ava-labs / avalanche-rs Goto Github PK

protoc-gen-prost-crate does support creating features for each proto automatically.

While researching the rust implementation, we noticed there’s a check in get_validator_set if reps.validators is empty in order to get the publick_key variable updated. But if resp.validators is empty, wouldn’t it skip the loop in the first place? We are assuming there it should check if validator.public_key is empty instead, because it seems it’s returned as a bytes array, and if the validator does not have a public key, the array would be empty. Are our assumptions correct or are we missing something?

Right now, one has to create an Rpc trait, define all the methods, then define the implementation. It would be way nicer if we could just have one impl block that both defined the trait as well as the implementation. This should be pretty easy to accomplish with our own proc-macro.

TODO:
write example of desired output

Building docs via cargo doc works just fine, but does not work on docs.rs.

See https://docs.rs/crate/avalanche-types/0.1.1/builds/899151 for the build error.

The linter just fails with clippy and fmt not installed, but then shows up as passing.

It would be much better if the commands were specified directly in the jobs and if they were made simpler.

For instance, we should only be running cargo clippy --all --all-features --examples --tests -- -Dwarnings. If anyone wants to see what's failing, it's easy for them to checkout the .github directory and see the commands all in one place without being lead to separate files.

And, in order to check specific rules, those rules can be specified in lib.rs for each individual crate. For example, just place !#[deny(clippy::pedantic)] at the top of lib.rs instead of calling cargo clippy -- -D clippy::pedantic

jsonrpc-http-server is no longer maintained (ref).

Same maintainers built jsonrpsee;

From my understanding (and comment)fxIDshould not be (de)serialized here:

See https://github.com/ava-labs/avalanche-rs/actions/runs/6040807095/job/16392431819 for an example flake.

Compiling avalanchego-conformance v0.0.0 (/home/runner/work/avalanche-rs/avalanche-rs/tests/avalanchego-conformance)
    Finished test [unoptimized + debuginfo] target(s) in 2m 36s
     Running unittests src/lib.rs (target/debug/deps/avalanchego_conformance-70b5c63bbd0deef4)

running 8 tests
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::key::bls::generate_bls_signature' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
test tests::key::bls::generate_bls_signature ... FAILED
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::key::certificates::load_certificate_to_node_id' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::key::certificates::load_certificate_to_node_id ... FAILED
thread 'tests::key::certificates::generate_certificate_to_node_id' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::key::certificates::generate_certificate_to_node_id ... FAILED
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::key::secp256k1::generate' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::key::secp256k1::generate ... FAILED
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::key::secp256k1::recover_hash_public_key' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::key::secp256k1::recover_hash_public_key ... FAILED
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::key::secp256k1::load' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
test tests::key::secp256k1::load ... FAILED
thread 'tests::packer::build_vertex::build_vertex' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::packer::build_vertex::build_vertex ... FAILED
[2023-08-31T18:03:21Z INFO  avalanchego_conformance_sdk] creating a new client with http://127.0.0.1:22342/
thread 'tests::ping' panicked at 'called `Result::unwrap()` on an `Err` value: tonic::transport::Error(Transport, hyper::Error(Connect, ConnectError("tcp connect error", Os { code: 111, kind: ConnectionRefused, message: "Connection refused" })))', /home/runner/work/avalanche-rs/avalanche-rs/avalanchego-conformance-sdk/src/lib.rs:54:72
test tests::ping ... FAILED

successes:

successes:

failures:

failures:
    tests::key::bls::generate_bls_signature
    tests::key::certificates::generate_certificate_to_node_id
    tests::key::certificates::load_certificate_to_node_id
    tests::key::secp256k1::generate
    tests::key::secp256k1::load
    tests::key::secp256k1::recover_hash_public_key
    tests::packer::build_vertex::build_vertex
    tests::ping

test result: FAILED. 0 passed; 8 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

error: test failed, to rerun pass `-p avalanchego-conformance --lib`

Here's the migration guide
https://hyper.rs/guides/1/upgrading/

Currently we use tcp as the network for the rpcchainvm as a default. This issue tracks adding unix abstract socket support to the Rust VM SDK.

ref.
[1] tokio-rs/tokio#4610
[2] ava-labs/avalanchego#1949

See @marun's PR as an example.

There are currently no public-facing documents that indicate the roadmap for this project. For the purposes of the community, there should be a clear list of issues and milestones that the project hopes to achieve. This way, issues can be picked up by external contributors and the project can continue to grow.

Transferring ava-labs/avalanche-types-rs#107 over to this new repository.

@Nuttymoon if you'd like to follow up regarding message signing, please do so in this new issue.

See https://github.com/neoeinstein/protoc-gen-prost/tags for an example of a workspace with multiple published crates. Each tag is prefixed by the name of the crate being released.

Requires updates to CI automation.

There seem to be some inconsistencies with Go and Rust with regards to JSON serialization.

AvalancheGo serializes a tx to JSON json.Marshal(tx) that looks like this:

{
  "unsignedTx": {
    "networkID": 1337,
    "blockchainID": "11111111111111111111111111111111LpoYY",
    "outputs": [
      {
        "assetID": "28VTWcsTZ55draGkmjdcS9CFFv4zC3PbvVjkyoqxzNC7Y5msRP",
        "fxID": "spdxUxVJQbX85MGxMHbKw1sHxMnSqJ3QBzDyDYEP3h6TLuxqQ",
        "output": {
          "addresses": ["P-custom18jma8ppw3nhx5r4ap8clazz0dps7rv5u9xde7p"],
          "amount": 19999999899000000,
          "locktime": 0,
          "threshold": 1
        }
      }
    ],
    "inputs": [
      {
        "txID": "CjisYCwF4j7zSyC25MWR21e5xxzJgwdaLEuf7oBXSGpe3oaej",
        "outputIndex": 0,
        "assetID": "28VTWcsTZ55draGkmjdcS9CFFv4zC3PbvVjkyoqxzNC7Y5msRP",
        "fxID": "spdxUxVJQbX85MGxMHbKw1sHxMnSqJ3QBzDyDYEP3h6TLuxqQ",
        "input": { "amount": 19999999900000000, "signatureIndices": [0] }
      }
    ],
    "memo": "0x",
    "validator": {
      "nodeID": "NodeID-JHD1JDUZYkiMdWPMDA9UJhWXR4wf25t39",
      "start": 1711142673,
      "end": 1711142913,
      "weight": 20,
      "subnetID": "CjisYCwF4j7zSyC25MWR21e5xxzJgwdaLEuf7oBXSGpe3oaej"
    },
    "subnetAuthorization": { "signatureIndices": [0] }
  },
  "credentials": [
    {
      "signatures": [
        "0x3265999bb1a3390f09ca28a75e293faeb2f5670d26d9e99dfef9db7ccf08ccbb35f8fe144e63c91f24d0eca57d9db9d8cf7c223362e58b89eb0d317d4c35ac1c00"
      ]
    },
    {
      "signatures": [
        "0x0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"
      ]
    }
  ],
  "id": "2EhnpHqoBEJYFmfBiBmvoykWLFvxAjMXuuLGjvSCuCWvMaHbo1"
}

Avalanchego generates these bytes for the above: 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

But the only structure I can come up with that this Rust SDK will deserialize looks like this:

{
  "base_tx": {
    "networkID": 1337,
    "blockchainID": "11111111111111111111111111111111LpoYY",
    "outputs": [
      {
        "assetID": "28VTWcsTZ55draGkmjdcS9CFFv4zC3PbvVjkyoqxzNC7Y5msRP",
        "fxID": "spdxUxVJQbX85MGxMHbKw1sHxMnSqJ3QBzDyDYEP3h6TLuxqQ",
        "output": {
          "addresses": ["P-custom18jma8ppw3nhx5r4ap8clazz0dps7rv5u9xde7p"],
          "amount": 19999999899000000,
          "locktime": 0,
          "threshold": 1
        }
      }
    ],
    "inputs": [
      {
        "txID": "2c1CbR7FGYdeFPB4WaeWphHZrChLH7TQ92FRW6U4mCWTnaxVsB",
        "outputIndex": 0,
        "assetID": "28VTWcsTZ55draGkmjdcS9CFFv4zC3PbvVjkyoqxzNC7Y5msRP",
        "fxID": "spdxUxVJQbX85MGxMHbKw1sHxMnSqJ3QBzDyDYEP3h6TLuxqQ",
        "input": {
          "amount": 19999999900000000,
          "signatureIndices": [0]
        }
      }
    ],
    "memo": "0x"
  },
  "validator": {
    "validator": {
      "node_id": "NodeID-111111111111111111116DBWJs",
      "start": 1710879913,
      "end": 1710966313,
      "weight": 20
    },
    "subnet_id": "2c1CbR7FGYdeFPB4WaeWphHZrChLH7TQ92FRW6U4mCWTnaxVsB"
  },
  "subnet_auth": { "sig_indices": [0] },
  "creds": []
}

The latest avalanchego release https://github.com/ava-labs/avalanchego/releases/tag/v1.10.13 bumped the protocol plugin version to 29. The client tooling here needs to be updated as well to match.

To improve UX for new Rust VM developers we could create cargo plugin that would initialize a new vm template, include examples etc.

cargo subnet new myvm

ref. https://www.youtube.com/watch?v=cUB4PXPKrnU

The latest avalanchego release https://github.com/ava-labs/avalanchego/releases/tag/v1.11.0-fuji bumped the protocol plugin version to 32. The client tooling here needs to be updated as well to match.

Port the Mempool type and API from avalanchego. See https://pkg.go.dev/github.com/ava-labs/[email protected] for package-level documentation.

Currently, cert-manager is maintained as a personal repo of @gyuho. It would be much easier to just copy the single module directly into this repo (with proper credit).

Right now, it's really hard to see how long a job like unit-tests takes for a normal workflow. There are always multiple workflows that are started per PR. It would be a lot nicer to have a single workflow with multiple jobs.

We should look into what's causing the tests to take so long, see if there's any way to speed them up or split them up.

We should only be running tests on a sub-crate that actually had changes applied, that would probably get us a nice speed up right off the bat.

Expected outcomes to close this issue:

• Identify slow tests and create individual issues specifying that they need to be sped up (probably one issue per slow test)
• Split up sub-crate tests based on diff.