Lock-duration-weighted coinholder voting ZIP draft
arya2 published a coinholder polling ZIP draft (zips PR #1433) that gives more voting weight the longer coins are locked after the snapshot, and asked for feedback.
The author, arya2, supported coinholder polling and still does, but believes that, since it was designed as one voice among several, it is not robust against manipulation. Today one ZEC counts as one vote at the snapshot time, so coins can be moved right after the snapshot, and someone who buys in vote week and sells the next week can hardly be called a real holder.
The idea of the draft ZIP (zcash/zips PR #1433) is to give a vote more weight the longer coins are locked after the snapshot. The weight grows linearly up to a cap of about 4 years, and unlocked coins still count at one tenth so that nobody loses their existing voting power. Voting eligibility, voting subjects and quorum do not change.
The problem is shielded coins. Shielded notes have no timelock, so the only locks usable today are transparent locks, whose amount and release height are visible on-chain. So the ZIP accepts any lock the protocol can prove, uses transparent locks as a stopgap, and leaves shielded locks to a later consensus ZIP. Even the smallest shielded version needs a new circuit, so it is not small.
The author asked for opinions on three points: whether the period for which coins have already been held should count, whether the one-tenth floor and the 4-year cap are appropriate, and whether any polling operator would actually add the transparent path. This is not a proposal to apply it to ongoing matters; it is a post asking whether it is feasible. The author said AI was used to draft the ZIP and polish the post's wording.
This summary is based on the original; check the original for exact details.