Solana's first binding on-chain governance vote has not closed. Coverage across the market, including our own earlier report, said voting ended around 15:30 UTC on Thursday. A direct query to a Solana mainnet RPC node shows epoch 1023 was only 55.8% complete at 20:01 UTC that evening, with roughly 190,800 slots left to run. At the measured slot time of 365 milliseconds, the epoch closes around 15:20 UTC on Friday, August 28, a full day later than reported.
That is the smaller of the two problems. The larger one is that when the count does finish, Solana's own published rules do not agree on what would count as passing.
The Vote Is Still Running
The voting window was announced as lasting until the end of epoch 1023, which Solana developers expected to fall on "Thursday at approximately 15:30 UTC" (Solana Developers). Epoch boundaries are not clock time. An epoch is a fixed 432,000 slots, and how long that takes depends on slot production, which drifts with network conditions and skipped slots.
| Measurement | Value |
|---|---|
| Current epoch | 1023 |
| Progress | 241,196 of 432,000 slots (55.8%) |
| Slots remaining | 190,804 |
| Measured slot time | 365 ms, sampled over 1,646 slots |
| Projected epoch end | ~15:20 UTC, Friday August 28, 2026 |
| Widely reported close | 15:30 UTC, Thursday August 27, 2026 |
Measured at 20:01 UTC on August 27, 2026, against the public mainnet RPC endpoint. Anyone can reproduce this; the commands are in the methodology section below.
This explains the absence everyone noticed. No outlet has published a final tally, and the Solana Foundation has announced no outcome, because the vote has not ended. The reporting was not late. It was early.
Three Sources, Three Quorum Requirements
When the count does close, the question of whether anything passed depends on which Solana document you read.
What the Repository Says
The Solana Foundation's governance proposal repository is the canonical home of the SGP process. Its README sets out the voting policy in a table, and on the quorum question it is unambiguous:
Quorum: None. There is no minimum turnout. Approval threshold: Supermajority. For must be at least two-thirds (66.67%) of For plus Against stake. Abstain is not counted.
That is a direct reading of the solana-governance-proposals README, retrieved from the repository rather than quoted from secondary coverage. The same table sets the bar to submit an SGP at a validator vote account holding at least 100,000 SOL, and the support threshold to trigger a vote at 15% of active stake.
What the FAQ Says
Solana's governance FAQ and the voting dashboard describe something stricter: one-third of network stake must participate, counting For, Against and Abstain alike, and two-thirds of that participating stake must vote For.
The difference is not cosmetic. It adds a turnout floor that the repository explicitly disclaims, and it changes the denominator for the approval test from decisive votes to all votes cast.
What the Portal Displayed
A third figure was visible to voters during the vote. The frontend at governance.solana.com contained an incorrect quorum calculation path in its production JavaScript bundle, displaying a 60% quorum threshold. The defect was reported unresolved as of August 22 (CryptoSlate). Pull request 170 proposes the frontend fix, and GitHub issue 141 tracked the vote parameters.
The Display Bug Is the Least of It
CryptoSlate's reporting notes that "the display issue does not affect on-chain vote verification or the final tally." That is correct and worth stating plainly. A wrong number on a webpage does not change what validators signed. The durable problem is the disagreement between the FAQ and the repository, because that one determines how a correct tally should be interpreted.
The Same Tally, Two Outcomes
The most recent published figures for SGP-0002, from an August 26 Validator Info snapshot, show 83.66 million SOL For, 12.01 million Against and 8.32 million Abstain, roughly 104 million SOL in total, with support at 87.45% of decisive votes (CryptoSlate).
Note what the second panel shows. The approval threshold is met under both readings, at 87.45% of decisive votes and 80.45% of participating stake. Only the turnout requirement produces a different answer, and only one of the two documents imposes one.
The Turnout Number Nobody Has Published
Reporting so far has described participation as being "below the one-third line" without quantifying it. We calculated it.
| Input | Value | Source |
|---|---|---|
| Votes cast on SGP-0002 | 103.99M SOL | Aug 26 snapshot, via CryptoSlate |
| Total active stake | 436.9M SOL | Solana mainnet RPC, Aug 27 |
| Turnout | 23.8% | Our calculation |
| One-third quorum | 145.6M SOL | Our calculation |
| Shortfall under FAQ rule | 41.6M SOL | Our calculation |
Active stake was taken from a getVoteAccounts call: 687 current validators holding 432.9 million SOL, plus 10 delinquent holding 4.0 million. The two inputs come from different moments, an August 26 vote snapshot against an August 27 chain reading, so treat 23.8% as close rather than exact. Stake moves slowly; the conclusion does not turn on the second decimal.
Turnout may still rise. The vote has roughly 19 hours left as of this writing, and clearing 33.33% would require another 41.6 million SOL to vote, which is more than a third again of everything cast so far.
Why This Matters Beyond One Proposal
SGP-0001 asks validators to ratify a constitution and switch on the governance program that every future proposal will run through. It is being decided by the same process whose rules are currently ambiguous.
There is also a step after this one that is easy to miss. SGP-0002 sets policy direction. The actual consensus-affecting change still has to move through SIMD-0550 and ship in a client release. An accepted SGP is a mandate, not an activation.
Solana is not alone in having a rough first outing. Cardano's parallel votes failed on straightforward arithmetic, with 43% DRep support against a 67% requirement and 15.1% stake pool operator support against 51%. That is a clean failure. A rule conflict is a different category of problem: the tally is not in dispute, only what it means.
What to Watch
- The actual epoch close, around 15:20 UTC on Friday August 28 on current slot times, not Thursday.
- Whether turnout moves in the remaining window, and how close it lands to the 145.6 million SOL that the FAQ rule would require.
- Which rule the Foundation applies when it announces outcomes, and whether the FAQ or the repository is corrected to match the other.
- SGP-0001 and SGP-0003 tallies, which have not been published separately at any point in this vote.
The Bottom Line
- The vote is still open. Epoch 1023 was 55.8% complete at 20:01 UTC on August 27 and projects to close around 15:20 UTC on August 28.
- Three quorum figures were published: 60% on the portal (a frontend defect), one-third in the FAQ, and none in the repository.
- Approval is not the issue. SGP-0002 clears two-thirds under either reading, at 87.45% of decisive votes.
- Turnout is 23.8% of active stake on the latest published snapshot, 41.6 million SOL short of the FAQ's one-third floor.
- Whether it passed depends on which document is authoritative, and that has not been settled publicly.
We will report the outcome when the epoch closes and a tally is published, including which rule was applied.
Sources and Methodology
Reproduce the Epoch Finding
Both chain readings came from the public mainnet endpoint and can be repeated by anyone:
curl -s -X POST https://api.mainnet-beta.solana.com \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"getEpochInfo"}'
curl -s -X POST https://api.mainnet-beta.solana.com \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"getVoteAccounts"}'
The first returns epoch, slotIndex and slotsInEpoch. The second returns per-validator activatedStake in lamports, which we summed across current and delinquent validators. Slot time was measured with getRecentPerformanceSamples over the last ten samples.
Corrections
Our article published earlier on August 27 stated that the governance vote had closed at approximately 15:30 UTC that day, following the timing reported across the market. The chain data above shows that was incorrect: the vote remains open. We are noting the error here rather than quietly amending it.
Market Context
SOL traded at $109.09 with a market capitalisation of $63.68 billion and $7.30 billion in 24-hour volume, up 12.74% on the day, as of 19:58 UTC on August 27, 2026 (CoinGecko).
Sources
- solana-foundation/solana-governance-proposals - voting policy table, read directly from the repository README (retrieved August 27, 2026)
- Solana mainnet RPC,
api.mainnet-beta.solana.com- epoch progress, active stake, slot timing (retrieved 20:01 UTC, August 27, 2026) - CryptoSlate - August 26 SGP-0002 snapshot, the FAQ versus repository conflict, Cardano comparison (August 2026)
- CryptoSlate - the 60% portal display defect, PR 170, issue 141 (August 22, 2026)
- Solana Developers - voting window and expected close (August 24, 2026)
- CoinGecko - price, market capitalisation, volume (retrieved 19:58 UTC, August 27, 2026)
Disclosure and Limitations
This article is news analysis, not investment advice. The turnout calculation combines a vote snapshot and a chain reading taken roughly a day apart and is presented as approximate. Projected epoch end assumes slot times hold at the measured 365 ms; if production slows, the close moves later. We have not seen separate published tallies for SGP-0001 or SGP-0003 and make no claim about either.
Background on the three proposals, the thresholds and the opposition to SGP-0002 is in our full breakdown of the governance vote.