yeah regular blockchain, with proof of relay (data transmit) rather than PoW. it is limited to ipv4 though. no it cannot be run on isolated networks because every node needs to be able to be reachable by a static IP.
thanks for mentioning yggdrasil, looks interesting.
This is not a 'message passing framework.' It's a program to establish direct TCP connections between peers. Existing solutions are largely based on mesh networks where messages reach nodes via one or more links. As you can imagine, this approach isn't suitable for all use-cases. Imagine a game server where you want your friends to connect to it. They should be able to make a direct TCP connection to the server with the lowest latency possible.
Currently, its hard to run servers on regular home networks where uses are behind NATs and firewalls. The goal of this project is to facilitate direct TCP connections. The software implements many improvements that are only in computer science papers at the moment. In the process of implementing such papers over 20k lines of code have been written, covering everything from NAT enumeration to automated port forwarding. Neither of the links you posted solve all these problems.
The blog goes into a higher overview of the project if you read it. Cheers.
yet another message passing framework? why reinvent the wheel? also, no need for all those protocols like STUN, MQTT etc. if you do it the right way.
here are some examples you can take inspiration from:
NKN (New Kind of Network, ~60k nodes in operation globally) [0] or something like Kameo [1]
[0] https://nkn.org/
[1] https://docs.page/tqwewe/kameo
Is NKN just a regular old Blockchain that doesn't work on isolated networks, or do they do something special to fix that?
Yggdrasil seems to be the best of this kind of thing, but it still has some issues
yeah regular blockchain, with proof of relay (data transmit) rather than PoW. it is limited to ipv4 though. no it cannot be run on isolated networks because every node needs to be able to be reachable by a static IP.
thanks for mentioning yggdrasil, looks interesting.
Hello Omani,
This is not a 'message passing framework.' It's a program to establish direct TCP connections between peers. Existing solutions are largely based on mesh networks where messages reach nodes via one or more links. As you can imagine, this approach isn't suitable for all use-cases. Imagine a game server where you want your friends to connect to it. They should be able to make a direct TCP connection to the server with the lowest latency possible.
Currently, its hard to run servers on regular home networks where uses are behind NATs and firewalls. The goal of this project is to facilitate direct TCP connections. The software implements many improvements that are only in computer science papers at the moment. In the process of implementing such papers over 20k lines of code have been written, covering everything from NAT enumeration to automated port forwarding. Neither of the links you posted solve all these problems.
The blog goes into a higher overview of the project if you read it. Cheers.