----- Forwarded message from Stephen Williams <sdw@lig.net> -----
On Sat, Feb 5, 2011 at 12:47 AM, Eugen Leitl <eugen@leitl.org> wrote:
...
Re Mesh - I have been mucking with it for many years and reading about it even longer. Seems solutions for getting what appears a group that Fun fact: a local /64 is fully enough to encode lat/long/alt (24, 24, 16 bit) with about 2 m resolution....
ORCHID address space is perfect for this type of thing, and there are particular extensions well suited to this. [0][1]
There are a couple main types of use cases for this kind of capability: A) emergency situations of various types like the Egypt blackout and ... There has been plenty of work with open source cell networks, etc. I haven't seen anything usable about hand-held mesh and/or store and forward capability yet.
this came up on the p2p-hackers list, which i can't seem to get mail through, [3] regarding Android in particular, sounds similar to Serval. [2] best regards, 0. An IPv6 Prefix for Overlay Routable Cryptographic Hash Identifiers (ORCHID) http://tools.ietf.org/html/rfc4843 [*this: just substitute the cryptographic with hashed coordinates or searchable prefixes, or both given privacy considerations.] 1. http://tools.ietf.org/html/draft-montenegro-6lowpan-dymo-low-routing-03 [*this: combine with geo-aware addressing.] 2. Serval Project http://www.servalproject.org/how-it-works 3. [p2p-hackers] What we should build for the Egyptian (and other) protesters http://lists.zooko.com/pipermail/p2p-hackers/2011-February/002829.html """ On Fri, Feb 4, 2011 at 10:54 AM, Alen Peacock <alenlpeacock@gmail.com> wrote:
I'd always hoped that a global ad-hoc wireless network would spring from something like MIT's RoofNet (http://pdos.csail.mit.edu/roofnet/doku.php).
There's still a lot of academic research into ad-hoc networks, but I'm not aware of anyone really pursuing something like this in the commercial space
not to any noteworthy degree. the attempts have been numerous in both commercial and research focus. given the number of ad-hoc routing protocols, developers capable of understanding the merits and detriments of general approaches and specific implementations, and certainly not least of all the difficulty in early-adoption of such a Metcalfe utility dependent "collaborative transport" is quite difficult. see http://en.wikipedia.org/wiki/List_of_ad_hoc_routing_protocols as for my druthers i am keenly interested in combined topologies and overlays leveraging the distinct advantages of particular transports together for a best effort, immediately accessible ad-hoc infrastructure. leased circuits and dedicated fiber continued to work in Egypt; a BGP4 peering blackout is just a small slice of the overall communication stacks and paths. (you can run E-Line/E-LAN over free space optics as easily as MPLS / WDM lambdas over fiber, multi-OFDMA wide band systems can be aggregated over polarizations and center bands, etc, etc. the possibilities are endless. and even packet over shortwave can carry signalling channels around the globe...) === On Fri, Feb 4, 2011 at 2:17 PM, David Barrett <dbarrett@quinthar.com> wrote:
... On this topic, it'd be great if you could simultaneously be connected to two networks (eg, a wireless DSL router *and* broadcast a mesh repeater). I don't think you can do this in the strict sense,
use madwifi-ng and a mobile IPv4/IPv6 transport. for example a custom SOCKS/HTTP/Transparent proxy to multi-homed SCTP over VPNs on virtual station madwifi-ng capable hardware. (i've used up to 6 clients per radio with success, and maximum 8 radios per PCI host bus.) lots of options in this space as well. just not cheap or easy yet... """
On Sat, 5 Feb 2011, coderman wrote:
Re Mesh - I have been mucking with it for many years and reading about it even longer. Seems solutions for getting what appears a group that Fun fact: a local /64 is fully enough to encode lat/long/alt (24, 24, 16 bit) with about 2 m resolution....
ORCHID address space is perfect for this type of thing, and there are particular extensions well suited to this. [0][1]
Why, in (egyption) situations like this, would you want location data embedded in your addressing scheme ? It seems like one would do hard work to ensure such encapsulation was NOT happening... Or is this just a scheme for fair and reasonable address allocation in an anarchy to avoid collisions ?
On Sun, Feb 6, 2011 at 3:44 PM, John Case <case@sdf.lonestar.org> wrote:
... Why, in (egyption) situations like this, would you want location data embedded in your addressing scheme ?
this is one of many potential addressing and routing optimizations for ad-hoc networks. for example, a dead-drop or multi-homed rendezvous point may not care about disclosing coarse geo coords, but clients sending to same may want much more stringent pseudonymous / anonymous protections. without context, this does seem counter intuitive for guerrilla comms. (locality constrained broadcast horizons are useful as well; anonymous wireless communication is indeed difficult; lots of problems and parts to play with :)
On Sun, Feb 06, 2011 at 11:44:39PM +0000, John Case wrote:
On Sat, 5 Feb 2011, coderman wrote:
Re Mesh - I have been mucking with it for many years and reading about it even longer. Seems solutions for getting what appears a group that Fun fact: a local /64 is fully enough to encode lat/long/alt (24, 24, 16 bit) with about 2 m resolution....
ORCHID address space is perfect for this type of thing, and there are particular extensions well suited to this. [0][1]
Why, in (egyption) situations like this, would you want location data embedded in your addressing scheme ?
Location is only available in point-to-point connections, of course you're free to run remixing layers on top of that infrastructure.
It seems like one would do hard work to ensure such encapsulation was NOT happening...
If you want to do cut-through at relativistic speed there are not many options left. If you want to obscure origins, use anonymization implemented in higher layers.
Or is this just a scheme for fair and reasonable address allocation in an anarchy to avoid collisions ?
That's one of the side effects: it cuts out the middlemen like IANA, RIRs and LIRs. Anyone can compute one's geographical positions either from global or local references, and you're free to refuse peering from position fakers. Faking disturbances takes ownership of the local loop, and is easily checked by referencing global beakons. -- Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org ______________________________________________________________ ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org 8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
participants (3)
-
coderman -
Eugen Leitl -
John Case