NU7 October at earliest, Sprout alternative
An outlook that NU7 comes in October at the earliest and a ZSA-based Sprout redemption idea were discussed.
Agenda summary 15
ECC update: Z|ECC offsite and Zallet
ECC · Kris
- The ECC team finished a productive in-house offsite and came up with several ideas to present to the community, including approaches to retiring the Sprout pool.
- Zallet, the wallet for full nodes, is quickly filling in the initial RPC service and the overall skeleton.
Zebra update
Zcash Foundation · Arya, Dodger
- getblockchaininfo now returns genesis block information when the state is empty, and the RPC client in tests was moved to the new JSON RPC crate.
- To check receivability, non-coinbase transactions are no longer converted to Zcash primitives types, and NU5 test block vectors were added.
- Preparation for deploying a long-running Zebra instance whenever the main branch is updated was merged.
- A PR is open to fix a concurrency bug related to the automatic verification of block transactions already verified for the mempool.
- Dodger said the project moved from the deprecated crate to the new JSON RPC crate, so the ecosystem might consider standardizing on one.
- Arya said they are applying this crate to Zallet and running into compatibility problems with zcash-cli, so they should discuss it separately.
zcashd retirement: outreach and schedule delay
Zcash Foundation, ECC, Zingo Labs · Pacu, Str4d, Za, Dodger
- Pacu received responses from one exchange and one mining pool and updated the public spreadsheet of RPC methods by partner, and is waiting for more responses.
- ECC reported that NU7 slips by about two months from August to October and that the retirement schedule could be affected.
- Str4d explained that August was the optimistic earliest schedule the ZIP editors gave in October, not a target deadline.
- Za said Zaino's Milestone 2 is late, with a best estimate of mid-May, and that the delay is due to the estimate having been optimistic about onboarding new staff.
- Pacu proposed gathering ECC's recorded code walkthroughs and adding reviewers to reduce the librustzcash review bottleneck.
Zaino scope and whether lightwallet features are needed
Zingo Labs, ECC, Zcash Foundation · Dodger, Za, Kris, Arya
- Dodger said that since Zebra already has a working lightwalletd, Zaino's lightwallet features are not on the critical path of the retirement, and proposed prioritizing the CLI wallet support portion.
- Za answered that the interface Arlo designed is structured as a library that other Rust crates can use and that the lightwalletd replacement part is nearly done.
- Kris said that lightwallet protocol APIs such as transparent UTXO access are needed by full-node wallets too, so this is not a detour.
- Zaino's high-level interface is backend-independent, so it can use both the read state service and RPC.
- Arya said that once the interface is set, Zaino does not block wallet development, and that Zaino's completion is needed only for integration tests and bug fixing.
Zallet development approach and distribution strategy
ECC · Str4d, Kris, Arya, Pacu, Dodger
- Str4d said the approach for Zallet is to first have a shell that can be put into build and distribution pipelines, ahead of features, so partners integrate and give feedback early.
- By the end of February the surrounding infrastructure will be in place and the features supported by Zashi and the mobile SDKs usable, and from March the zcashd wallet's own features will be exposed in turn.
- Zaino and Zallet are expected to communicate with each other within the first quarter, and Kris said they would access it directly through the Zaino API rather than the lightwalletd RPC.
- Str4d said it is desirable for the wallet and node to be separate services, and that if needed, a shim that splits RPC across the two services is not a big task.
- In RPC test integration, the Python node abstraction acts as the shim by bringing up zebrad and Zallet together and routing methods appropriately.
- Kris said the goal is Docker Compose first deployment, and that production deployment such as Helm also starts as soon as Zallet responds to RPC.
ZSA integration ZIP editors meeting
ECC · Str4d, Dodger
- At the ZSA integration meeting earlier this week, an audit concern about how a certain constant is generated came up and was resolved by improving the API documentation.
- The ZSA ZIP integration process has already started and the Zcash note encryption changes have been incorporated, and some PRs reach 20,000 lines.
- They discussed putting in advance, into the v6 transaction format, the changes for atomic swaps proposed by QEDIT.
- They passed to QEDIT the ZIP editors' sighash design, which means the swap sighash need not be defined in NU7 because its semantics are not yet stable.
- Dodger lightly mentioned the idea of defining the transaction format in a standard language such as XML to make it easier to support new formats.
Crosslink update
Shielded Labs, Zcash Foundation · Jason McGee, Dodger
- Zooko and Nate were absent, and Jason McGee introduced Shielded Labs' new engineer Sam Smith.
- Sam and Nate are focused on building the Crosslink prototype and will give an update on the February 6 call.
- Dodger said that, as post-NU7 work, ZF is discussing refactoring Zebra so it can plug in various consensus mechanisms.
- The purpose is to make it easy to create testnets that try ZSA or atomic swaps without mining.
FROST demo, PCZT, community call
Conrado, Kris, Str4d, Pacu, Dodger
- Broadcast channel support was added to the FROST client demo and the documentation updated, and the signing tool that had stopped working after NU6 was also fixed by raising the ECC dependencies.
- Around the first week of February, a community call will present the status of the demo and its overall structure and aim to draw wallet developers' interest in integrating.
- Kris proposed moving the FROST signing tool into a zcash devtool command and supporting PCZT in addition to the Ywallet format, so it can also be used with hardware wallets.
- Str4d and Conrado said FROST logic can be carried with PCZT v1 and proprietary extension fields, and dedicated fields will be considered in PCZT v2 after the v6 format is finalized.
- Pacu said this work is useful for the guardian signing of Red Bridge, an Avalanche-Zcash bridge that uses Orchard and FROST.
- Dodger set expectations that after the community call, FROST will briefly take a back seat in order to focus on zcashd retirement and NU7.
NSM update
Shielded Labs · Jason McGee, Kris, Dodger
- To address concerns about the term burn, all burn wording was removed from the ZIPs.
- The description was changed so that ZIP 233 is a mechanism for removing coins from circulation and ZIP 234 is a mechanism that flattens the issuance curve and recycles the removed coins as future block rewards.
- ZIP 235 removes 60% of transaction fees from circulation to support network sustainability.
- After revising the ZIPs with comments from Str4d and Daira Emma, the team moves to integration work and wants the NSM direction decided in the NU7 schedule delay and NU 6.5 discussions.
- Kris said merging into zebrad and wallet code behind a feature flag as soon as possible is the way to reduce conflicts.
- Dodger set expectations that PR review time may be limited because of priorities.
NU7 schedule delay analysis
ECC · Str4d
- Str4d explained that the schedule the ZIP editors gave in October was the shortest possible schedule for the minimum steps before activation, and that things went smoothly up to the semantics-stable deadline.
- After that, other work was pushed back by ECC's shift to the Keystone hardware wallet.
- In early December the ZIP editors issued a safety, feasibility and usefulness assessment, but with too little community feedback it is uncertain what will go into NU7.
- They said that without feedback, silence should be taken as consent, and that the point at which an audit starts should effectively be the deadline for changes.
- The bottlenecks are PR review of a lot of code including ZSA, and the fact that zcashd retirement must be sufficiently ahead before the mainnet activation height is set.
- Both paths are about two months behind the optimistic schedule, so Str4d believed an upgrade containing ZSA is unlikely before October.
Options for reducing NU7 scope and audit impact
ECC · Str4d, Arya, Dodger
- Arya asked whether it is acceptable to request the audit for both the feature-flag-on and feature-flag-off states.
- Str4d answered that independent features such as ZIP 235 are no problem, but that the ZSA note encryption changes and the memo bundle changes are tightly coupled in the code implementation, so the combination must be settled before the audit.
- On the premise that including ZSA is central, ZIPs 2002 and 2003, which touch the transaction format only slightly, and the field addition in ZIP 233 are cheap.
- There is also the option of putting in only the format and setting it to zero in the consensus rules, and QEDIT's action group changes are being handled the same way.
- The memo bundle is entangled with ZSA and essential for uses such as authenticated senders, so if it were dropped, the decision would need to be quick.
- Most of the code is implemented, and the core concern is PR review staffing.
Lockbox extension discussion and consensus branch ID
ECC, Zcash Foundation · Josh, Str4d, Arya, Dodger
- Josh said NU7 is October at the earliest and could slip past November, and that there is a problem in that the Dev Fund going to the lockbox and ZCG ends in November.
- The options are to stop the funding, to send part of it to NSM at activation, to include a one-year lockbox extension in NU7, or to split the extension out as a separate upgrade ahead of NU7.
- ECC currently has about two years of operating runway and the timing of funding matters for its planning, so it asked for opinions on the forum.
- Str4d said funding stream changes are easy, and that three months between upgrades is the shortest period the ecosystem, such as exchanges, needs to respond.
- Str4d said parallel preparation is possible and the order does not matter, so people should not get hung up on numbers like 6.5.
- Dodger relayed an opinion that the approach of changing the consensus branch ID at every upgrade should be reconsidered, and Str4d explained the two-way replay protection motivation in ZIP 200.
Public announcements: Pop-up City, Zcash Dev Summit, Z|ECC Prague
Zcash Foundation, ECC · Za, Str4d, Kris, Dodger, Josh
- To address the problem of librustzcash knowledge being concentrated in a few developers, Za floated the idea of holding a knowledge-spreading cluster at a Pop-up City event.
- Str4d and Kris said the hacker house after this ECC summit was beneficial.
- ZF's Zcash Dev Summit will be held in Sofia on Saturday, March 22, just before ZK Proof and Real World Crypto, with attendees setting the content and the next day a joint work day.
- Josh said the next Z|ECC summit is planned for the first week of July in Prague, with structured sessions and a hacker house.
Open discussion: ways to retire the Sprout pool
ECC · Kris, Str4d, Arya, Dodger
- Sprout coinholders should move as soon as possible, and at the hacker house new ideas were raised for the problem that removing v4 transactions would make Sprout funds unspendable.
- A group of trusted signers verifies proofs and issues a ZSA representing Sprout funds, and within the consensus rules burning this ZSA lets it be exchanged for ZEC from the Sprout pool balance.
- Because retiring v4 freezes the Sprout pool balance, the maximum issuance of the synthetic asset is the remaining balance, and if the community decides, ZSA issuance can be ended.
- The only consensus changes needed are the asset definition and burn rules, so there is no new circuit and, using NU7's ZSA circuit, the risk of a counterfeiting bug is lower than with direct ZEC issuance.
- Arya added that if the proofs are put on chain via the memo bundle, anyone can verify that the issuer is behaving correctly.
- Dodger welcomed it as a better alternative for closing the bad-idea issue they filed six years ago. Str4d said Sapling could be retired the same way, and judged that dealing with it in this NU is hard given the schedule.
Open discussion: Equihash PR and calendar notice
Zcash Foundation, ECC · Arya, Str4d, Dodger
- When Arya asked about the status of the librustzcash Equihash PR, Str4d answered that it is on the review list along with the ZSA and NSM PRs.
- Dodger said they are looking for a fix to the Arborist call calendar problem, and in the meantime warned that the schedule registered on Zoom could be wrong.
This summary is based on the original meeting notes; check them for exact details.