← All calls
ARBORIST#97

ZIP security and privacy sections debated

The lack of security and privacy sections in most NU7 candidate ZIPs was discussed.

Agenda summary 9

  1. Crosslink Deployment Milestone 1 complete

    Shielded Labs · Nate

    • The first milestone set in the roadmap is done; as a foundation-laying step, Zebra, Zaino and zcash devtool were modified to work together.
    • The team began building a new RPC interface related to finality and showed a demo in which a light client, in regtest fashion, turns a Zebra node's notion of finality into a specific block hash.
    • The milestone was carried out as a vertical slice producing thin results in each part of the stack, so it included documentation updates and ZIP placeholders.
    • The NSM PR for Zebra was split into a PR containing only the transaction format change and a PR adding the ZIP 233 field.
    • Through this the team contributes upstream, and it is thinking about how to coordinate and ship new transaction formats earlier in the future.
  2. Crosslink Milestone 2 plan and request for collaboration

    Shielded Labs · Nate

    • Milestone 2 is a separate BFT that integrates the Malachite SDK, a Rust implementation of Tendermint, into Zebra without modifying PoW consensus.
    • A BFT component agrees on the order of PoW blocks across nodes, and the goal is a mainnet shadowing mode in which anyone can connect it read-only to mainnet and try it out.
    • The target date for Milestone 2 is June 1.
    • Architecture decisions, including with the Zebra team and also Zaino, zcash devtool, librustzcash and the RPC connection, must be coordinated early.
    • The team plans to reach out actively to its list of stakeholders, such as wallet vendors and development teams, for early feedback on the milestone.
    • The detailed roadmap can be seen on the blog at shieldedlabs.net.
  3. ECC update: Zallet and Zaino integration and crate releases

    ECC · Kris

    • Str4d integrated Zaino into Zallet as a library so syncing works, and is filing what is needed going forward as issues in the Zaino repository.
    • New releases of the Zcash keys, client backend and client SQLite crates and the ECC wallet SDK are coming soon.
    • These releases include transparent gap limit handling for Bitcoin-style address rotation recovery and will be in next week's Zashi release.
    • A bug fix for a possible note commitment tree corruption in Zcash client SQLite also goes out with them.
    • The team also worked on the ZIP specification for the future community governance model Josh will announce soon.
  4. NSM update

    Shielded Labs, Eiger · Jason McGee, Kris

    • They could not make more progress over the past month and are waiting for PR 9300, which raises Zebra's dependencies to the latest librustzcash, to be merged.
    • Arya said the ZIP 233 implementation review is on hold until PR 9256, which introduces NU7 constraints aligned with ZSA, is merged, so they asked when the two PRs would be merged.
    • Kris answered that they had not had time to look at these PRs in recent weeks but would prioritize them.
    • Kris asked that such matters keep being flagged on Discord or Signal.
  5. Public announcement: replacing the community notetaker

    ZCG · Jason McGee, Nate, Dan

    • That day was the last day for Jason Rogers, who kept the Arborist call records for two years, and ZCG is looking for a successor.
    • Artkor will fill in temporarily until a successor is chosen.
    • Nate thanked Jason Rogers, saying the notes thread was Nate's only link to Zcash while on leave.
  6. Review of the Zenith FROST integration grant proposal

    Zenith, ZCG · Jason McGee, Rene, Nate, Dan

    • Rene of Zenith, a Zebra full-node wallet, proposed a grant for FROST support and Keystone integration, and ZCG asked whether the core development team uses Zenith and whether funding it is worthwhile.
    • Rene asked the FROST team for its view on whether carrying out the FROST multisig procedure serverlessly over decentralized, end-to-end encrypted P2P through Tor meets the FROST book's broadcast channel criteria.
    • This approach ports the Cwtch protocol to Haskell, embeds a node in each wallet, and also verifies the counterparty's identity through mutual authentication.
    • Nate said it was interesting and that they would like to be an alpha or beta tester.
    • At Dodger's suggestion, Rene will post the details on the forum and tag Conrado.
  7. Open discussion: improving the ZIP process and security and privacy sections

    Shielded Labs, ECC · Jason McGee, Kris, Str4d, Nate

    • Jason McGee, drawing on the experience of the NSM ZIP taking longer than expected, said that for the Crosslink ZIP they will have three Shielded Labs engineers attend the ZIP meetings and grow into ZIP editors.
    • Jason McGee pointed out that the security and privacy considerations section required by ZIP 0 appears in only 2 of the 10 NU7 candidates, and that it is also unclear who is responsible for the assessment.
    • As an example, they said ZIP 231 creates a size-leakage risk through variable-length memos and could conflict with the wallet threat model document, yet there is no record of analysis.
    • Kris acknowledged that ECC engineers discussed ZIP 231 extensively but did not record it adequately in the ZIP, and said they would fix that.
    • Str4d explained that ZIP 0's should is not mandatory and that much of the security and privacy rationale is in the Rationale section, and said responsibility ultimately lies with the author.
    • Nate proposed making the section mandatory and, where it does not apply, writing a sentence or two saying why.
  8. Open discussion: when to do security analysis and the ZIP 226 example

    ECC, Shielded Labs · Jason McGee, Str4d, Nate

    • Jason McGee asked whether security analysis of consensus changes such as ZIP 234 should be done at submission, at implementation, or both.
    • Str4d said that by the norms, consensus-level changes must be finished before the mainnet activation height is set, and that in practice it is better done before a specification audit or early in implementation.
    • Nate said that having security and privacy rationale in place before community ZIPs are selected can reduce cases where a team builds for a long time and is then rejected, and pointed out that the selection process is unclear.
    • Str4d introduced that, in the action groups and ZSA burn change proposed by QEDIT, the ZIP editors found a privacy trade-off and suggested a way for issuers to mitigate it outside the protocol.
    • This matter is being discussed in PR 991 and relates to ZIP 226.
  9. Open discussion: ZIP 0 revision, summary documents, governance

    ZIP editors, Shielded Labs · Nate, Str4d, Zooko, Kris, Jason McGee, Dan, Mark

    • Str4d said changes to ZIP 0 are easier to discuss if proposed as a concrete diff PR like any other ZIP, and Jason McGee said they would prepare a proposed change by the next call.
    • Jason McGee said most community members vote without reading ZIPs, so material summarizing the trade-offs and the affected stakeholders is needed.
    • Str4d explained that an exact specification and easy-to-understand writing are in tension, and that they have polished the motivation section to be the most readable.
    • Kris said every ZIP has an abstract, so the abstract should be improved rather than a separate summary document being created, and Mark of Shielded Labs volunteered to help with plain explanations.
    • Zooko said ZIP 226 shows its security rationale but for other ZIPs there is no way to tell whether analysis was done, and they want to separate decision-making from information gathering and track it in a public document.
    • Kris said the ZIP editors check that the analysis section exists but the analysis is the author's responsibility, and acknowledged that a formal governance process is now needed.

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