[Battlemesh] LibreMesh

[willi uebelherr]

Dear Paul and Stefan,

many thanks for your open answer. I also ask me, can the grafic explain 
something? Can "How it works" help us to better understand?

Of course, yes. If we have a our own clear understanding. It is not 
necessary, that it is more true then the others. But it is a condition 
to reflect and come in to a interactive process.

"Mesh networks are such networks where all participants (nodes) are able 
to route traffic from other participants". It is depend, what we mean 
under "route". It can be only a transport of traffic, if this node have 
no alternatives. If we cascade nodes with his access point spaces, then 
every node is also a router.

For me, the primary difference in the terms, that we have to make is 
node interconnection, a p2P-connection, and the access-point space. And 
this is also valid if the device is a node and a user enddevice in one box.

Mostly in our reality only this AP-function is used. Like a extender 
oder WDS-device.

The examples, what Dave Taht referenced, is part of that. A compact 
connectivity extender with wireless technology.

Paul wrote about ad-hoc and mesh network. Ad-hoc means for me, it is 
dynamic integratable. But it says nothing about the inner structure.

MeshNet or MeshNetwork or Meshed Network? Maybe, it comes from my german 
language. A mesh (masche) is like a circle. A Meshnet is a 2 dimensional 
arrangement of meshes. The net is a 2-dimensional interconection of 
nodes. It can look like many meshes, if you forget the interconnection 
points.

Under network we understand to act in a connected environment. But it 
says nothing about the inner structure of this interconnection.

If we use the term Battlemesh, what do we mean by that? Battle is clear. 
If we use the inner net structure and mechanism as a black box, then we 
check the transfercapacity and the different latencys and the error 
densities in our access between the user devices in our testnet and to 
user devices outside.

We know, the result is a accumulation of many different steps and 
phases. But maybe, this can be the best practice for a BattleMesh.

but for our middle and longtime future? Can we stay always in wait 
position and are driven from any new market product? In the text to the 
Google WiFi kit the jounalist use 20-30 and sometimes more WiFi-devices 
for his text environment in his small house. But this can never be our 
base, or?

In the Guifi-map we see many points in North Spain, some lines (maybe 
cables), some lines go to Pakistan and Colombia, but without to 
understand, what is inside. You have a connection to Frankfurt. Is that 
your directly access to North America and Asia connection over the IXP 
in Frankfurt? And the technical specification of this links.

This map is not directly helpful for this discussion. Also not the map 
from Ninux.

Can we say, we don't have Net-structure in the telecommunication in the 
existing Community Networks like in the private systems? In freifunk 
also i don't found it.

Please, i hope, you are not to much angry with me. I want to understand, 
how you think about and how you see your own activities, that are really 
very valuable.

many greetings, willi


On 16/12/2016 12:23, Pau wrote:
> IMO there is not a single definition for mesh network. It depends on the
> context. For my understanding the definition from libremesh.org is quite
> correct in our context "Mesh networks are such networks where all
> participants (nodes) are able to route traffic from other participants".
> So all nodes are able to route. Thus IMO when we talk about mesh network
> we talk about layer 3 (or layer2 in case of Bat-adv), but not layer 1.
>
> Guifi.net is not a mesh network (or at least we do not define it as a
> mesh network), but a community network using mainly BGP/OLSR following a
> star topology. So there are SuperNodes (which are able to route) and
> standard nodes connected to some SuperNode (nodes not able to route
> packets from the main network). However inside Guifi.net there are some
> mesh clouds which are actually small mesh networks where all
> participants are able to route using some dynamic routing protocol.
>
> Regarding the image, I don't get your point Willi. You can understand
> the libremesh.org graphic as a small piece of the Wikipedia graphic you
> attached. In any case it is just a cosmetic thing, the image is not
> trying to explain anything technical. You should go to "How it works" to
> get more into detail.
>
> Cheers.
>
> On 16/12/16 08:38, Paul Gardner-Stephen wrote:
>> Hello,
>>
>> For better or for worse, "mesh network" and "ad-hoc network" have become
>> almost synonyms, even though a mesh network correctly requires that every
>> node is connected to every other node.  The colloquial and common usage of
>> a mesh network in most cases includes ad-hoc networks, where connectivity
>> does not exist between every possible pair of nodes.
>>
>> Paul.
>>