NU6.1 lockbox multisig keys requested
Testnet multisig key generation for the NU6.1 one-time lockbox disbursement was requested.
Agenda summary 9
Zebra update: CI, NU6.1, DB upgrade
Zcash Foundation · Arya
- Several CI improvements were merged, including fork PRs and cached state management, and a change letting the debug crate compile with a different version of its dependencies went into main.
- A NU6.1 network upgrade variant, not yet used, was added to main, and the LightwalletD version in CI was pinned.
- A major DB format upgrade was added for the new fields in the getaddressbalance and getblock RPCs.
- Regtest now allows local outbound connections defined in the configuration, an addnode RPC was added, and an initial framework for testing RPC methods was merged.
- A problem of redundantly serializing Zebra types into librustzcash types during transaction verification was fixed, and a setting for mining to a shielded address was added, though shielded coinbase creation is not yet done.
- A received field was added to getaddressbalance, and the new script trait from zcash_script was merged into zebra-script.
Zebra update: open PRs and the NU6.1 ZIP request
Zcash Foundation · Arya
- A PR implementing the one-time lockbox disbursement mechanism for NU6.1 is open, and the team asked for the ZIP for it to be finalized quickly.
- A PR is also open that cleans up the API so the Zebra RPC client types can be used without depending on the Zebra state or server code.
- A PR adding value pools and block size to the getblock RPC is due to be merged soon.
ECC update: NU6.1 lockbox disbursement and Zallet
ECC · Kris
- The NU6.1 consensus change for the one-time lockbox disbursement was implemented in zcashd.
- A new release of the whole librustzcash stack of crates was made to improve Tor support and advance Zallet features.
- New RPC methods such as z_gettotalbalance were implemented in Zallet and development keeps being pushed forward.
zcashd retirement: Zallet alpha scope and key setup for testing
ECC, Zcash Foundation · Kris, Alfredo
- The team is at the stage of sorting out the remaining items for the Zallet alpha release, which will include basic features such as zcashd wallet migration, z_sendmany, sync and transaction history, centered on modern APIs.
- Some old APIs inherited from Bitcoin will remain incomplete in the alpha; there is no date yet, and they hope to settle timing and scope by the end of this month or the next call.
- The only key import the alpha supports is the zcashd wallet import; transparent and legacy Sapling key import is supported by the backend but not exposed over RPC.
- Alfredo said that in Python tests they need a way to put keys into the wallet DB so regtest mining funds go into the wallet.
- Kris said it would be better to create a CLI command that adds an account to the wallet without starting Zallet, and also proposed a flag on wallet init for whether to create a default account, and Alfredo will open an issue.
NSM update: status of the ZIP 233 PRs
Shielded Labs · Mark, Kris
- Three ZIP 233 PRs, in librustzcash, Zcash test vectors and Zebra, were rebased, and Mariusz addressed Daira Emma's comments.
- Mariusz said the digest header change in the librustzcash PR breaks several of Zebra's tests that handle old block test vectors, and asked whether the ZIP 233 amount recording should depend on transaction version.
- Kris answered that it should depend on transaction version, and that if not, it is a bug.
- Once reviewed and merged, the plan is to move on to ZIP 234 and 235.
Crosslink update: Milestone 3 design and objective validity
Shielded Labs · Nate
- The Milestone 2 mainnet shadowing workshop network was fragile enough that all nodes had to have the same peer configuration and it could hardly survive nodes dropping off.
- Milestone 3 modifies PoW consensus to implement the basics of Crosslink consensus, with BFT and PoW blocks referencing each other; there is no proof-of-stake, delegation or bonding, and BFT is a PoA based on a configuration file.
- The goal is a node on which the design, implementation and security issues of this consensus can be reviewed.
- The team researched obtaining objective validity, so that a snapshot of PoW and BFT blocks received from a source that is hard to trust can be checked for internal consistency and rule compliance.
- Malachite by default starts from a checkpoint and receives BFT blocks and off-chain signatures from trusted nodes, and the team is also exploring putting the signatures on chain.
Crosslink update: staking design up front and testing approach
Shielded Labs · Nate, Dan
- The goal at the end of Milestone 4 is a testnet with finalizer operation, staking, delegation, fees and penalties, and to that end the design is being brought forward in this milestone to gather early feedback.
- The team compiled questions from validators on other PoS networks into an FAQ, and also talked with Informal Systems, the developer of Malachite, and Christopher of Anoma.
- Next it plans to study Cosmos Hub, Namada and Penumbra and reflect them in the design.
- It is preparing a testing method that replays a linear-sequence test vector mixing PoW blocks and BFT blocks into a hybrid Zebra node and verifies the state.
- This test keeps its logic as long as the serialization does not change, and the team wants to share it with the Zebra team.
- The roadmap is on shieldedlabs.net and follows the existing schedule.
Open discussion: passkey key derivation and the FROST key derivation spec
ZCG, ECC · GGuy, Kris
- GGuy shared a proposal on whether ZCG could support passkey-based key derivation and asked whether Zashi could use it.
- Kris explained that FROST key derivation is not yet fully specified, and that ECC has proposed a FROST key derivation that has a recovery path even if quantum computers appear.
- Kris said this spec must be finalized before people start using FROST-derived keys, and is needed before NU7.
Open discussion: keys for the lockbox disbursement and a FlyClient question
ECC, ZCG, Zcash Foundation · Kris, Nate, GGuy, Arya
- Kris asked the initial key-holding organizations for the one-time lockbox disbursement to generate testnet multisig P2SH keys.
- With those keys, they can be put into the ZIP and a testnet release can be prepared in which both Zebra and zcashd test the lockbox disbursement; the people responsible are Josh and Jason, and the ZF lead.
- GGuy thanked the people who helped grantees define their milestones more clearly.
- Zooko asked whether Zebra verifies the FlyClient bit in block headers, and Arya answered that, with 95% confidence, it does not.
- The next call is June 26 at 15:00 UTC.
This summary is based on the original meeting notes; check them for exact details.