| 26 Jan 2022 |
|  Anon10100101 joined the room. | 18:25:03 |
 albert | How does yggdrasil manage on top of mesh networks? | 20:17:51 |
| albert | Or does it require fixed underlying topology? | 20:18:28 |
 Arceliar | it is a mesh network | 20:18:52 |
| Arceliar | you could run it on top of another one, if it supports link-local multicast. batman-adv would probably work. i'm not sure it would help any to do that | 20:19:21 |
| albert | I'm thinking in terms of hardware and hardware-link topo that shifts arbitrarily. Maybe I'm misunderstanding the relationship between yggdrasil and the underlying hardware? | 20:30:32 |
| Arceliar | it can automatically discover peers and connect over link-local IPv6, which includes things like wifi running in ad-hoc mode | 20:32:25 |
 Evanito | Wonder what kind of LoRa implementations can be made to work with Ygg.
Meshtastic with the fastest configuration is able to get up to 18 kbps... probably not enough | 20:38:52 |
 mark22k | In reply to @Arceliar:matrix.org
it can automatically discover peers and connect over link-local IPv6, which includes things like wifi running in ad-hoc mode So you could form a network with Yggdrasil like Hyperboria actually wanted? | 20:39:27 |
| Arceliar | In reply to @evanito:matrix.org
Wonder what kind of LoRa implementations can be made to work with Ygg.
Meshtastic with the fastest configuration is able to get up to 18 kbps... probably not enough you could link 2 nodes with 18kbps if that's continuously available.... if it's in infrequent bursts at higher bandwidth, that average out to 18kbps, then that probably wouldn't work | 20:47:14 |
| Arceliar | In reply to @mark22k:tchncs.de
So you could form a network with Yggdrasil like Hyperboria actually wanted? that's the idea. or rather, figuring out how to do that is the idea, but it's an ongoing research project / experiment, so i can't claim that we're there yet | 20:47:48 |
| albert | It could work (theoretically) on a cellular hardware network? That's almost by definition a network with hardware topo that changes continuously. The hardware doesn't manage peer to peer, though, and my interest is in a wholly peer to peer driven hardware layer. | 20:54:56 |
| Evanito | If you find any such hardware, do share! I'm guessing that it doesn't exist at the minute though, and the gap has to be made up for in software :z | 20:56:38 |
| Evanito | I mentioned Meshtastic because it is a current P2P LoRa project that has an alpha LoRaWAN implementation, so it seems prime to try to fit with Ygg | 20:57:31 |
| Evanito | * I mentioned Meshtastic because it is a current P2P Mesh LoRa project that has an alpha LoRaWAN implementation, so it seems prime to try to fit with Ygg | 20:57:44 |
| Arceliar | actual cell phones are too locked down / not designed to directly communicate without going through a cell tower | 21:00:16 |
| Arceliar | other than short range things like bluetooth | 21:00:30 |
| Evanito | Ah yeah I was going to mention the regulatory issues any GOOD radio technology requires (curse you FCC) | 21:00:51 |
| Arceliar | if LTE had peer-to-peer connections as an option, then there's no technical reason why it shouldn't work. but that's not the world we live in | 21:01:00 |
| Evanito | I wonder if Wifi-6E will bring us closer to that world 🌥️ | 21:02:17 |
| albert | 80211s might be of interest: https://openwrt.org/docs/guide-user/network/wifi/mesh/80211s | 21:04:15 |
| albert | The docs say it is a Layer2 implementation and that Layer3 ought to work well (yggdrasil??) | 21:05:28 |
| albert | Expanding a node that operates on standard internet architecture (your wifi router of choice) to accept Meshtastic comms shouldn't be incredibly difficult. | 21:08:16 |
| albert | "Shouldn't be", lol, I'm an engineer, I should know better. | 21:08:49 |
| Arceliar | ygg over 802.11s should work. i have never tried it, so YMMV. i believe the toronto mesh people tested it, but hey had the mesh routing features in 802.11s disabled (they were just using it because it's better as a 1-hop protocol than adhoc mode, since you can do things like set a lower signal-to-noise threshold before a connection counts) | 21:10:32 |
| Evanito | Limiting low level connections to 1 hop could be a neat case for Ygg! | 21:44:04 |
 Tom | In reply to @albert:lathamco.net
It could work (theoretically) on a cellular hardware network? That's almost by definition a network with hardware topo that changes continuously. The hardware doesn't manage peer to peer, though, and my interest is in a wholly peer to peer driven hardware layer. some writing about this here;
https://read.cash/@TomZ/mesh-networking-with-yggdrasil-e300d18e
| 21:53:08 |
 neilalexander | In reply to @Arceliar:matrix.org
ygg over 802.11s should work. i have never tried it, so YMMV. i believe the toronto mesh people tested it, but hey had the mesh routing features in 802.11s disabled (they were just using it because it's better as a 1-hop protocol than adhoc mode, since you can do things like set a lower signal-to-noise threshold before a connection counts) I think if you disable the built-in 802.11s forwarding, which just gives you direct neighbour reachability, then it basically works the same as AWDL and that works very well | 22:23:58 |
| 27 Jan 2022 |
|  @xvczcio:libera.chat joined the room. | 00:10:22 |
| @xvczcio:libera.chat left the room. | 00:14:31 |
