Sergi Delgado Segura b200968da9 Removes snap catching until a way of doing it is found.
Currently it fails on restoring the cache due to trying to resto /snap
2020-03-18 16:26:52 +01:00
2020-03-18 16:26:48 +01:00
2020-03-17 15:47:38 +01:00
2020-03-18 12:09:58 +01:00
2020-03-17 15:47:38 +01:00
2020-03-18 12:09:58 +01:00
2020-03-18 16:26:52 +01:00
2020-03-18 16:26:48 +01:00

State of the code

Currently working on updating the software to match BOLT13 rev1.

The Eye of Satoshi (TEOS)

The Eye of Satoshi is a Lightning watchtower compliant with BOLT13, written in Python 3.

TEOS consists in three main modules:

  • teos: including the tower's main functionality (server-side)
  • cli: including a reference command line interface (client-side)
  • common: including shared functionality between teos and cli.

Additionally, tests for every module can be found at tests.

Running TEOS

In order to run teos you will need to create a configuration file (follow INSTALL.md) and run teosd.py. Currently you will also need a set of keys (to sign appointments) stored in your data directory.

You can use generate_keys.py to generate those keys.

Interacting with a TEOS instance

You can interact with a teos instance (either run by yourself or someone else) by using teos_cli under cli.

Since teos_cli works independently of teos, you will need a separate configuration file for this (follow cli/INSTALL.md).

teos_cli will also need an independent set of keys (that can be generated using generate_keys.py) as well as the public key of the tower instance (teos_pk.der). The same data directory can be used for both if you are running things locally.

Notice that teos_cli is a simple way to interact with teos, but ideally that should be part of your wallet functionality (therefore why they are independent entities). teos_cli can be used as an example for how to send data to a BOLT13 compliant watchtower.

Dependencies

Refer to DEPENDENCIES.md

Installation

Refer to INSTALL.md

Contributing

Refer to CONTRIBUTING.md

Description
No description provided
Readme 1.8 MiB
Languages
Python 99.7%
Shell 0.2%