Key Takeaways
Zebra: Three releases landed since the previous Arborist Call. Versions 6.4.0 and 6.4.1 were recalled, with operators advised to upgrade to 6.4.2 following a high-severity denial-of-service vulnerability.
NU7: Work is moving into its final stages, including 25-second blocks, the Network Sustainability Mechanism, removal of V4 transactions, and a state database upgrade.
Light Clients: Zebra now includes an experimental lightwalletd-compatible gRPC server, allowing it to serve Zcash light clients directly through the CompactTxStreamer interface.
Zodl: New Orchard and librustzcash releases are being prepared for NU7, alongside Zallet, Ledger, multisig, performance, and formal verification work.
NSM: ZIP 235 and ZIP 237 have been merged into Zebra and Zakura, bringing the first deployment of the Network Sustainability Mechanism closer to mainnet.
Crosslink: A new FeatureNet includes implementations of all consensus components currently planned for the design, shifting the focus toward testing, refinement, and design review.
Fees: Following the move to a 1,000-zatoshi marginal fee, further reductions are being discussed, including a possible move to 400 zatoshis.
Development Process: The call closed with a broader discussion about shipping Zcash protocol improvements faster through smaller upgrades, stronger automation, and careful ecosystem coordination.
Zebra: Security, Performance & NU7
Three Zebra releases have landed since the previous Arborist Call: 6.4.0, 6.4.1, and 6.4.2.
Versions 6.4.0 and 6.4.1 have since been recalled, and node operators running either release should upgrade to Zebra 6.4.2, which contains the latest fixes and improvements.
The release addresses a high-severity, remotely triggerable denial-of-service vulnerability involving malformed V6 transactions. The issue was discovered through Google OSS-Fuzz and independently reported by v12security and SphereDonout.
Beyond the security fix, the 6.4.x series introduces several performance improvements aimed particularly at mining and block processing.
getblocktemplate is faster through pre-computed templates, while shielded proof and signature verification can now be cached. Block submission validation has also been improved, and transactions already verified in the mempool can avoid unnecessary re-verification after being mined.
Together, these changes reduce redundant work and improve the efficiency of block processing.
Zebra Light Client Support
Zebra now includes an experimental lightwalletd-compatible gRPC server, allowing it to serve Zcash light clients directly through the CompactTxStreamer interface.
This is an important infrastructure step because light-client functionality that traditionally depends on separate lightwalletd infrastructure can begin moving directly into Zebra.
The implementation remains experimental, but it lays groundwork for a simpler Zcash infrastructure stack.
NU7 Readiness
NU7 preparation is now moving into its final stages across Zebra and the broader Zcash development ecosystem.
Current work covers several of the upgrade’s major changes, including:
25-second target block spacing — ZIP 218
Network Sustainability Mechanism
Disallowing V4 transactions — ZIP 2003
State database upgrade to v29
Zodl is also preparing new Orchard and librustzcash crate releases, with multiple changes being merged ahead of NU7.
At the same time, work continues to stabilize testnet and coordinate NU7 activation across ecosystem development teams.
librustzcash & Zallet
Development around librustzcash and Zallet continues alongside the NU7 work.
Zodl is upstreaming performance improvements into librustzcash, together with ongoing formal verification work.
PCZT development is continuing around Ledger integration, while Zallet multisig support is also progressing. CI improvements are underway to strengthen the development and testing workflow.
A new Zallet release is expected as the updated NU7-compatible crates become available.
Network Sustainability Mechanism Heads Toward Mainnet
The Network Sustainability Mechanism (NSM) has reached a major implementation milestone ahead of NU7.
Two key components have now been merged into both Zebra and Zakura:
ZIP 235, which removes 60% of transaction fees from circulation
ZIP 237, which implements the halving-preserving issuance approach
This moves the NSM toward its first mainnet deployment as part of NU7 and marks an important milestone for Shielded Labs’ work on the mechanism.
The implementation follows the direction established through the recent NU7 polling and subsequent protocol discussions, turning the proposed monetary-policy changes into code for the upcoming network upgrade.
Crosslink Reaches a Feature-Complete FeatureNet
Crosslink has also reached an important development milestone.
Its persistent FeatureNet provides a live environment where participants can run infrastructure, experiment with the proposed finality design, and provide feedback, with real mainnet ZEC rewards available for participation.
A new FeatureNet has now launched with implementations of all consensus components currently planned for the Crosslink design, making it feature-complete in terms of the currently proposed consensus functionality.
Two notable additions are multi-phase activation logic and finalizer commission fees.
With the core components now implemented, the focus can increasingly shift from adding functionality toward refinement, testing, and design review.
User-Activated Slashing
The Crosslink FeatureNet is also providing a live environment for testing User-Activated Slashing (UAS).
UAS is a permissionless mechanism designed to penalize finalizers that misbehave or remain persistently unavailable through a coordinated hard fork.
Testing has already resulted in both failed and successful coordination attempts, providing practical lessons about how participants coordinate around slashing under real-world conditions.
These experiments are helping expose operational and coordination challenges before the design moves closer to production use.
Dynamic Fee Discussions
Zcash’s transaction fee policy continues to evolve.
Following the move to a 1,000-zatoshi marginal fee, participants are already discussing whether fees could be reduced further.
One idea raised during the call was lowering the marginal fee to 400 zatoshis, potentially beginning October 12.
This remains a discussion rather than a finalized network change. A major consideration is how wallets, nodes, and transaction-relay policies can remain coordinated as fee parameters evolve.
The discussion reflects a broader effort to lower transaction costs while ensuring that infrastructure across the ecosystem remains compatible with the new fee policy.
Can Zcash Ship Faster?
The call closed with a broader discussion ahead of Zcon: How can Zcash ship protocol improvements faster without sacrificing safety, quality, or ecosystem coordination?
One approach discussed was moving toward smaller, more frequent network upgrades, allowing features to ship when they are ready instead of waiting for many unrelated changes to be bundled into a larger upgrade.
The goal is not simply to increase the pace of releases.
Smaller and more clearly defined changes, combined with strong testing and careful coordination, could potentially increase development velocity while preserving the reliability expected from consensus-critical network upgrades.
Preparing for Zcash at Global Scale
Faster development also introduces a coordination challenge.
As the Zcash ecosystem grows, manually tracking every wallet, node implementation, exchange, infrastructure provider, and integration becomes increasingly difficult.
The discussion highlighted a longer-term need for more automated upgrade and coordination processes, reducing the amount of manual work required to safely deploy protocol changes across the ecosystem.
This becomes increasingly important if Zcash adoption expands significantly and the network moves toward operating as widely used, unstoppable internet money.
Looking Ahead
The October 1 Arborist Call showed several major efforts moving from development toward deployment and testing.
NU7 preparation is entering its final stages, while the Network Sustainability Mechanism is approaching its first mainnet deployment. Zebra continues to mature with security, mining-performance, and light-client infrastructure improvements, and Crosslink has reached a feature-complete FeatureNet for its currently planned consensus design.
At the same time, work on Zallet, librustzcash, Ledger integration, dynamic fees, and formal verification continues across the stack.
Beyond individual features, the discussion is also shifting toward a larger question for Zcash development: how to ship protocol improvements faster while maintaining the safety, reliability, and coordination required for consensus changes.
With NU7 approaching, that question is likely to become increasingly important for the upgrades that follow.