lightning-bolts/01-messaging.md

# BOLT #1: Base Protocol

## Overview
This protocol assumes an underlying authenticated and ordered transport mechanism that takes care of framing individual messages.
[BOLT #8](08-transport.md) specifies the canonical transport layer used in Lightning, though it can be replaced by any transport that fulfills the above guarantees.

The default TCP port is 9735. This corresponds to hexadecimal `0x2607`: the Unicode code point for LIGHTNING.<sup>[1](#reference-1)</sup>

All data fields are big-endian unless otherwise specified.

## Table of Contents
  * [Connection Handling and Multiplexing](#connection-handling-and-multiplexing)
  * [Lightning Message Format](#lightning-message-format)
  * [Setup Messages](#setup-messages)
    * [The `init` Message](#the-init-message)
    * [The `error` Message](#the-error-message)
  * [Control Messages](#control-messages)
    * [The `ping` and `pong` Messages](#the-ping-and-pong-messages)
  * [Acknowledgments](#acknowledgments)
  * [References](#references)
  * [Authors](#authors)

## Connection Handling and Multiplexing

Implementations MUST use a single connection per peer — channel messages (which include a channel ID) are multiplexed over this single connection.

## Lightning Message Format

After decryption, all Lightning messages are of the form:

1. `type`: 2 byte big-endian field indicating the type of message.
2. `payload`: variable length payload which comprises the remainder of
   the message and conforms to the format matching the `type`.

The `type` field indicates how to interpret the `payload` field.
The format for each individual type is specified in a specification in this repository.
The type follows the _it's ok to be odd_ rule, so nodes MAY send _odd_-numbered types without ascertaining that the recipient understands it.

A sending node:
  - MUST NOT send an evenly-typed message not listed here, without prior negotiation.

A receiving node:
  - upon receiving a message of _odd_, unknown type:
    - MUST ignore the received message.
  - upon receiving a message of _even_, unknown type:
    - MUST fail the channels.

The messages are grouped logically into 4 groups, ordered by their most significant set bit:

  - Setup & Control (types `0`-`31`): messages related to connection setup, control, supported features, and error reporting. These are described below.
  - Channel (types `32`-`127`): messages used to setup and tear down micropayment channels. These are described in [BOLT #2](02-peer-protocol.md).
  - Commitment (types `128`-`255`): messages related to updating the current commitment transaction, which includes adding, revoking, and settling HTLCs, as well as updating fees and exchanging signatures. These are described in [BOLT #2](02-peer-protocol.md).
  - Routing (types `256`-`511`): node and channel announcements, as well as any active route exploration. These are described in [BOLT #7](07-routing-gossip.md).

The size of the message is required to fit into a 2 byte unsigned int by the transport layer; therefore, the maximum possible size is 65535 bytes.

A receiving node:
  - MUST ignore any additional data within a message, beyond the length it expects for that type.
  - upon receiving a known message with insufficient length for the contents:
    - MUST fail the channels.

### Rationale

The standard endian of `SHA2` and the encoding of Bitcoin public keys
are big endian, thus it would be unusual to use a different endian for
other fields.

Length is limited to 65535 bytes by the cryptographic wrapping, and
messages in the protocol are never more than that length anyway.

The _it's ok to be odd_ rule allows for future optional extensions
without negotiation or special coding in clients. The "ignore
additional data" rule similarly allows for future expansion.

Implementations may prefer to have message data aligned on an 8 byte
boundary (the largest natural alignment requirement of any type here);
however, adding a 6 byte padding after the type field was considered
wasteful: alignment may be achieved by decrypting the message into
a buffer with 6 bytes of pre-padding.

## Setup Messages

### The `init` Message
Once authentication is complete, the first message reveals the features supported or required by this node, even if this is a reconnection.

[BOLT #9](09-features.md) specifies lists of global and local features. Each feature is generally represented in `globalfeatures` or `localfeatures` by 2 bits. The least-significant bit is numbered 0, which is _even_, and the next most significant bit is numbered 1, which is _odd_.

Both fields `globalfeatures` and `localfeatures` MUST be padded to bytes with zeros.

1. type: 16 (`init`)
2. data:
   * [`2`:`gflen`]
   * [`gflen`:`globalfeatures`]
   * [`2`:`lflen`]
   * [`lflen`:`localfeatures`]

The 2 byte `gflen` and `lflen` fields indicate the number of bytes in the immediately following field.

#### Requirements

The sending node:
  - MUST send `init` as the first Lightning message for any connection.
  - MUST set feature bits as defined in [BOLT #9](09-features.md).
  - MUST set to zero any feature bits that are not defined.
  - SHOULD use the minimum lengths required to represent the feature fields.

The receiving node:
  - MUST wait to receive `init` before sending any other messages.
  - MUST respond to known feature bits as specified in [BOLT #9](09-features.md).
  - upon receiving _odd_ feature bits which are non-zero:
    - MUST ignore the bit.
  - upon receiving _even_ feature bits which are non-zero:
    - MUST fail the connection.

#### Rationale

This semantic allows both future incompatible changes and future backward compatible changes. Bits should generally be assigned in pairs, in order that optional features may later become compulsory.

Nodes wait for receipt of the other's features to simplify error
diagnosis, where features are incompatible.

The feature masks are split into local features (which only affect the
protocol between these two nodes) and global features (which can affect
HTLCs) and are thus also advertised to other nodes.

### The `error` Message

For simplicity of diagnosis, it's often useful to tell a peer that something is incorrect.

1. type: 17 (`error`)
2. data:
   * [`32`:`channel_id`]
   * [`2`:`len`]
   * [`len`:`data`]

The 2-byte `len` field indicates the number of bytes in the immediately following field.

#### Requirements

The channel is referred to by `channel_id`, unless `channel_id` is zero (i.e. all bytes are zero), in which case it refers to all channels.

The funding node:
  - for all error messages sent before (and including) the `funding_created` message:
    - MUST use `temporary_channel_id` in lieu of `channel_id`.

The fundee node:
  - for all error messages sent before (and not including) the `funding_signed` message:
    - MUST use `temporary_channel_id` in lieu of `channel_id`.

A sending node:
  - when sending `error`:
    - MUST fail the channel referred to by the error message.
  - SHOULD send `error` for protocol violations or internal errors which make channels unusable or further communication unusable.
  - MAY send an empty `data` field.
  - when failure was caused by an invalid signature check:
    - SHOULD include the raw, hex-encoded transaction in reply to a `funding_created`, `funding_signed`, `closing_signed`, or `commitment_signed` message.
  - when `channel_id` is zero:
    - MUST fail all channels,
    - and MUST close the connection.
  - MUST set `len` equal to the length of `data`.

The receiving node:
  - upon receiving `error`:
    - MUST fail the channel referred to by the error message.
  - if no existing channel is referred to by the message:
    - MUST ignore the message.
  - MUST truncate `len` to the remainder of the packet (if it's larger).
  - if the string is composed solely of printable ASCII characters (For reference: the printable character set includes byte values 32 through 127, inclusive):
    - SHOULD only print out `data` verbatim.

#### Rationale

There are unrecoverable errors which require an abort of conversations;
if the connection is simply dropped, then the peer may retry the
connection. It's also useful to describe protocol violations for
diagnosis, as this indicates that one peer has a bug.

It may be wise not to distinguish errors in production settings, lest
it leak information — hence, the optional `data` field.

## Control Messages

### The `ping` and `pong` Messages

In order to allow for the existence of very long-lived TCP connections, at
times it may be required that both ends keep alive the TCP connection at the
application level. Such messages also allow obfuscation of traffic patterns.

1. type: 18 (`ping`)
2. data:
    * [`2`:`num_pong_bytes`]
    * [`2`:`byteslen`]
    * [`byteslen`:`ignored`]

The `pong` message is to be sent whenever a `ping` message is received. It
serves as a reply and also serves to keep the connection alive, while
explicitly notifying the other end that the receiver is still active. Within
the received `ping` message, the sender will specify the number of bytes to be
included within the data payload of the `pong` message.

1. type: 19 (`pong`)
2. data:
    * [`2`:`byteslen`]
    * [`byteslen`:`ignored`]

#### Requirements

A node sending a `ping` message:
  - SHOULD set `ignored` to zeros,
  - but MUST NOT set `ignored` to sensitive data such as secrets or portions of initialized
memory.
  - if it doesn't receive a corresponding `pong`:
    - MAY terminate the network connection,
    - and MUST NOT fail the channels in this case.
  - SHOULD NOT send `ping` messages more often than once every 30 seconds.

A node sending a `pong` message:
  - SHOULD set `ignored` to zeros,
  - but MUST NOT set `ignored` to sensitive data such as secrets or portions of initialized
 memory.

A node receiving a `ping` message:
  - SHOULD fail the channels if it has received significantly in excess of one `ping` per 30 seconds.
  - if `num_pong_bytes` is less than 65532:
    - MUST respond by sending a `pong` message, with `byteslen` equal to `num_pong_bytes`.
  - otherwise (`num_pong_bytes` is **not** less than 65532):
    - it MUST ignore the `ping`.

A node receiving a `pong` message:
  - if `byteslen` does not correspond to any `ping`'s `num_pong_bytes` value it has sent:
    - MAY fail the channels.

### Rationale

The largest possible message is 65535 bytes; thus, the maximum sensible `byteslen`
is 65531 — in order to account for the type field (`pong`) and the `byteslen` itself. This allows
a convenient cutoff for `num_pong_bytes` to indicate that no reply should be sent.

Connections between nodes within the network may be very long lived, as payment
channels have an indefinite lifetime. However, it's likely that for a
significant portion of the lifetime of a connection, no new data will be
exchanged. Also, on several platforms it's possible that Lightning
clients will be put to sleep without prior warning. Hence, we use a
distinct `ping` message, in order to probe for the liveness of the connection on
the other side, as well as to keep the established connection active.

Additionally, the ability for a sender to request that the receiver send a
response with a particular number of bytes enables nodes on the network to
create _synthetic_ traffic. Such traffic can be used to partially defend
against packet and timing analysis — as nodes can fake the traffic patterns of
typical exchanges, without applying any true updates to their respective
channels.

When combined with the onion routing protocol defined in
[BOLT #4](https://github.com/lightningnetwork/lightning-rfc/blob/master/04-onion-routing.md),
careful statistically driven synthetic traffic can serve to further bolster the
privacy of participants within the network.

Limited precautions are recommended against `ping` flooding, however some
latitude is given because of network delays. Note that there are other methods
of incoming traffic flooding (e.g. sending _odd_ unknown message types, or padding
every message maximally).

Finally, the usage of periodic `ping` messages serves to promote frequent key
rotations as specified within [BOLT #8](https://github.com/lightningnetwork/lightning-rfc/blob/master/08-transport.md).

## Acknowledgments

[ TODO: (roasbeef); fin ]

## References
1. <a id="reference-2">http://www.unicode.org/charts/PDF/U2600.pdf</a>

## Authors

[ FIXME: Insert Author List ]

![Creative Commons License](https://i.creativecommons.org/l/by/4.0/88x31.png "License CC-BY")
<br>
This work is licensed under a [Creative Commons Attribution 4.0 International License](http://creativecommons.org/licenses/by/4.0/).