Small NU7 proposal draws pushback
ECC-side participants opposed Shielded Labs' proposal for a smaller NU7.
Agenda summary 8
Zebra update
Zcash Foundation · Arya
- Zebra 2.5.0 was released, deploying NU 6.1 on testnet.
- CI was improved by adopting nextest and optimizing Docker test builds, and the minimum required version was raised to 0.1.89.0.
- Funding streams can now be configured on custom testnets, and regtest was changed to do real batch verification with librustzcash's Orchard batch verifier.
- PRs in progress include completing the raw transaction RPC fields, backing up the state cache to restore non-finalized state on restart, and replacing the Sapling verifier.
ECC core update (Zallet, P2SH, librustzcash)
Electric Coin Company · Kris, Daira Emma, Pacu
- Changes are being added to the Zcash script crate for P2SH support, and the team has gone back to work toward the Zallet alpha launch.
- Zaino was updated to the latest Zebra release and librustzcash crates, and Str4d opened a draft PR in the zips repository with the PCZT format specification.
- The confirmations policy feature was merged into zcash_client_backend so Zallet and Zashi can support low-conf and zero-conf shielding and sending.
- For the NU 6.1 mainnet release, the key-holding organizations must finish the P2SH setup for the lockbox disbursement address, and the team reviewed key-generation methods that Ledger and Keystone can support.
- Three issues remain before the Zallet alpha: completing z_listunspent, changing the zcashd wallet import, and updating z_viewtransaction.
- The review of Qedit's Halo 2 PR was finished and merged, and the review of the Orchard PR for ZSA resumed.
- The OP_RETURN feature contributed by the Maya protocol team was also merged into the transaction builder.
Zaino update
Zingo Labs · Za
- The chain index API for Zallet to use is being finalized, aiming for full support of the non-finalized state that matters for Crosslink.
- About a week remains until the grant covering this work is complete, but a bottleneck is delaying progress.
- A redesign of the core code to improve code quality has also begun, and work on RPCs for block explorers will follow.
zcashd retirement DAG and priorities
Electric Coin Company · Pacu, Kris, Daira Emma, Alfredo
- The priority is completing the Zallet MVP item in the zcashd retirement DAG, with the chain index refactoring and block explorer API work in parallel.
- The testnet block explorer is currently running on Zebra as a stopgap, and Zingo Labs is finishing the official RPCs.
- Work on the compact block protocol's handling of transparent data and fixing privacy side effects is also in progress.
- Daira Emma pointed out that it is inaccurate for the DAG to mark setting the NU 7 activation height as a prerequisite for replacing the zcashd wallet, and Kris said they would fix it.
- The DAG is updated every two hours.
- Conrado updated the RPC method spreadsheet for retiring zcashd to identify features missing from Zebra.
- Daira Emma said zcashd's getrawtransaction-style implementations do not account for shielded transactions and contain wrong fields, and raised a policy discussion on whether bug compatibility is needed even for obvious errors.
- For lack of time it was postponed to next time.
Zcash Shielded Assets (ZSA) update
Qedit · Vivek
- A PR containing all the ZSA changes to Halo 2 was merged, and the Orchard PR supporting it is awaiting further review.
- The SIGHASH signature versioning change was merged into the ZIP, and the encoding changes for the issuance authorization signature and verification key were reflected in the test vectors and the Python reference implementation.
- The Orchard integration is nearly done.
- Kris said the problem of exposing internal types in the Orchard bundle type is resolved and ECC is satisfied.
- The PR is about 30,000 lines of code and 20,000 lines of verification keys, so review takes time.
- librustzcash has mostly caught up with upstream changes such as PCZT, and the Zebra fork supports ZSA based on 2.4.2 except for the state management (finalization) part.
- The tool for generating v6 transactions is also maintained.
- Alfredo said ZSA wallet support needs work in zcash_client_backend and asked for Qedit developers to take part, and Vivek accepted.
- Kris proposed designing the ZSA issuance proposal API and the PCZT flow top-down.
Network Sustainability Mechanism (NSM) update
Shielded Labs · Mariusz
- The ZIP 233 PR was rebased and updated onto the latest librustzcash and is awaiting review.
- Review comments on zcash-test-vectors PR 101 were addressed, and this PR is also awaiting review.
- Kris could not find the ZIP 233 PR in the zips repository, and it turned out that PR was in the test vector repository.
- Daira Emma will review it.
Crosslink update (Trailing Finality Layer)
Shielded Labs · Nate
- The team is wrapping up Milestone 3, running and connecting the two consensus protocols together, and has a test case showing that finality prevents rollback before a node's last finalized PoW block.
- What remains is writing the technical documentation.
- The plan is to include an ADR (architecture decision record) as an appendix in a fork of the Zebra book, publish it at a public URL, and create a release and tag.
- The team wants to show a sturdier testnet at next week's Shielded Labs in-person workshop; the current approach is fragile because every node must change its configuration file at the same time when nodes are added or removed.
- Several attempts at formalizing Crosslink were discussed together, and Shielded Labs, in consultation with Informal Systems, which uses Quint, wants to try running automated proofs by the end of this year.
- Daira Emma plans to work on a Lean formalization to match.
Open discussion: small NU 7 proposal
Shielded Labs · Jason, Nate, Daira Emma, Kris, Arya, Str4d
- Jason proposed that Shielded Labs lead a small, fast NU 7 containing NSM, explicit fees and all the transaction format changes brought forward.
- The goal is a six-month upgrade cycle, and they rebutted concerns about ZSA delays and ecosystem breakage.
- Daira Emma opposed.
- They said issuance smoothing will be pushed to the next interim point (November next year) anyway, the effect of fee burning is long term, and explicit fees are a cleanup with almost no user impact.
- Daira Emma argued that changing the transaction format before the features are settled is risky, and amounts to slipping in an unrequested NU 6.2.
- Kris said putting the full format change, including the memo bundle, into zcashd is too risky, and Jason answered that keeping only the changes needed for NSM would also be fine.
- It was also pointed out that, with the 16-week sunset grace period after the NU 6.1 mainnet height is set, activation is hard before the end of Q1 2026.
- Arya said it is inappropriate for this NU 7 but that the project should return to regular, incremental upgrades in the future.
- Str4d cited past experience of failing to keep a six-month cycle and asked what would be different this time.
- Daira Emma answered that six months is not realistic and once a year is appropriate.
- Jason said the current process of repeated schedule slips needs to change and that they are seriously considering pushing the proposal.
This summary is based on the original meeting notes; check them for exact details.