best* use for mesh networks is as a complimentary network. Because bandwidth rises in a geometric relationship with the number of nodes (x^n is the limit where x is link bandwidth, but it won't happen due to technology limitations and a bunch of other things) as a very high capacity bulk carrier it probably cant be beaten. Combine it with a smart information distribution system and information redundancy* and you could do some very impressive things.

That way you free up you *real* bandwidth, the latency middle ground of mesh-to-wireless-direct-backhaul, for web browsing as all of the porn and youtube junk is transiting the bulk delivery system. Add on one-hop-to-wired-PoP for latency critical apps such as VoIP and you gain a seamless, layered, high capacity and high performance network.

*I wrote a paper looking at using freenet plus an overlying signing mechanism and co-ordinated seeding as a reliable distribution system in 2003, I'll have to dig it ou and publish it some day.

Re:Free Wifi? How about free Tibet?

[hclyff]

Oh no, here we go again.

"Instead of technology, we should [provide food for the starving children / free the oppressed people / solve everyones problems]."

I know this will generate many doh's, but you are missing the point: with access to the internet comes actual freedom.