← All calls
ARBORIST#66

First mainnet block mined with Zebra

Marek shared the first case of mining a mainnet block with Zebra, and FROST 1.0.0-rc.0 was released.

Agenda summary 6

  1. ECC update: Rust library refactoring and decentralizing the Zcash GitHub organization

    Electric Coin Company · Str4d, Kris

    • Two tasks are underway in the core Zcash Rust libraries: separating the Sapling code into a standalone structure at the same level as the Orchard code, and applying the type safety developed in Orchard to Sapling as well.
    • Work is also underway to introduce Orchard support into zcash_client_backend and zcash_client_sqlite, the light client wallet components. The Sapling separation is planned to be attempted next week.
    • Kris said that because of Orchard support, the API on the current branch will be a moving target for 1-2 weeks. Published crates are fixed and unaffected.
    • After the Z.cash website, the next step is decentralizing responsibility for the Zcash GitHub organization and Zcashd. Participants will soon be recruited for a new group that will decide who is responsible for Zcashd and the Zcashd wallet after NU6.
  2. Zebra 1.5.0 and Zebra scanning

    Zcash Foundation · Teor

    • Work on shielded scanning in Zebra has started. It is still very experimental, and new crates and changes will appear in the release.
    • 1.5.0 is mostly a fix that avoids a panic when reading Sprout or history trees cached by earlier versions, plus network fixes responding to the Ziggurat red team report, and it comes out next week.
    • For the scanning data store, its own implementation is being used for now, with reference to the zcash_client_backend scanning foundation. Str4d proposed holding a knowledge-sharing session on the R&D Discord about separating wallet elements for use in a full-node environment.
  3. FROST 1.0.0-rc.0

    Zcash Foundation · Conrado

    • The first release candidate of the FROST crate, 1.0.0-rc.0, was published. Feedback from users and interested parties is being collected to stabilize the API, and the timing of the official 1.0.0 is undecided. Work on adding socket communication to the terminal demo is about half done.
    • Because the current structure assumes the signing coordinator is a server, the team is considering changing the architecture or releasing it as a graphical UI application. Work is also underway on no_std support for embedded environments.
    • The draft ZIP for using FROST in Zcash is complete but on hold awaiting Chelsea's security proof. The proof is ready and will be posted to ePrint soon, and the team intends to resume the ZIP process at ZIP sync.
  4. Trailing Finality Layer update

    Electric Coin Company · Nate

    • TFL book 0.1.0 was released about two weeks ago. The next milestone is to integrate the Crosslink documents on HackMD into the book and publish a comprehensive next release. The review of the safety argument with Daira is finished.
    • In Crosslink, the whole is considered safe if either one of the two subprotocols is safe, and the team judges that this resolves the concern about weaker-side security raised at Zcon4.
    • A prototype combining Zebra with CometBFT from the Cosmos ecosystem is being discussed. The goal is not to implement Crosslink but to gain experience in modifying Zebra to bind the two protocols together.
    • They said many questions, such as staking, yield and delegation approach, are still undecided and that they are far in the future.
  5. Year-end Arborist Call schedule and Zebra's first mainnet block

    Zcash Foundation · Teor, Marek

    • The next Arborist Call is the early slot on December 14, the December 28 call is canceled, and the first call of 2024 is January 11 at 15:00 UTC. Marek mined it personally, producing the first known Zebra mainnet block.
    • Marek connected an s-nomp mining pool to Zebra and routed ports through a VPS to connect ASICs located in Slovakia. Marek calculated odds of roughly one in 10,000 blocks (about 9 days), but it actually took about a month.
    • The miners are immersed in oil, with the heat used for home heating and hot water and the electricity offset by solar power. Marek plans to write a blog post over the Christmas holiday on the operating economics and setup.
  6. Mining luck, Zebra latency and sandblasting status

    Marek, Str4d, Teor

    • Marek asked whether the distribution of mining luck (deviation from expected time) can be estimated. Str4d answered that block finding is a Poisson process independent of the past, and that block propagation delay would affect smaller miners.
    • They discussed whether Zebra's getblocktemplate latency is better than Zcashd's. Teor confirmed that it responds immediately when a new block arrives and checks after 5 seconds only when the mempool changes.
    • Marek observed that the global hashrate fell from about 11 Gsol/s to 8 Gsol/s, raising ViaBTC's share.
    • Str4d explained that sandblasting has weakened, for instance with the introduction of ZIP-317, and that the mempool has been quiet since November 16. Str4d said it may start again, however.

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