Crosslink staking design update
Crosslink adds a finality gadget on top of PoW to prevent chain reorganizations. This post explains how the staking design has changed since the previous post.
Shielded Labs updated the Crosslink staking design. Crosslink layers a rollback-preventing finality gadget on top of PoW; all economic activity and block validation continue on the PoW chain, and the gadget anchors blocks in parallel to prevent chain reorganizations. After the previous tokenomics post there was a lot of feedback that the original proposal was too rigid to support liquidity, with particular criticism that the 30-day unbonding and 10-day epoch were unnecessarily long.
The analysis found ways to shorten the epoch and unbonding periods while keeping the privacy and security benefits, and it says a 5-day epoch with 1-epoch unbonding is probably feasible. Staking takes place only within the Orchard pool, and unstaked funds go through 1 epoch (5 days) of unbonding before they can be used, so they are released 5 to 10 days later depending on when the request is made. For security and transparency, staked amounts are public so that total stake and the distribution across finalizers can be verified.
The design has mechanisms that balance privacy and usability. First, staking and unstaking are processed in batches at fixed 5-day epochs, which reduces information leakage and makes it harder to link to individuals, and stake can also be reallocated to other finalizers during this period. Second, staked amounts are quantized in powers of ten, such as 1, 10, 100 and 1,000. Third, the unbonding period gives time to detect and respond to attacks and lets participants leave some funds liquid in Orchard, strengthening the anonymity set.
On security, it explains response scenarios split into PoW compromise, PoS compromise and both compromised. As part of the incident response plan, there is an idea of preparing in advance an emergency network upgrade patch that temporarily freezes only the PoS part while PoW and normal transactions continue, which is one reason it considers 5-day unbonding sufficient. On privacy, it examined Penumbra-style shielded staking but left it out of V1 because of complexity and latency. Liquid staking tokens are also outside the scope of V1. It also moves away from Malachite and keeps its own lightweight Tendermint, Tenderlink.
The design also aims to reward long-term holders and have ZEC holders take part directly in consensus as Protocol Guardians running finalizer nodes. It admits that although liquidity was improved to attract new capital, the epoch windows and quantization mean wallets must think about usability, such as showing schedules. Shielded Labs plans to deliver a working testnet prototype by the end of the year, and said it will then check community and coinholder sentiment again to decide whether to proceed. This design is not final.
This summary is based on the original; check the original for exact details.