← All calls
ARBORIST#102

Crosslink Milestone 2 complete

Crosslink Milestone 2 is done, creating a one-way link from PoS to PoW.

Agenda summary 11

  1. Zebra update

    Zcash Foundation · Marek

    • As in the past few months, a lot of effort went into the RPC work needed for zcashd retirement.
    • Work continues on shielded coinbase transaction support, and the hotfix release procedure was updated.
    • The team is moving Python tests from zcashd to the z3 stack and working on Docker packaging of the z3 stack.
  2. ECC update: Zallet testing and the changed meaning of z_viewtransaction

    ECC · Str4d

    • Over the past two weeks the team reviewed PRs and worked on bringing librustzcash changes proposed by third parties into the mobile SDKs.
    • Bug fixes found earlier were merged into Zebra and Zaino so initial Zallet testing went further, and newly found Zebra and Zaino bugs were reported as issues and PRs.
    • zcashd's z_viewtransaction decrypted and showed transactions on the fly, but since Zallet stores decrypted information in the wallet DB, the team is considering changing its meaning to show already-decrypted information.
    • In that case the behavior would differ slightly from the old one where importing a viewing key made the wallet show all information immediately, so it will be documented along with a change to speed up the reflection.
  3. ECC update: Zewif import, Tachyon on-site, Tor HTTP API

    ECC · Str4d

    • Kris refined the Zewif changes with Blockchain Commons so they can be used in Zallet, and a draft PR converting an existing zcashd wallet into a Zallet wallet is up and will be tested over the coming weeks.
    • This week the ECC core team did three days of Tachyon R&D on site with Sean Bowe, 12 hours a day, making some progress and also finding new issues.
    • The zcash client backend crate's Tor API now exposes a Rust API for sending GET and POST requests directly, in addition to the HTTP GET used for exchange rate lookups.
    • With it, API queries to services that run an HTTP server can be made privately over Tor.
  4. Zingo Labs: Zaino update

    Zingo Labs · Arlo

    • The error propagation fix is about to be merged and other issues are being handled.
    • To check that the implemented RPCs work correctly, the team is running zcashd's RPC tests, and with the local cache reworked there are several PRs awaiting review.
    • Zingo Labs released Excavator v0 this week, available for early testing and use.
  5. zcashd retirement update: Excavator and block explorer

    Zcash Foundation, Zingo Labs · Pili, Pacu

    • ZF is finishing the last missing parts of the RPC and testing the new stack with the RPC test framework, while Gustavo is packaging everything into a single Docker image.
    • They asked to be told if there is anything to help with in the Zaino or Zallet implementation.
    • Pacu said Excavator is a tool for rescuing ZEC from wallet files and asked people to test it with old wallet files such as SQLite files.
    • Pacu has started work on the block explorer and is trying it with Zebra and Zaino, and Pili asked to be told the minimum amount of work among RPCs not yet implemented such as getnetworkinfo and getinfo.
  6. Zallet integration tests and import issues

    Zcash Foundation, ECC · Alfredo, Str4d

    • Alfredo opened a PR extending atomic tests that run Zebra, Zallet and Zaino together, and since there is no addnode RPC for connecting nodes to each other, it needs to be implemented.
    • Alfredo asked how to put the miner address used in tests into Zallet.
    • Str4d said it works as in zcashd: create the miner address in the wallet, then load it in the node configuration, and that the blockers now are Zallet issues 56 (import address) and 57 (import privkey).
    • These two can only be implemented once the backend changes that zcashd wallet import PR 152 will bring have landed.
    • Zallet has a lot to implement, and features that mobile wallets did not need must also be added, so anyone who wants to contribute was told to contact Kris or Str4d on the R&D Discord.
  7. Zcash Shielded Assets (ZSA) update: PR status

    QEDIT · Pablo

    • A small ZIPs PR correcting terminology was opened, and most of ZSA has already been merged in the ZIPs repository, so the scope is narrow.
    • The Halo 2 PR, preparatory work for ZSA, was approved by Daira Emma, with only final feedback from Kris and Str4d remaining.
    • The Orchard ZSA implementation is a large change and is being discussed with ECC, awaiting further feedback.
    • Daira Emma's quantum-resistance draft PR was also reviewed and comments left.
    • Considerable effort is going into catching up with the recent large changes in LibRustZcash.
  8. ZSA change propagation structure and the first Zebra PR

    QEDIT · Pablo

    • ZSA changes must propagate through the whole official stack, from note encryption to Orchard, librustzcash and Zebra, and also to txtool, which builds test transactions for Zebra.
    • The change replacing the asset description with a hash has been reflected in Orchard ZSA and is being reflected in Zebra and txtool; the Python reference implementation and test vectors will also be updated where needed.
    • The memo bundle is implemented on top of Orchard ZSA, can be done by QEDIT or ECC, and support in Zebra and txtool is needed separately.
    • A few hours earlier the team opened the first Zebra PR verifying v6 ZSA transactions, hoping for early review from ZF.
    • This PR is based on versions of Orchard ZSA and librustzcash ZSA not yet merged upstream, so it cannot be merged.
  9. NSM update: awaiting review and feature flags

    Shielded Labs, ECC, Zcash Foundation · Mariusz, Jason McGee, Pili, Str4d, Marek, Alfredo

    • Mariusz rebased the ZIP 233 implementation PR for Zebra, which needs the librustzcash PR to be merged first, and the librustzcash PR and the Zcash test vector PR are awaiting review.
    • Str4d said Kris had left a blocking comment that seems to have been addressed two weeks ago, and that they would put it back on the review list.
    • Once merged into librustzcash, it goes behind the zcash_unstable flag in the next crate release so it can be used downstream in unstable form.
    • Pili said that, without assuming it goes into NU7, they would merge the Zebra PR if it can be merged.
    • Str4d said the Zebra PR must be behind an unstable flag to avoid the risk of a mainnet consensus fork, and Marek confirmed that NSM is effectively safe because it is behind a feature flag.
    • When Alfredo suggested a dedicated NSM flag instead of the transaction v6 flag, Str4d said that for now a consistent flag that does not touch mainnet is enough.
  10. Discussion of when NU7 is decided

    ECC, Shielded Labs · Jason McGee, Str4d

    • Jason McGee asked whether the point at which the content of NU7 is decided is three to six months away.
    • Str4d answered that changes are locked once the mainnet activation height is set and a release containing that height is made, and that before that the broad outline is agreed before the testnet activation height is set.
    • Testnet can be rewound, so bugs found can be fixed, but they must be found before the mainnet activation height is set.
  11. Crosslink update: Milestone 2 complete

    Shielded Labs · Mark, Str4d

    • The second milestone is done, creating a one-way link from the PoS chain to the PoW chain.
    • After the BFT network votes on and pre-commits the hash of the PoW block it judges final, it is stored in the PoS block header.
    • The next milestone touches PoW chain consensus for the first time to complete the link in the other direction.
    • At the Zcash Global Discord workshop, real participants mined and produced blocks with the PoA roster.
    • The biggest feedback from staking partners of various sizes was the need for a tokenomics design for delegation, staking and slashing, which became the top priority.
    • Because this is the late part of an issuance schedule where 16 million have already been mined, there is no precedent, and the plan is to get feedback on the prototype with the most basic rules.

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