← All calls
ARBORIST#98

Note construction change for quantum readiness

A proposal to change note construction so notes remain usable after quantum computers was introduced.

Agenda summary 11

  1. Zebra update

    Zcash Foundation · Marek, Arya

    • CI and Docker were substantially restructured and the Zebra Docker configuration method changed; as a breaking change, the RPC endpoint is not on by default and users must turn it on themselves.
    • The getblock, getrawtransaction and getpeerinfo RPCs were extended, and the invalidateblock and reconsiderblock RPCs were added.
    • Negative height support was added to RPCs that zcashd supported, and a trusted chain sync bug that had been bothering the Zaino team was fixed.
    • The internal miner used for testnet and regtest block generation was restored, and the coinbase transaction output order, which was the opposite of zcashd's, was corrected, ending confusion in block explorers.
    • The internal database handling was refactored to make it easier to maintain network upgrades.
    • Arya thanked Str4d and teor for finishing the Equihash PR and helping restore the internal miner.
  2. ECC update: crate stack release and Zallet

    ECC · Kris

    • A new release of the whole librustzcash crate stack was made, centered on Bitcoin-style wallet recovery based on the transparent gap limit.
    • Zallet has a branch that syncs the chain using a Zingo-side service as a data source, and z_send is being worked on.
    • After discussion with developers, an importer that brings a zcashd wallet into Zallet is being implemented.
  3. Proposal to change note construction for post-quantum readiness

    ECC · Daira Emma, Kris

    • The current note commitment is not quantum binding, so a quantum attacker could violate balance conservation.
    • It is a relatively simple change that only alters how notes are created; it does not change the consensus protocol itself but does affect note encryption and decryption.
    • New Orchard notes created by a wallet supporting this change can later be used with a post-quantum protocol, avoiding a situation where the whole existing protocol is turned off and the notes become unusable.
    • The analysis is being finalized, and the team thinks it can be introduced together with the memo bundle or note encryption changes.
    • Kris said this change also affects FROST key generation and must be reflected before the spec is finalized.
  4. QEDIT news (reported by proxy)

    QEDIT · Arya

    • There were no presenters from Zingo Labs or QEDIT, so Arya reported on their behalf.
    • QEDIT is mainly focused on revising ZIPs based on the ZIP editors' comments.
  5. NSM update

    Shielded Labs · Jason McGee, Arya

    • They are waiting for Zebra PRs 9300 and 9256 to be merged, and ZF replied on Discord that it hopes to handle them this week.
    • PR 9300 raises Zebra's dependencies to the latest librustzcash so the librustzcash PR can be rebased onto the latest main.
    • Arya said they would look at the ZIP 233 implementation after PR 9256 is merged, and that PR was merged that morning, so review can begin this week.
  6. Zingo Labs: Zaino update

    Zingo Labs · Arlo

    • The per-hash compact block lookup that Zallet needs was implemented and works.
    • The PR fixing the entry kind bug in Zcash history is ready for review, and once it is merged, the remote read state service PR will be nearly ready.
    • A detailed post on the staffing needed for the follow-up grant and the reasons was published on the forum and feedback was requested.
  7. Crosslink Milestone 2 progress: Malachite integration and visualization tool

    Shielded Labs · Nate, Andrew

    • Milestone 2 is to build a network that agrees on PoW blocks using Malachite, a Rust implementation of Tendermint BFT, and write initial documentation on the BFT block format and validation.
    • Andrew showed that, as a result of Milestone 1, a Zebra mainnet node prints finalized blocks to the terminal, with only the minimum number of characters needed to uniquely identify a hash highlighted.
    • A GUI visualization tool showing the PoW and PoS chains side by side was built, where circles are blocks, lines are parent links, size is the number of transactions, and distance represents work.
    • Using the RPC added in Milestone 1, they moved finality to an arbitrary point and demonstrated finalized and non-finalized blocks displayed differently.
    • They also showed creating an arbitrary node on a parallel BFT chain and adding a cross-link connection to the PoW chain.
  8. Crosslink milestone roadmap and testing discussion

    Shielded Labs, ECC · Nate, Daira Emma, Arya

    • Milestone 2 aims at a mainnet shadowing mode, connecting read-only to mainnet to see whether participants finalize the same blocks, with the BFT side as PoA.
    • In Milestone 3 the PoW header is modified without changing transaction semantics to produce the first complete Crosslink prototype, and the consensus change that decides which blocks are finalized also goes in at this point.
    • Milestone 4 implements the actual staking rules.
    • Daira Emma pointed out that in the full protocol each node's finalized state must not roll back, and that the BFT chain may also have parallel blocks not yet finalized.
    • Nate said that since BFT finality interacts with Zebra's existing rollback limit, they would write rollback tests carefully, and Arya recommended using the new invalidateblock RPC.
  9. FROST audit follow-up and the post-quantum key generation change

    Zcash Foundation, ECC · Natalie, Kris, Daira Emma

    • Natalie is responding to the results of the Least Authority audit of the FROST demo and will handle a bit more and reply.
    • ZF is gradually stepping back from FROST work, so whether further work happens will be discussed separately, and external PRs are welcome.
    • Daira Emma explained that for post-quantum readiness the way key generation computes the rcm value changes, so that a note must commit to g_d, pk_d, the value and the asset base.
    • This does not force a particular key generation algorithm, and if needed Daira Emma will reflect the change in the FROST spec personally.
  10. Open discussion: request to review the Zenith grant

    ZCG · Jason McGee, Marek, Arya

    • Jason McGee asked for opinions on whether ecosystem development teams use Zenith, regarding Zenith's grant proposal for FROST multisig and Keystone integration.
    • One attendee said they had never used Zenith, and Marek said they do not use it directly in Zebra either, but the issues the Zenith team opened were useful.
    • Arya said ZF would look in more detail if ZCG needs opinions, and Jason McGee said they would send the link.
  11. Open discussion: post-quantum protocol and scalability

    ECC, Shielded Labs · Arya, Daira Emma, Nate

    • Daira Emma said the core algorithm of a post-quantum protocol uses only symmetric cryptography and could use an interactive method that passes notes point to point, so diversified addresses may not be needed.
    • This method also has a scalability advantage because note scanning is unnecessary.
    • Nate said Sean Bowe published a blog post on a new scalability architecture that includes changes to the payment interaction protocol.
    • Daira Emma answered that these changes are compatible with each other and that both scalability and post-quantum readiness are motivations.

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