← All calls
ARBORIST#88

ZSF renamed NSM, NU7 candidate

ZSF was renamed NSM, and there was an intention to put it forward as an NU7 candidate.

Agenda summary 12

  1. NU6 update

    ECC, Zcash Foundation · Kris, Marek

    • zcashd 6.0.0, which supports NU6, was released, and everyone must upgrade.
    • The previous release series will hit its end-of-service halt around November 5, and NU6 mainnet activation is around November 23.
    • A release candidate for Zebra 2.0.0 (v2.0.0-rc.0) was published.
  2. Zebra update: Zebra 2.0 / security audit

    Zcash Foundation · Marek, Pili

    • They resolved the testnet problems, and PRs are in progress for orphan mempool transaction validation, nullifier tracking, cookie-based RPC authentication, and Dockerfile and script refactoring.
    • Least Authority's NU6 audit produced two very minor comments about documentation and panics, which were filed as issues.
    • They plan to fix them within a few days, get re-verification and then publish the final report.
  3. ECC update: PCZT and Zashi delay

    ECC · Kris

    • Str4d made a lot of progress on PCZT, the partially constructed Zcash transaction format.
    • The format takes into account FROST signing, hardware wallets and cases where several people build a transaction together, and there is a draft PR in librustzcash.
    • In Zashi they implemented minimal note management that holds notes in a mixed state so that consecutive transactions are possible.
    • This can fix the tip the barista problem.
    • The release is delayed because Apple App Store review is blocked by the Coinbase integration, and the time of resolution is uncertain.
  4. zcashd deprecation progress

    Zingo Labs, Zcash Foundation · Pacu, Arlo, Marek

    • Pacu contacted block explorers and began collecting RPC requirements for moving their services to Zebra.
    • Kris is working out with ChainSafe how to bring the in-memory wallet work they contribute upstream.
    • Arlo plans to open a small PR to Zebra early next week that adds getters to private structs so Zebra primitives can be used as the Zaino backend.
    • Zaino will finish a test framework that compares behavior with lightwalletd around next week, and over the following weeks will finish implementing lightwalletd features and integrate with the Zebra read state service.
    • Marek said a PR for ZIP 233, which covers the network sustainability mechanism, is up and they will start reviewing it soon.
  5. Zcash Shielded Assets (ZSA) update

    QEDIT · Vivek

    • For asset swaps, they reflected improvements found during implementation in the ZIP, and put the first draft of the Orchard crate changes needed for swaps up for review.
    • The Orchard and librustzcash work for the ZSA protocol and the v6 transaction format is done.
    • Because asset descriptions are specified only as needing to be UTF-8, they stopped using a string type.
    • In Zebra the main work is consensus rule validation and updating global state for ZSA, and after a meeting with ZF they are splitting the work into small PRs and starting review.
    • The transaction format ZIPs 226, 227 and 230 are ready for ZIP editor review, and revisions to the ZIP 317 fee and the ZIP 315 recommended wallet behavior are in progress.
    • They plan to share a design document for the trade acceptance feature around next week.
  6. FROST update

    Conrado

    • They added cryptographic authentication, the last piece of the FROST server.
    • Work remains to bring it to production level.
    • They plan to release FROST 2.0.0 soon, while reviewing contributor PRs.
    • They resolved a small problem with the key share refresh feature when using DKG.
  7. ZSF renamed NSM, and the ZIP overhaul

    Shielded Labs · Jason McGee, Mariusz

    • To avoid being mistaken for a fund, ZSF was renamed Network Sustainability Mechanism (NSM).
    • Instead of deposit and distribution, the terms burn and mint are used, and the 21 million cap is maintained.
    • ZIP 233 became a voluntary burn mechanism, and the ZSF deposit field was replaced with burn amount.
    • ZIP 234 defines the money reserve as max money minus chain value, and ZIP 235 specifies burning 60% of transaction fees.
    • All three ZIP implementation PRs for zcashd are up, Zebra has the ZIP 233 PR up, and 234 and 235 are to be posted within a few days.
    • They plan to propose it as an NU7 candidate.
  8. NSM terminology and consensus rule concerns, and NU7 priorities

    Daira Emma, Kris, Josh, Jason McGee, Zooko

    • Daira Emma, Kris and Str4d worried that burn implies a permanent reduction of the money supply and causes misunderstanding.
    • It was also pointed out that the block balance rule in ZIP 236 conflicts with the burn amount field, so a separate meeting between the ZIP editors and Jason McGee was set.
    • Daira Emma said that if ZSA is included in NU7, the complexity budget is close to its limit, and the community's wishes must also be checked.
    • Jason McGee said they would hold a vote after the November ZIP selection process.
    • Josh asked why implement in zcashd when NU7 centers on Zebra, and why it is a priority now.
    • Daira Emma answered that the window for getting things into zcashd is already closing.
  9. Crosslink 2 security proof improvements and first deployment preparation

    Shielded Labs · Daira Emma, Zooko, Nate

    • Daira Emma said they improved the Crosslink 2 security proof so that confirmation latency can be halved at the same security.
    • Previously, with a confirmation count of sigma the finality latency was 2 sigma + 1 blocks, but it is now possible to reach sigma blocks plus only the BFT overhead.
    • Zooko said they will write up a short requirements and tradeoffs document, post it on the Shielded Labs GitHub and take comments until October 31.
    • Nate, who recently joined, is responsible for fleshing out and deploying the Crosslink design, is hiring a small engineering team, and intends to start a prototype before the design is complete.
    • Nate will also work out how to separate the TFL book repository's roadmap from their own plan.
  10. Crosslink staged deployment and economic security

    Shielded Labs · Daira Emma, Nate, Josh

    • Daira Emma proposed an intermediate-stage deployment that turns off the finality gap rule and treats finality as advisory information only.
    • Nate preferred staking real ZEC on mainnet to gauge interest.
    • Asked by Josh about economic security, Nate said they would do an analysis comparing security relative to issuance with the existing PoW.
    • Daira Emma explained that the no free lunch is that for safety only one of the two needs to be safe, but for liveness both are needed.
    • The new risk of introducing BFT is that the chain could halt, and Nate judged a halt to be better than a long-range rollback because it makes everyone take the same response.
  11. PoW/PoS economic design discussion

    Shielded Labs · Pacu, Nate, Daira Emma, Kris, Josh, Dodger, Zooko

    • Pacu worried that if existing miners also act as validators, rewards could become concentrated and make it hard to judge the security state, and Nate said they have not yet found a benefit but will analyze it further.
    • Kris said that if validators are the same organization as a mining pool it is not decentralization, and pointed out that a 50-50 split is not essential.
    • Kris said that at first an incremental approach that relies entirely on the finality point could safely start an unequal split.
    • Citing the drop in hash rate when the coin price fell to around $18, Josh asked about the price range at which PoW miners leave and the appropriate PoS yield.
    • Dodger pointed out that a price floor may come from holders' choice to hold rather than from miners leaving.
    • Daira Emma said whether private delegation is possible is what matters and suggested bringing in an economist, and Zooko suggested using the public comment period.
  12. Open discussion: ZCG process and upstream contributions

    ZCG · Pacu, Daira Emma, Pili, Dodger, Jason McGee, Kris

    • Citing ChainSafe's in-memory wallet case, Pacu raised the problem that grant recipients create forks of several tools while upstream contribution is outside the grant scope.
    • Daira Emma said a fork without a PR is effectively the same as nothing, and that the team must be told in advance when a PR is opened on a secondary repository, and Pili agreed.
    • Dodger proposed that when a grant involves changes to software maintained by others, whether the changes will be accepted and whether an audit is needed should be discussed first at the evaluation stage.
    • Jason McGee said that the last ZCG meeting decided to include this in the application process.
    • ChainSafe is expected to apply for an additional small grant because the librustzcash work turned out bigger than expected.
    • Kris announced that to be included in NU7, ZIPs must be semantically stable and submitted as PRs by November 5, and Jason McGee and Nate agreed to review the NSM ZIPs together afterward.

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