Community Update 54

General

• Sprint and next release

  • 2-week sprint started Jan. 7. Priority is platform stability, bug fixes, and feature specification.

  • RNode v0.8.2 released on Jan. 7. At a high level this release delivers bug fixes and optimizations contributing to bugs related to platform stability, expanded test coverage for Casper, and a new implementation of the metrics API to benefit debugging, reporting, and RNode operators. Please see RNode-0.8.2 release plan for more details.

• Mercury requirements and acceptance criteria

  • Ongoing effort to ensure completeness of feature-based user stories in the Mercury features doc (https://github.com/rchain/rchain/blob/dev/docs/features.md). We have tooling in place to show the status of the definition of the feature (Can we describe a test scenario to demo the feature?) and the status of the feature's completeness if the test is implemented. Demo of tool.

• Work with Whiteblock

  • Whiteblock team continues work to implement the RChain test plan into the testing environment (https://github.com/rchain/rchain/blob/dev/docs/whiteblock-test-plan.md). Expect to be able to see tests running in the environment today.

• Test net status

  • A public test net is not yet available. 

  • A view of the issues blocking launch of a public test net is available here.

  • Platform stability is priority #1 for most people on the team.

Rholang

• PR for the block store is in and is in review. Integration tests passing.  Work continues to improve checkpoint behavior for the blockstore. https://rchain.atlassian.net/browse/RCHAIN-652

Casper 

• Discussing Validator Incentives and Casper and implementation of cost accounting to support Mercury requirement work.

• Implementation of RhoSpec complete. This is a class that collects failures from Rholang tests and reports them as scalatest failures. Next step is to complete work so we can use this with TestResultCollector in Casper tests.

• Bug investigation that looked like an issue with validator bonding, and now appears to related either to the code or the algorithm. We have a unit test to recreate the issue and investigation is underway. Plan is to try a fix to the code: changing the assertion so that all blocks are the same instead of ensuring there is only one  block in the list.  https://rchain.atlassian.net/browse/RCHAIN-2805

Node 

• Completing work for stream API on client queue with retries. Last PR in review. https://rchain.atlassian.net/browse/RCHAIN-1904

• PR in review to stop receiving stream blobs if they don't fit in memory. https://rchain.atlassian.net/browse/RCHAIN-1615

• Updates to DAG grapher to group blocks by validator. This tool has been very helpful in debugging.

• Profiling of RNode in progress. https://rchain.atlassian.net/browse/RCHAIN-2765

SRE team 

• Priority is stability work and testing.

  • v2 of RDoctor used in this week's community test.

  • Improving delivery of data from the Rundeck testnet

  • Supporting Whiteblock team

  • Optimizations to integration test coverage

Developer website

• https://developer.rchain.coop 

• Work completed to update https://developer.rchain.coop with RNode v0.8.2 release information.