This setup is fantastic! I say that because it's essentially what I and several others designed for many of the oceanogrpahic resesarch ships operated by Columbia Univ. Oregon State Univ. Scipps, WHOI, Univ. Hawii, U. of Wash. etc. to name a few. Our system is called SWAP (http://sssg1.whoi.edu/swap/) - "Ship to Ship/Ship to Shore Wireless Access Procool" and consists of Pebble Linux on Soektris boxes (or other similar) with PCMICA wireless cards. We predated OSLR and so used an open source routine called "aladin" to provision IP addresses to WDS links between units and OSPF to maintain routes between units. I'm really interested in what you all have done, because you've provided all the user-friendly bits that we haven't had time to develop.
I some questions:
1) Have you set up your system on nodes that are mobile? If so, how did it go?
2) How does OSLR perform when 1 or more links are of poor quality (i.e. the link comes up and drops frequently)? Are the OSLR daemons robust in this environment?
3) Suppose you have a meshed network similar to the one you've described in this article, but had more than 1 uplink to the wired world. How does OSLR know where to send the packets? Is it shortest hop? If so, what if the topolgoy changes? Do the packets get re-routed seamlessly?
Thanks for the article - very cool.