Update Summary


Previously, the co-op’s financials were made publicly available. Greg Meredith expressed the importance of Ravi Srivatsav’s role in bringing in equity; community members were more concerned about minimizing expenses.

This week, Kelly Foster assures the community that the dev team is committed to delivering the platform with resilience. They have a sense of obligation to investors thereby ensuring codes are performant, reliable, and secured. Greg informs the community that sensitive matters regarding co-op financials will be discussed in a separate closed session on Fridays.

Greg mentions four upcoming podcasts that address the tokenomics of proof of stake at macro scale, the concept of COMM event and transactions, technical discussions on execution models, and general operations and modifications of the co-op. Transcripts will be made available.

Ravi Srivatsav has been officially appointed as COO and board member; he will not be seeking compensation beyond the $1 signing bonus until more funds are secured. Miller Nash has taken a role as counsel. Greg and community continued conversations comparing the functionality of execution models (WASM,  Erlang, Rosette, Rholang) on the criteria of scalability behaviour. The meeting ended with Ed Eykholt’s question on validators.


Announcements



Content in Brief


Technical update: Kelly Foster discusses dev team commitments

Questions from the community regarding dev team commitments have been continual. Kelly Foster assures the community that the team has been resilient, focused, and committed to delivering the platform. They take the responsibility to RHOC and potential REV holders by developing code that is performant, reliable, and secure.

Sprint and Release

  • Two-week sprint ends March 11. Priorities: block store functionality, wallet and cost accounting implementation, and public test net infrastructure
  • RNode v0.8.4 will release. It was blocked earlier this week by  RCHAIN-3073 – Developer test net shows a deadlock in RSpace DONE . Kudos to the team for the dedication to resolve this bug! RNode-0.8.4 release plan

Mercury requirements and acceptance criteria

Test net status

  • A public test net is not yet available; the plan is to launch the week of March 25. Please see RChain public test net information to learn about the rollout plan, as well as an FAQ. At this time there are no known blockers to the delivery of Testnet-1
  • Tuesday’s test cancelled due to a critical bug
  • There will be a testing session on Thursday, March 7 at 15:00 UTC. We plan to demo how the block store allows a node to restart and catch up with the support of the block store and demo block finalization. Worksheet and details will be posted in #node-testing on RChain Discord
  • Upcoming changes to start time for community testing related to US move to daylight savings time (DST) on March 10. Starting with the test planned for March 12 the start time will be 16:00 UTC

Block storage

  • Work complete to resolve  RCHAIN-3074 – Restoring a node causes LatestMessagesLogIsMalformed error DONE. We can now demonstrate killing and restarting a node that catches up to the state of the network with the support of the block store. Many thanks to the team for their focused work on this! Next steps are to harden this feature through internal and community testing efforts

Consensus

  • Work in progress to improve performance of the process to prevent replay attacks. There is an open PR and Kent led a walkthrough of the change earlier today. RCHAIN-3070 – Limit deploy history duplication check depth TO DO
  • Work in progress to remove the BlockStore from the Estimator. PR in review.  RCHAIN-3069 – Make estimator return block hashes instead of full block TO DO
  • Cost accounting: Work in progress to implement common cost accounting across different layers of the platform: reducer, spatial matcher, and RSpace match component. Work described in Jira

Node

  • Collaboration with SRE and Whiteblock teams to support efforts to report data from Whiteblock test infrastructure

Test infrastructure and performance testing

  • Testnet support: Work in progress to finish tooling to launch and monitor the test net. View all issues in JIRA
  • Testing in Whiteblock: Collaboration in progress to get data out of Whiteblock for analysis and to support the operation of tests in Whiteblock to run in CI
  • Performance testing: Work in progress to develop a contract that best supports the measurement of COMM events/second and to establish a benchmark for performance. The goal for Mercury is 1,000-1,500 COMM events/second

Wallet

  • Work in progress to deliver the wallet feature. View all issues in JIRA

Business operations: Ravi Srivatsav signs on as COO; Miller Nash becomes counsel

After the technical update, Greg announces the signing of Ravi Srivatsav as COO. A community member recommended making a short introductory video about Ravi. Greg also informs the community of the board’s appointment of Miller Nash as counsel. They will address the board’s recommendation for creating an equity vehicle to incentivize investors.

Marketing update

Greg announces four upcoming podcasts to address the tokenomics associated with proof of stake at macro scale, the concept of COMM event and transactions, the idea behind technical decisions (such as the choice of execution model), and the general operations and modifications of the co-op (the shift in the management structure).

Comments/QA

  • Why wasn’t Erlang chosen as a smart contract language?
    • Greg: “The biggest reason is the correct by construction. Erlang does not have an independent operational semantics. The second point in this correct by construction argument is very important because if you want to do static analysis you need a static analysis that brings the concurrent considerations to the fore, you the concurrency to be reflected in the syntax and not hidden behind libraries or other aspects of the execution…”
  • Greg and community discuss execution models ; making comparison between models. (WASM, Rosette, Erlang, Rholang, LLVM)
    • He mentioned that execution models without concurrency by design will suffer.

Ed Eykholt’s questions about validators

  • What’s the number of validators that are committed?
  • At what point will anonymous validators be able to join the network?