Author: Marc Fawzi
Twitter: http://twitter.com/#!/marcfawzi
License: Attribution-NonCommercial-ShareAlike 3.0
The One Laptop Per Child (OLPC) project has brought some interest to mesh networking.
In theory, the XO laptop has the ability to form a wireless mesh together with other XO laptops in its vicinity. Each laptop extends the mesh further, like a link in a long chain.
Such mesh technology, when supplemented by signal repeaters, can theoretically cover entire villages. Since villages, towns, cities are connected to each other via the Internet these meshes come together in one world-wide mesh.
The Web and the Internet as a result is constrained (by many economic, political and technical factors) to working within the client-server model. The inability to establish direct communication between applications on different PCs, without having to go through difficult -and sometimes- unreliable paths (think: UDP hole punching, NAT traversal, uPnP etc), combined with ISPs’ tendency to throttle and even block P2P traffic has resulted in an unhealthy environment for P2P applications.
The XO laptop maybe the first sign of a global shift from the client-server model of the Web to the peer-to-peer model of wireless mesh technology.
The current Web architecture is bound to evolve over the next few decades as the architecture of global communication moves from the network-centric model to the peer-to-peer model, enabled by wireless mesh technology.
Related
Evolving Trends 
I hope and wish that someone will start writing a similar program for MESH for regular laptops and release it under GPL License very soon, and create a world wide mesh also.
There can be a security layer added to it, which can secure the data between certain computers, the rest of the people can take advantage of the shared network.
Brilliant Idea Marc.
“OpenMoko is a project to create a software platform for smartphones, using free software.”
http://www.openmoko.com/
http://www.openmoko.org/
Mesh computing sounds like an excellent idea. Does anyone know how slow such a network would be in practice? For example, if it takes 20 mesh hops to reach an actual Internet access point, does that create significant latency?
Andrew, a lot of smart people are working on this and the problem is practically solved, depending on your criteria. There is a guy in South Africa who is promoting mesh transceivers for cell phones that enable phone-to-phone calls for free. Sure it wouldn’t work where there are no repeaters/relays and it has a longer call setup time but when you’re in an area with many peers the actual call itself should flow smoothly. Take a look at the 802.11s standard and ongoing progress that has been made.