This is a system for exposing an interface to a user’s lightning node to a web application. The seamless workflows for creating a secure, authenticated connection make use of DNS and certificate authorities. Bitcoin users should not need to register with these systems if there is another way.
Experience has shown that the main barrier to linking a user’s self-hosted service to an application is the size of the parameters for establishing the secure connection and having a secure way to give the application an authorization credential. LNC uses password authenticated key exchange so that users only need to furnish the web application with address parameters and a low-entropy password. Establishing the session includes sending an encrypted payload, and LNC uses this to give the application a macaroon.
LNC takes it a step further and integrates a message broker operated by lightning labs. This eliminates the requirement that the node be reachable on the internet in some way, at a cost of relying on a single service cluster for the entire ecosystem.
One cool thing about this is that it provides further examples of how to build out production systems using open standards like the noise protocol.
Apparently there is a Twitter discussion of how to think about splitting funds across several mints.
This project exposes a really interesting interface:
ln-sendPay an LN invoice with e-cash.
ln-receiveGenerate an invoice to redeem for e-cash upon payment.
From the outside, the mint looks like a lightning node. This interface also makes sense for users internally, instead of having a reminting operation.
The deposit mechanism addresses two requirements:
The scheme has users tweak a fixed descriptor, then provide the tweak to the federation along with the raw transaction creating the output and an inclusion proof.
The lightning payment flow requires the sender and receiver to be online at the same time. This post suggests a method for mitigating this.
Setup. Both sender and receiver have an LSP, which can store data to forward later.
There would need to be a mechanism for the CTV on the first HTLC to be large while subsequent hops can have standard timeouts.
The two main open source implementations of threshold ECDSA (one is maintained by Binance) used the vulnerable protocol. However, the authors disclosed their attack and the library developers hardened their code to get rid of the attack vector.
It looks like the attackers compromised Badger DAO’s CDN layer. They got access to an API key and spun up CloudFlare workers to edit requests on the way to users.