← All calls
ARBORIST#60

Deirdre leaves; SDK release candidate imminent

The mobile SDK release candidate is imminent, and Deirdre said they are leaving the Zcash Foundation.

Agenda summary 9

  1. ECC core update: mobile SDK release candidate

    Electric Coin Company · Kris

    • The team is quickly approaching a release candidate for the next Android and iOS mobile SDKs. The Zcash core library changes in zcash_client_backend and zcash_client_sqlite are all done.
    • Final testing is underway, and the SDK release candidate was expected the next day. Putting non-linear sync, order-independent sync, and the ability to spend funds before the wallet is fully synced into the mobile SDK was the whole of the recent work.
  2. Zebra 1.2 release

    Zcash Foundation · Marek, Teor

    • Zebra 1.2 was released last Friday. It includes note commitment tree deduplication together with an internal framework for DB upgrades, so users do not have to resync the node from scratch when the DB is upgraded.
    • With this framework, subtree roots were added for the new RPC method z_getsubtreesbyindex. It is being tested and will be in the next release. Teor warned that this upgrade is subtly incompatible with earlier versions.
    • If a DB upgraded to 1.2 is used with an earlier version, it may panic or some RPCs may return wrong results.
  3. FROST audit and documentation update

    Zcash Foundation · Deirdre

    • Final fixes from the FROST implementation audit and improvements to the library usage documentation are underway. Zcash-specific usage, the shielded signing process and the relation to the key tree are also being merged into the docs.
    • The follow-up fixes from the audit change the shape of the public API, but it is still a 0.* beta, and the goal is a stable 1.0 whose API does not change afterward. The Zcash-specific documentation will soon be merged into the Zebra Book.
    • A FROST and wallet developer community call is being prepared for the in-depth discussion with wallet developers that did not happen at Zcon.
  4. Trailing Finality Layer research

    Electric Coin Company · Nate

    • After the Zcon workshop presentation, an update was posted on the forum. The key concern is whether, when the PoW and PoS subprotocols are combined, an attacker needs to break only one of them, and whether the attack cost becomes lower than pure PoW or PoS.
    • The first milestone is to define this security problem clearly and write a Trailing Finality Layer design that corresponds exactly to the ebb and flow paper.
    • The initial design treats PoS as a black box, leaving out detailed mechanisms such as issuance rate and staking bonds. The second milestone is writing code in parallel with the design, with the first target being a simulation of the attack under analysis.
    • The team plans to give regular updates going forward.
  5. Zcash Wallet Community Developer update

    Zcash Wallet Community · Pacu

    • Milestone 1 was completed and an update was posted on the forum. R&D on the Second Flush of Enthusiasm dataset is finished, and the Nighthawk app PR was also reviewed. Milestone 2 is short, in the first half of September.
    • The plan includes integration testing with Zingo Labs, Nighthawk review, support for Unstoppable and Edge adopting the ECC SDK, and porting reorg tests to Regtest. They will be away from September 14 to 28.
  6. ZSA ZIP editor update

    QEDIT / ZIP editors · Deirdre, Str4d

    • The QEDIT team is writing a v6 transaction draft that includes ZSA, Transfer, Issuance and Burn. Updated circuit information for Transfer and Issuance was also received.
    • QEDIT plans to develop on Zebra wherever possible from now on. However, Zebra has no integrated wallet, so this is being discussed in line with the Foundation's roadmap. Zcashd is heading toward deprecation.
    • A way to separate the ZSA issuance signature from the Orchard parameters was discussed. A Schnorr-based signature in place of the current redpallas was sketched and is not yet final.
    • Str4d said BIP-340 (secp256k1 Schnorr) is a candidate because of ease of HSM deployment. People with constraints on asset issuance were asked to give feedback on Discord or the forum.
  7. ZIP editor process changes

    ZIP editors · Str4d, Teor, Nate

    • A proposal was made to add ZIP editors. The intent is to formalize the loosely defined auxiliary staff and spread the review and decision burden. Editor consensus would be required to accept or reject a ZIP and for major changes to approved ZIPs.
    • The direction is that adding or removing editors is decided by consensus minus one, with details to be documented in a ZIP.
    • The meeting quorum is being considered as one Foundation editor and one ECC editor, with a ZIP secretary who posts the agenda and decision summaries. Content review and style and typo review will also be separated.
    • Those interested in becoming a ZIP editor can say so in the #zips channel on the R&D Discord.
  8. Deirdre Connolly leaves the Zcash Foundation

    Zcash Foundation · Deirdre, Jack Gavigan

    • Deirdre is leaving the Zcash Foundation, with their last working day on September 13. Their ZIP editor role also ends then, and they will work on Zcash and other areas of cryptography.
    • They said that as a FROST co-author they will continue to take part in threshold signature discussions. Jack Gavigan thanked Deirdre for their major role in Zebra and FROST.
  9. Detection keys and out-of-band notes / Liberated Payments

    Jack Gavigan, Str4d, Nate, Deirdre, Kris, Zooko

    • Because scanning the full chain burdens wallets, the team discussed passing detection keys and note information out of band. A detection key tells you a transaction is yours but does not reveal the amount or memo.
    • Liberated Payments (URL-encapsulated payments) pass the spending key of a temporary wallet holding a set amount via a URL. Payment is immediate without knowing the recipient address, and the sender can reclaim it for a period.
    • Detection keys need extra key material in the UA and are opt-in at first. Fuzzy detection keys and PIR-based approaches were mentioned as promising, and full rollout requires a network upgrade.
    • Nate said to first experiment with approaches that can be tried without protocol changes where possible. Str4d thought that combining DAGSync with direct wallet-to-wallet transfer would let most users go without scanning the main chain.
    • Server-side detection costs money, but economies of scale may make it more efficient than scanning on each device. A model in which the server sends real-time notifications of new transactions was also raised. Use cases where an address is published, such as a donation QR code, still need a full scan.
    • Str4d said detection keys can be imitated outside consensus using the memo field of dummy notes, and that out-of-band notes can be experimented with right away by wallet developers. They encouraged submitting Minor Grants applications.

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