Orbs is a proof-of-stake network whose token holders have governed by signed off-chain vote since 2022, and which spent August 2026 formally constituting itself as a DAO. Its governance is worth reading closely for two reasons that have nothing to do with its size: it runs delegation by default, a design most token-weighted systems avoid, and its public proposal record has fallen years behind the votes it is supposed to record.

The DAO decides staking rewards, delegation rules, reserve-pool spending, and now protocol oversight. Everything below is read from the project's own Snapshot space, its OIP repository, and its governance blog.

Guardians vote unless you stop them

Orbs weights votes by staked ORBS, and stake is delegated to operators called Guardians. The distinctive choice is the direction of the opt: in the words of the OIP-9 ballot, "the default option is that the Guardians are entitled to vote on behalf of their entire delegated stake, unless a delegator opts out." A delegator who wants their own voice must take an action to reclaim it; one who does nothing has already voted, through someone else.

Most token-weighted systems run the opposite convention. Delegation elsewhere is opt-in and revocable: an undelegated holder is simply absent from the tally. Orbs converts absence into assent-by-proxy, which raises participating weight and lowers the number of signatures behind it. Both effects are visible in the record below, and both are the reason a headline vote count on this space cannot be read the way it would be read on Uniswap or ENS.

Voting itself is gasless: ballots are off-chain signed messages on Snapshot, verifiable on-chain if contested, with weight fixed at the block the proposal was created. The space resolves weight through a voting-proxy strategy over two Orbs delegation contracts, one on Ethereum and one on Polygon.

Nine proposals in four years

Governance opened on 11 August 2022 with the blog post "Introducing Orbs Governance," and the first ballot – reinstating a minimum self-delegation requirement for Guardians – followed on 23 August. Nine proposals have run since, read from the Snapshot space on 14 August 2026:

• Turnout is thin in signatures and heavy in weight. The nine proposals drew between 7 and 49 voting addresses. OIP-8 (December 2023 staking rewards) passed on seven signatures carrying 151.1m ORBS. Guardian-by-default delegation is what makes those two numbers compatible.
• Almost nothing is contested. Seven of the nine closed with zero weight opposed. The quorum bar is 100m staked ORBS including abstentions – roughly 10% of stake, and about 1% of the 10bn total supply.
• The exception is instructive. OIP-4 (December 2022), which proposed allocating 6.3m ORBS to holders hurt by the FTX collapse, is the only proposal ever rejected: 107.6m ORBS against to 60.2m for. The one time the DAO was asked to spend the treasury on a bailout with no protocol return, it said no – and the GitHub record still carries it, retitled "REJECTED".

OIP-9 and the founding of the DAO

OIP-9, "Establishing the Orbs DAO," ran from 30 July to 6 August 2026 and closed 351,971,661 ORBS for, 0 against, 0 abstaining, on 23 voting addresses – about 3.5% of total supply, comfortably over the 100m quorum. It creates the DAO's initial multi-sig wallets and delegates a first tranche of authority to it: network operations, Guardian oversight, major upgrades, and selected protocol parameters. Its own text is explicit that this is a first step rather than a finished handover, the progressive-decentralization framing that most protocols of this vintage now use.

One detail is worth correcting, because the wire coverage of the vote got it backwards. OIP-9 was widely described as the project's first formal community governance vote. It is the ninth. The proposal itself says so – "nearly four years since Orbs governance was officially launched" – and the eight ballots before it are public on the same Snapshot space. The vote also closed before much of that coverage ran.

The record stopped before the votes did

Orbs keeps its proposal archive as GitHub issues in orbs-network/OIPs, retitling each one "Accepted:" or "REJECTED:" once its vote settles. The convention held for OIP-0 through OIP-8. It has not been applied to OIP-9, which sat open and still titled "DRAFT" more than a week after passing unanimously, even though the ballot committed that "the Orbs OIP Github repo will be edited in accordance with the governance vote results."

The archive around it is quieter still. The repository contains one file – a README – and its last commit landed on 3 October 2022, so four years of proposals live entirely in issue titles. The governance blog, the other public record, stops at OIP-6 in September 2023; OIP-7, OIP-8 and OIP-9 were never written up there at all. Nothing here suggests a vote was mis-executed. It is the narrower and more common failure: a DAO whose decisions have outrun the record a newcomer would read to understand them, which is one of the quieter ways DAOs decay long before anything dramatic happens.

How Caper approaches this

A caper makes the opposite choice on both counts. There is no delegation at all: voting weight is (t · v) / (V · T), where v is a soulbound vote token minted one per ballot you cast yourself. Weight cannot be handed to an operator, and a holder who does not vote does not silently count as agreeing – their weight is zero, because the two terms multiply.

The record problem is handled by not keeping a separate one. Proposals, ballots and settlements are contract state, and a member's exit is sized by the same expression that decided the vote, so there is no archive to fall behind. That is a narrower guarantee than Orbs is reaching for – a caper governs one treasury, not a network with Guardians – but it removes the specific gap between what was decided and what a reader can find.