#Meshtastic and #Meshcore are the gateway drug. #Reticulum (or something similar) should be the endgame.

If we're serious about network resilience and real independence from third-party infrastructure, not just being hobbyist cool, we need a technology-agnostic convergence layer rather than depending on LoRa for everything.

https://www.youtube.com/watch?v=XTnYVh7K6xQ

The Internet, Reinvented.

Unstoppable mesh.In this video, I build a Reticulum RNode and prove that completely different radios — LoRa and Wi-Fi — can communicate through a hardware-ag...

YouTube

@py2toz Well, what about plain old IP protocol?

It does not depend on the media (there are wifi, ethernet, fiber, microwave...) and it can do dynamic routing (we use it every day in the form of BGP, but others are possible).

Even the old HAM packet radio supports IP over AX25 over HF or VHF / UHF.

@marsik That's a fair point. But as far as I can tell, Reticulum goes a step further by implementing cryptography at the network layer by default. Not to mention the multiple encapsulation every time we traverse different interfaces (LoRa then IP then serial then LoRa again...)

So it's a matter of taste and use cases, as always (hence the "something similar" in my original post).

@py2toz Yep, it is a matter of taste and a specific use case.

Reticulum also uses encapsulation, how else would it go through all the various underlying protocols like LoRa or Wifi.

IP is rather heavy (even though lwIP exists) and the existing routing protocols are tailored towards faster links with higher MTUs.

But the cryptography address and no record of packet sender that Reticulum uses are interesting concepts for sure.

@marsik Sorry, I meant that IP would require different software independently implemented from each other for encapsulation, whereas Reticulum bridges a number of interfaces with encapsulation implemented within a single, opinionated framework.

Still a good fall back.

@py2toz Hmm, I am not sure I follow here.

Are you talking about the link layer (L2) protocols for the various transports that IP (L3) sits on top? There are of course a lot of them (starting with SLIP, through Ethernet and ending with complex Wifi and BT protocols).

But surely Reticulum (L3) must have that as well. And this list even says so and shows how to implement your own https://markqvist.github.io/Reticulum/manual/whatis.html#interface-types-and-devices

Data (L4) you transfer on top are link agnostic (per the ISO/OSI model).

What is Reticulum? - Reticulum Network Stack 1.1.3 documentation

@py2toz If the argument is that Reticulum actually implements all of the pieces in a single package, then yes, you are correct.

However many of those transports depend on the existing protocols anyway, so I honestly do not see the value in this specific aspect of Reticulum. Using standardized and well known link protocols is easy.

It is the routing where the main difference lies. And the crypto of course.