← All calls
ARBORIST#85

NU6 testnet activated, mining stopped

NU6 activated on testnet, but the miners disappeared and the chain stalled.

Agenda summary 9

  1. NU6 testnet activation results

    ECC, Zcash Foundation · Kris, Conrado

    • The NU6 testnet activation was successful, zcashd 5.10.0 was released, and it hits its end-of-service halt in early November.
    • Next is the zcashd 6.0.0 release, which sets the NU6 mainnet activation height.
    • At activation they ran a coordinated test of sending and receiving transactions, and Kris confirmed sending and receiving on testnet with the local zec-sqlite CLI.
    • The update to that crate will be published soon.
  2. Testnet mining stoppage and recovery

    ECC, Zcash Foundation · Kris, Alfredo, Dodger, Pacu, Arya, Conrado

    • The party supplying hash power to the testnet did not upgrade, ECC and ZF had turned off all their miners, and Zebra's internal miner is not in an official release, so the chain stalled.
    • With zcashd, a block interval that is too long requires human intervention, and they decided to wait for Str4d, who knows how to bypass the maximum elapsed time.
    • It was found that mining stopped when Michael Harms took down the faucet because of a zcashd build error.
    • ECC and ZF agreed to run minimal testnet mining infrastructure and to document the recovery procedure.
    • When Arya merged the branch with the miner enabled into main and brought up a Zebra instance, the testnet started moving again.
    • Michael Harms had been mining with zcashd Gen=1 on a T3 large instance.
  3. zcashd deprecation: replacement wallet architecture and progress tracking

    ECC, Zcash Foundation · Kris, Pacu, Conrado, Dodger

    • ECC and ZF held intensive discussions this week on the architecture of the components needed for the replacement wallet, and a plan was set.
    • They plan to move the sync code from the internal test CLI to the Zcash client backend and build a production CLI skeleton.
    • Because the ECC DAG and the ZF tracking issues are separate, Pacu is trying a GitHub project board that shows everything at a glance.
    • Conrado opened a PR moving DAG generation, which broke after a Zenhub API change, to the new API and showing ECC and ZF issues together.
    • Dodger said the architecture agreement on the end state must be documented so external contributors can take part, and the discussion is taking place in librustzcash issue 1373.
  4. Zingo Labs grant and how to handle external contributions

    ECC, Zcash Foundation, ZCG · Kris, Jason McGee, Dodger, Pacu

    • On the grant proposal Zingo Labs revised, Kris said the direction is right but it needs further refinement.
    • Jason McGee said they asked Zancas for prior approval from ECC and ZF but did not get it, and that a meeting of the three organizations could be held.
    • Dodger pointed out that bringing in outside organizations has an onboarding cost, and noted the tradeoff between strict specifications and outcome-based delegation.
    • Pacu said they need to survey the RPCs used by exchanges, explorers and wallets to set the indexer specification.
  5. ECC core team update

    ECC · Kris, Alfredo

    • They opened a draft PR moving Zcash Client SQLite's test infrastructure into the Zcash client backend.
    • The work is generalizing it for reuse by the Brave, Chainsafe and Keystone wallet teams.
    • This PR replaces a stopgap PR that Alfredo had opened.
    • They are fixing one remaining wallet sync bug, and also progressed on NU6 specification work and the release process.
  6. Zebra update: Zebra 1.9 / NU6 testnet and audit

    Zcash Foundation · Alfredo, Arya, Jason McGee

    • They finished the private NU6 testnet work with ECC, bumped ECC dependencies to the latest, and released Zebra 1.9.0 including the testnet activation height.
    • They fixed a panic edge case in state after activation and a Docker image problem.
    • To run zcashd's Python RPC tests against Zebra they added Regtest-only stop and generate RPCs, so merging the tests can begin.
    • The NU6 audit kickoff meeting took place within the past two weeks and they handed over the code to be audited.
    • There is also a change in which sendrawtransaction returns validation errors, and they asked Hanh to test it.
    • ZSF has no update this week, and the next update is on the 19th.
  7. Trailing Finality desirable properties document

    Shielded Labs · Zooko

    • They updated the properties document reflecting comments from Josh Swihart and Daira Emma and meetings with Chris Goes and Nate Wilcox.
    • They organized privacy in staking and whether a validator can steal delegated ZEC as the core security issues.
    • They leaned toward excluding even distribution of staked ZEC from the first deployment scope.
    • This is because even if one party holds a large stake the damage is at the level of a protocol halt, and it is also hard to prevent concentration through the protocol.
    • Shielded Labs does not receive the Dev Fund and operates on project-specific donations, so it places the most importance on time to launch.
  8. Crosslink implementation strategy and prototype stages

    Shielded Labs · Zooko, Alfredo

    • Whether to reuse existing designs or implementations or build its own is a strategic fork, and Cosmos Hub was mentioned as a design reference and Namada and Penumbra as Rust sources.
    • Stage 1 is a Crosslink in which the PoS component always fails, and stage 2 is a PoA prototype on a few servers.
    • Stage 3 is CometBFT with public keys and voting power hardcoded, and there is also a later option of putting weighted votes into the Zcash ledger.
    • When Alfredo asked about the relationship to Daira Emma's TFL book, Zooko answered that they will use that book's protocol (Crosslink v2) almost as is.
    • Alfredo suggested updating the book, and Zooko took note of it.
  9. FROST update

    Conrado

    • Progress is slow because of the NU6 work and vacations.
    • For the FROST server's cryptographic authentication, they are building a simplified PGP-like contact export and import feature.
    • They are finishing the key share refresh feature for DKG.
    • The randomizer generation method proposed by Daira Emma must be reflected in the ZIP.
    • FROST creates only a single spending key, so a way to generate a Zcash wallet's full keys must be added to the specification.
    • The transaction proposal format is likely to be worked on by the Keystone team and is also needed for FROST.

This summary is based on the original meeting notes; check them for exact details.