On Tue, Oct 29, 2013 at 1:30 PM, Tony Anderson wrote:
> Hi,
>
> There was discussion of this at the SF sprint.
>
> As I understand it, openWRT (from the Shuttleworth project) can be installed
> on
> a TP-Link router.
As with many open source projects the lineage can be a bit confusing.
The Mesh Potato is a appliance developed and sold by
http://villagetelco.org/ . Villagetelco is a the Shuttleworth project.
the software side of the project is a thin and clever mesh layer +
confiuration GUI running on top of openWRT. In the past year terry ( I
forget his last name) has been repurposing it software to provide mesh
data service.
>It can be configured to serve connected XOs (such as in a
> classroom)
> on a mesh. It could also serve as a gateway to the school server's LAN.
>
> If this could work, then the schoolserver would not provide DHCP (since the
> mesh potato routers would
> have fixed addresses on the schoolserver LAN) and would not provide ejabberd
> (since that function is
> served by the classroom-level mesh).
>
> Presumably each classroom mesh would be a subnet of the LAN so that openWRT
> would act as a gateway
> for messages directed at the schoolserver.
I don't know the technical details. OLPC-AU has been piloting some
systems at the classroom level they might have more information. Terry
is clever enough to solve this level of issues in a couple of hours if
necessary.
> Is my understanding in the ballpark?
>
> Is there someone in the community who is working on this capability or is an
> appropriate reference for further
> information?
I spent a couple of weeks testing the system. The mesh aspect brings
many advantages in terms of set up because it requires no wiring.
Theoretically you could place a potato anywhere with in range of an
existing mesh and it would just work. The initial mesh is just a
matter or setting up the master potato.
My challenge was contention. Many inexpensive routers are single
frequency and many existing laptops use the 2.4 GHz range. That give
you 4 non-overlapping channels with with to work.
My conclusion was that while the villagetelco work is potentially very
valuable, tooling which makes it easier for unskilled or semiskilled
facilitators to figure what is wrong with a network and then fix it
were of greater priority. If you or anyone else want to look into that
would be most welcome.
Much of the value of enterprise routers, which can cost more then my
car, is on the software side.
Please see http://villagetelco.org/
> Tony
> ___
> Server-devel mailing list
> Server-devel@lists.laptop.org
> http://lists.laptop.org/listinfo/server-devel
--
David Farning
Activity Central: http://www.activitycentral.com
___
Server-devel mailing list
Server-devel@lists.laptop.org
http://lists.laptop.org/listinfo/server-devel
