Can you explain with some level of detail how routing works on cjdns, Reticulum or Yggdrasil?
Discussion
No. Next question
I like your style. How are you doing today?
I'm having my fourth cup of coffee. Heading out to buy a baguette in a minute and make me a sandwich with leftover bbq chicken. I'm doing well.how about you, Scoundrel?
I got up bright and early to do something productive. Ugh, how irritating. It is nice to be back after Thanksgiving though. Enjoy your sandwich nostr:nprofile1qqsq0zknzdj3lhk7dds9v3dzrd3m87ytkeu6s6rpxd2gsne6ey56wqcpzpmhxue69uhkummnw3ezumt0d5hszrnhwden5te0dehhxtnvdakz7qgawaehxw309ahx7um5wghxy6t5vdhkjmn9wgh8xmmrd9skctcw8a4n7
So apparently on cjdns your node keeps track of routes to all other notes that have addresses close to yours -- even if that means they are actually very far away from you on the network (because addresses are just public keys). Then when you want to contact someone you find a route to the closest address of those you know and send your packet there, then the receiver has to find another route and so on -- which sounds terribly inefficient as these packets will travel through the entire globe many times before they reach the destination, and that doesn't make sense, so I am obviously missing something.
On Yggdrasil (by the way both cjdns and Yggdrasil supposedly do "source routing" and maybe it's even onion-routing which means the sender computes the route and sends routing information inside the packet) there is supposedly a spanning tree with a root node which to me doesn't make sense and then somehow something happens and you calculate a route?
Reticulum I have no idea whatsoever.
the problem with source routing is failures and congestions on paths that your router selects, and this is due to the inherent latency and anonymity issues of keeping the whole network state updated... especially the congestion issue
i've thought about a few ways to deal with that congestion stuff, and i know that tor solves a lot of it by requiring "telescoping" a connection ahead of using it
source routing sucks
the main reason why lightning has issues at times, when your node or your recipient node constructs a route that includes ded nodes in between
i've thought about ways to alleviate this problem with redundancy added to AMP but it's just inherent in source routing
for payments, it's a really hard problem to solve because path selection and the cryptography limit what you can do, but for network traffic aside from trust issues i don't see why it can't be done with path routing
tor uses source routing too but it adds a delay to the process by checking the path is working before handing it back to the proxy code as a working path, probably similar solutions might work for LN and CJDNS
I cannot.
CJDNS usa endereços criptografados e roteia via DHT.
Reticulum foca em redes resilientes e roteia pacotes independentes.
Yggdrasil cria uma topologia dinâmica e descentralizada com endereços criptográficos.
Hardest Arkanoid level