As Bob is dialing Alice, we now ensure that we are connected to Alice
at each step that needs communication.
If we are not connected, we proceed with dialing.
In an attempt to improve libp2p usage, we also add known address of
Alice first and only use peer_id to dial.
This ensures that we use the expected peer id.
This module was intended to contain helper functions for each step.
However, those are not needed except for the negotiate step.
A dedicated module is not needed for one function.
Previously the libp2p swarm had to be manually polled within the
protocol execution code to execute actions such as sending a
message. The swarm is now wrapped in SwarmDriver which polls the
swarm in a seperate task
Move state machine executors into seperate files
Remove check for ack message from Alice. Seems like a bad idea to
rely on an acknowledgement message instead of looking at the
blockchain.
Fix warnings
Consolidate and simplify swap execution. Generators are no longer
needed. Consolidate recovery and swap data structures. The
recursive calls can be replaced with a loop if returning prior to
completion is desired for testing purposes.
Fill out alice abort path
Move state machine executors into seperate files
Not compiling due to recursion/async issues
Fix async recursion compilation errors
Fix Bob swap execution
Remove check for ack message from Alice. Seems like a bad idea to
rely on an acknowledgement message instead of looking at the
blockchain.
Fix Bob abort
Fix warnings
Xmr lock complete
Add TxCancel submit to XmrLocked
Bob swap completed
Remove alice
The numerous tor conditional compile flags were removed by
extracting transport creation to the main statement. A tor
transport is created if Alice specifies a tor port using the CLI.
Before this patch Bob is not sending message 3. This is because we are not
polling Bob's swarm correctly. To fix it we can just mimic the other NB's and
bubble up an event when Bob receives message 3 response from Alice, this way we
can `await` upon this event which triggers polling, making Bob's swarm send the
message.
Also use cosntant backoff retry strategy as opposed to exponential
backoff. This is in case retrying several times quickly causes the
retry intervals to become large enough that the test is very slow
and/or the Bitcoin lock transaction expires.
The current problem occurs on the last message i.e. Bob sending
tx_redeem_encsig to Alice. The action is yielded for Bob to do it, but
Alice appears to never receive it (unconfirmed claim, requires more
logging).
Unfortunately, I had to put the wrap the swarm in Alice's `Network`
struct in an `Arc<Mutex<T>>` in order to be able to use `backoff` to
control the retry mechanism. This is because the stream of events
cannot be turned into a `SharedFuture` (unlike Bob's).
It would be good to find an alternative solution.
Alice does not respond with anything when receiving message 2 from Bob. We don't
want to leave Bob's request/response protocol waiting so send an empty response
back.
In order for Alice to complete the handshake she needs to transition to state 3,
for this she needs message 2 from Bob.
Send `bob::Message2` to Alice and transition to `State3` - completing the
handshake.
We model the getting of amounts as a network behaviour even though conceptually
it is a protocol. Refine/refactor the code a bit to make this more obvious.
- Use `Amounts` instead of `Messenger`
We only ever connect to a single peer, update peer tracker to reflect this. This
is a single patch because the handling of the two network behaviours is a
intertwined.
- Only track one peer connection
- Track the peer id and the multiaddr of the counterparty
- Emit an event for connection established on Alice's side as well as Bob's side
Add a binary crate `swap` that implements two nodes (Alice and Bob). With this
applied we can start up a node for each role and do:
- Bob: Requests current amounts using BTC is input
- Alice: Responds with amounts
- Bob: (mock) get user input to Ok the amounts
... continue with swap (TODO)
