Issue #20 closed like this: “The deadline is September 1. There are roughly ten days left.”
Seven of those ten days have passed. This week, the votes really did come in. Pools voting yes on the Constitutional Committee renewal went from 32 to 179, and SPO support went from 8.07% to 35.47%. DReps moved from 37.58% to 64.24%. On the minPoolCost side, pools went from 60 to 103 and SPO support from 15.78% to 26.11%. The participation we described in #20 as “what is missing is not persuasion but participation” is happening.
And still, neither action has crossed its threshold. The Constitutional Committee renewal is 2.76 points short on DReps and 15.53 points short on SPOs. The deadline does not move. In Japan Standard Time it falls at 06:44:51 on the morning of September 2, and fewer than four days remain. The thing worth thinking about this week is not how many points are missing but where those points are sitting. Break the numbers apart and they all collect in a single box: not voted.
Executive Summary
- The votes came in. The Constitutional Committee renewal went from SPO 8.07% to 35.47% (32 → 179 pools) and DRep 37.58% to 64.24% (89 → 129 votes).
minPoolCostwent from SPO 15.78% to 26.11% (60 → 103 pools) and DRep 43.65% to 61.73% (97 → 125 votes). - The thresholds still stand. Ratification requires 67% of DReps and 51% of SPOs. The committee renewal is 2.76 points short on DReps and 15.53 points short on SPOs.
- The deadline is 21:44:51 UTC on September 1 — 06:44:51 JST on September 2. Intersect’s own figure and the epoch 652 end time returned by Koios agree to the second. Some write-ups give September 6; the primary source gives this timestamp.
- The missing share is sitting in “not voted.” To clear the SPO threshold on the committee renewal takes roughly 1,737,664,166 ada, which is 24.45% of the 7.1 billion ada of delegation that has not voted. On the DRep side it takes roughly 141,204,790 ada, or 8.46% of the DRep stake that has not voted.
- One action drew opposition; the other barely did. Pools voting no on
minPoolCostwent from 4 to 28. Pools voting no on the committee renewal are still just one. Same deadline, different reasons for being stuck. - The consequence of failure is now spelled out. Intersect writes that without ratification the Constitutional Committee falls from seven members to three, below the required minimum of five, leaving only info actions and update-committee actions available.
- “von Bergen” did not pass. The info action naming protocol version 12 reached its deadline in epoch 651 and expired.
- Leios published its Earth phase numbers. Leios carried 54% of everything reaching the chain, moving 18x the transaction count and 3.3x the bytes of real mainnet traffic, and the red team found that “none of it was a protocol failure.”
- NIGHT is now listed in Japan for the first time. On August 24, S.BLOX — a Sony group company — began handling NIGHT and ADA. For NIGHT this is the first listing by a registered Japanese exchange.
- The market went the other way from last week. ADA fell 10.52% on the week and NIGHT 11.38%, against BTC −0.97% and ETH −2.90%.
1. Market Pulse
Morning fixed-point readings from August 22 to August 29.
| Asset | 8/22 | 8/29 | Week |
|---|---|---|---|
| ADA | $0.226213 | $0.202420 | −10.52% |
| BTC | $78,459 | $77,696 | −0.97% |
| ETH | $2,511.72 | $2,438.85 | −2.90% |
| NIGHT | $0.02188096 | $0.0193916 | −11.38% |
The August 29 values for all four were taken together at 08:20 JST.
In #20 we wrote that “ADA’s +26.70% sits inside the movement of the market as a whole.” That relationship broke this week. BTC and ETH are roughly flat, and only ADA and NIGHT are down by double digits. Last week all four pointed up together; this week the group split in two.
What deserves care here is that a split does not hand us a reason. Cardano produced materials pointing in different directions this week at once: a governance deadline, a strong Leios result, and a first domestic listing for NIGHT. Which of those moved the price cannot be read off the price. The posture is the same one we took in #20, where we said governance progress could not be used to explain the price or the other way around.
The one thing that carries over from #20 is that NIGHT moved within a hair of ADA. Even in a week with a Japan-specific catalyst of its own, NIGHT’s weekly change stayed close to ADA’s.
2. The Deadline Is 06:44:51 JST on September 2
Let us start with the most operational item.
Articles are circulating with September 1 and with September 6. Read from Japan, a time-zone offset sits on top of that. For anyone deciding when to vote, the difference is a full five days, so we checked it as far as it can be checked.
Intersect’s weekly update #126, published August 28, states that both actions expire at “September 1, 2026 at 21:44:51 UTC (Epoch 653).”
That timestamp has independent corroboration. The end time Koios returns for epoch 652 is 1788299091 in UNIX time, which converts to 21:44:51 UTC on September 1, 2026 — matching Intersect to the second. So the deadline is the instant epoch 653 begins, which in Japan Standard Time is 06:44:51 on September 2.
The voting record confirms the same reading from another direction. The von Bergen action we tracked through last week carried an expiration recorded as “epoch 651.” Sorting all 190 votes cast on it by timestamp, the final vote landed at 21:20:09 UTC on August 22 — twenty-four minutes before epoch 651 began at 21:44:51 UTC. Not one vote arrived after that. An action whose deadline is written as “epoch N” closes the moment epoch N starts.
So for anyone weighing a vote right now, the real cutoff is the morning of Wednesday, September 2 in Japan, not Sunday night. That said, a vote has to land in a block as a transaction, so designing for the last minutes of the boundary is not the safe plan.
Intersect also notes this week that the SPO voting documentation on the Cardano Developer Portal has been updated to cover voting on several governance actions in a single transaction. The page explains that a vote is scoped to one action but a transaction is not: you create one vote file per action and put as many of them in one transaction as you like, and they are all witnessed by the same cold key signature, so voting on five live proposals costs exactly the same key handling as voting on one. With two actions sharing a deadline, that applies directly.
3. The Votes Came In — What Moved in One Week
The thing we pressed hardest in #20 was how far SPO support sat from the threshold. Here are this week’s numbers, captured at 08:17 JST on August 29, in epoch 652.
| Action | SPO yes | Pools (yes・no・abstain) | Issue #20 |
|---|---|---|---|
| Update Constitutional Committee 2026 | 35.47% | 179・1・4 | 8.07% (32・1・1) |
Reduce minPoolCost to 75 ada + Plutus memory limits (Part 2) | 26.11% | 103・28・37 | 15.78% (60・4・28) |
DReps moved in the same direction.
| Action | DRep yes (threshold 67%) | Votes | Issue #20 |
|---|---|---|---|
| Update Constitutional Committee 2026 | 64.24% | 129 | 37.58% (89 votes) |
minPoolCost and Plutus | 61.73% | 125 | 43.65% (97 votes) |
The committee renewal in particular added 147 yes pools in a week, and the percentage rose 4.40x. DReps climbed more than 26 points, from 37.58% to 64.24%. The path we described in #20 as “not natural growth” has now appeared two weeks running.
The shape of the increase needs a note again. On the committee renewal, the pool count rose 5.59x while the percentage rose 4.40x. On minPoolCost it was 1.72x on pools against 1.65x on the percentage. The reason is the same as in #20: ratification counts delegated weight, not ballots. But note that in #20 the minPoolCost figures ran the other way — 2.5x on pools produced 6.00x on the percentage. The pools that moved this week carried less delegation each than the pools that moved last week. How many pools moved does not translate into how much closer the action got.
Outreach moved in the same week. Intersect covered both actions in its August 28 DRep monthly call, with members of the Technical Steering Committee and the Parameters Committee taking part, and its August 25 X Space put governance actions on the agenda alongside the Board and Constitutional Committee elections. We are not asserting causation; the pattern of outreach and movement landing in the same week is the same shape as #20.
4. And Still Short — 2.76 Points and 15.53 Points
This is the heart of the week. “It went up” is not the end of the sentence.
| Action | Body | Now | Threshold | Short by |
|---|---|---|---|---|
| Update Constitutional Committee 2026 | DRep | 64.24% | 67% | 2.76 pt |
| Update Constitutional Committee 2026 | SPO | 35.47% | 51% | 15.53 pt |
minPoolCost and Plutus | DRep | 61.73% | 67% | 5.27 pt |
minPoolCost and Plutus | SPO | 26.11% | 51% | 24.89 pt |
Ratification requires both bodies to clear their thresholds. One alone does not pass the action.
Converting the shortfall into delegated stake makes the picture more concrete.
| Action | Body | Shortfall (stake) | Stake that has not voted | Share of non-voting stake needed |
|---|---|---|---|---|
| Committee renewal | DRep | ~141,204,790 ada | ~1,668,373,863 ada | 8.46% |
| Committee renewal | SPO | ~1,737,664,166 ada | ~7,107,843,373 ada | 24.45% |
minPoolCost | DRep | ~264,116,858 ada | ~1,539,720,568 ada | 17.15% |
minPoolCost | SPO | ~2,767,536,634 ada | ~7,705,950,866 ada | 35.91% |
Laid out this way, it is clear where the missing share is sitting. For the committee renewal, 8.46% of the DRep stake that has not voted would carry it over the line — slightly less than a tenth. On the SPO side, a quarter of the non-voting delegation would do it.
In other words, what is holding these actions is not opposition. It is delegation that has been left unvoted and therefore stays in the denominator. In the ratification calculation only explicit abstention leaves the denominator. Doing nothing keeps the stake in the denominator and counts it as “not yes.”
In #20 we wrote that “what is missing is not persuasion but delegation taking its seat.” This week 147 pools took a seat. It is still 35.47%. The seats still empty outnumber the ones that filled.
5. One Action Drew Opposition, the Other Barely Did
The two actions share a deadline and share thresholds, but they are not stuck for the same reason.
Pools that have cast a no vote on minPoolCost went from 4 in #20 to 28 this week — 4.12% of the delegated weight, or roughly 458,047,686 ada. DReps have added 26 no votes (4.58%). In the week the yes votes arrived, no votes arrived too.
On the committee renewal, the number of pools voting no is still exactly one (0.54%), and DReps have four (0.24%).
That difference maps onto what the two actions are. Lowering minPoolCost from 170 ada to 75 ada shifts the economics differently for different pools. And as we noted in #20, this action bundles an increase to the Plutus memory limits alongside the fee change, with Intersect writing that voters should weigh both. There is structural room for genuine disagreement.
The committee renewal is different. As the next section sets out, failing to pass it shrinks the governance procedure itself. That opposition amounts to a single pool tells us this is not an action people are split on. It sits at 35.47% not because of division but because of absence.
Even with the same four days left, minPoolCost carries a question of agreement and the committee renewal is almost purely a question of participation. Reading them as one thing loses that.
6. If It Fails, the Committee Goes From Seven Members to Three
At the time of #20, what we could establish was Intersect’s line that failure “could interrupt governance continuity.” This week that was given a number.
Intersect’s weekly update #126 states that if the action is not ratified before the current members’ terms expire, the Constitutional Committee will fall from seven members to three. Three is below the required minimum of five. In that state, the update explains, only info actions and update-committee actions remain available.
Put differently, failure does not simply mean “this proposal was rejected.” The machinery for bringing the next proposal narrows first. Treasury withdrawals, parameter changes, hard fork initiations — every action type that needs Constitutional Committee approval stops moving. What is left is the non-binding info action and the action that rebuilds the committee.
The Dijkstra timeline is directly affected. In #20 we wrote that “Dijkstra’s timeline carries a risk note for the first time,” and traced that risk to node release candidates. This week the stated reason has changed. Update #126 writes that timelines currently face a potential risk of delay because of the ongoing on-chain vote to update the Constitutional Committee. In one week the subject of the risk moved from the node side to the governance side.
One more asymmetry matters as the days count down: of the two actions, the committee renewal is the harder one to recover from. If minPoolCost fails, the status quo of 170 ada simply continues. The committee has no status quo to fall back on, because terms do not wait.
7. “von Bergen” Did Not Pass
The action we described in #20 as being at its deadline has an outcome.
The info action naming the protocol version 12 hard fork “von Bergen” reached its deadline in epoch 651 and expired. In the ledger record it expired in epoch 651 and dropped out of the proposal list in epoch 652.
At the time of #20 it stood at 59.5% DRep support (117 votes) against a 100% requirement for info actions. With 36.14% not voted, no path to 100% remained, and we said as much. The ending is the expected one.
What is worth holding onto is that this did not fail because of opposition. Votes cast against it numbered two (0.04%). When the requirement is 100%, a single unvoted stake is enough. An action that only assigns a name failed on participation.
How the hard fork name gets handled from here does not appear in this action’s record. Whether it is resubmitted or settled another way is a question for the next batch of material.
8. Leios Published Its Earth Phase Numbers
On August 27, IOG published a consolidated report on the Earth phase of the Leios testnet MusashiNet. #20 carried figures that came through the weekly development report; this is the results write-up.
The main measured values:
| Item | Value |
|---|---|
| Duration | 41 days from the June 23 launch |
| Blocks | more than 127,000 |
| Endorser blocks announced | 30,000 |
| Certificates carried on chain | nearly 8,000 |
| Participating pools | 3 → 63 registered (peak of 40 producing blocks in a single epoch) |
| Throughput | 26.8 TxkB/s, against the 4.51 TxkB/s mainnet reaches with Praos alone |
| Releases | 12 in eight weeks |
Three numbers are new.
First, in the stable final days of the phase, Leios carried 54% of everything reaching the chain. Until now the figures described total output; this one describes what share of it came through Leios. It is over half.
Second, the comparison with real mainnet traffic. Against actual mainnet over the same period, Earth moved 18x the transaction count and 3.3x the bytes. The transaction multiple being the larger of the two tells us the test transactions were lighter than mainnet’s. Capacity to clear many transactions and capacity to clear heavy ones are different things, so the two multiples should be read separately.
Third, the red team result. Against attempts to find weaknesses, IOG writes that “none of it was a protocol failure,” and that performance degraded in proportion to stake rather than breaking the chain. It bent instead of snapping.
And the Water phase has opened on a fresh chain. Where Earth validated the design, Water varies parameters and measures behavior. IOG states the objective as a mainnet-ready Leios by the end of the year.
The follow-on to what we wrote in #20 — that the ledger had begun carrying BLS keys — did not appear in this week’s development report. Whether SPO-facing key registration steps get documented remains a watch item for next week.
9. Core Protocol Work Moves to BlockPQR
On August 26, IOG announced that part of Cardano’s core protocol development is now carried by a new firm, BlockPQR.
Four workstreams moved:
- Ledger — transaction validation rules and node engineering
- Blockchain Indexer — off-chain data querying for wallets and dApps
- Formal Methods — machine-checkable specifications
- High Availability — block production infrastructure and monitoring
BlockPQR was founded and is led by Leonard Hegarty, described as the former director of technical delivery for the Cardano blockchain at IO, who held responsibility for infrastructure readiness during the van Rossem upgrade. The entity changes; the hands do not — that is the core of the announcement.
Structurally, IO Labs continues to support delivery and holds primary responsibility for the initiatives and milestones. The work happens in open repositories, and Intersect gates each milestone and keeps oversight of long-term continuity. IOG states that the protocol is already in production and continues to operate exactly as it does now.
From an SPO’s seat, what matters is that the number of places responsible for things has grown. More than one entity writes node code, and milestone approvals stitch them together. This runs in the same direction as node implementation diversity, of which Amaru is the visible case, and IOG has published a comparable transfer to ICAN Group. Spread implementations out and the impact of any one company stopping gets smaller. On the other hand, who guarantees what becomes harder to see. Since Intersect holds the gate, the governance blockage described above lands on this structure too.
On node diversity itself, this week gave dates. Amaru met its Q3 goal of delivering a relay node binary and is targeting block production on mainnet in Q4 2026, with a block-producing pre-release targeted for late September. Node diversity workshops are set for Singapore on October 6 during Token2049 and London on November 12–13, with SPOs among the participants.
10. The Constitution Portal — the Test Submissions Are Gone
In #20 we wrote that the “six in consultation” shown on the Constitutional Amendment Portal (cap.intersectmbo.org) was really four, because the rest were posts made to exercise the portal itself.
That has now been cleaned up. Five items sit in consultation, and the apparent test submissions are gone from the list. The caveat we gave in #20 — that reading the on-screen “6” as a proposal count misstates the situation — is not needed this week.
The breakdown as of August 29:
| # | Type | Class | State | Title | Filed | Comments |
|---|---|---|---|---|---|---|
| 9 | CAP | Procedural | in consultation (author-ready) | Remove immutable storage requirement for governance metadata | 8/23 | 4 |
| 4 | CIS | Procedural | in consultation (author-ready) | The Bundling Gap in Constitutional Amendment | 8/07 | 5 |
| 3 | CAP | Substantive | in consultation | Evolve the Net Change Limit into a Strategic Spending Plan | 8/06 | 11 |
| 2 | CAP | Procedural | in consultation | Remove Article II section 7.4 | 8/06 | 2 |
| 1 | CAP | Substantive | in consultation | Require Milestone-Based Treasury Disbursement and Return of Undistributed Funds | 8/06 | 2 |
The new arrival is CAP #9. Filed on August 23, it reached author-ready in five days. It proposes removing the clause that requires governance action metadata to be hosted on immutable storage.
The reasoning is practical and worth reading. The on-chain hash is already a cryptographic commitment to the final version of the document. If the document behind the URL is later altered, its hash no longer matches and the mismatch is detectable. Requiring the storage itself to be immutable therefore adds no protection. Meanwhile content-addressed storage such as IPFS does not guarantee availability once nobody pins it, while ordinary mutable web hosting can stay reliably reachable. So the Constitution should specify the property that matters — cryptographic integrity of the final document — rather than a characteristic of the storage mechanism.
For anyone submitting governance actions this is a procedural question directly. It is also a more general one: how far into implementation detail should constitutional text reach? That runs in the same direction as CIS #4, which we have tracked since #19 and which asks whether one proposal may carry several changes. Two author-ready items now sit side by side.
Intersect also records this week that the CAP phase 2 deliverable was approved by the Civics Committee on August 20. The portal has entered alpha operation, and the working group’s regular calls pause in favor of biweekly public CAP calls.
11. Midnight Watch
Midnight produced material this week that lands directly for readers in Japan.
NIGHT Is Handled in Japan for the First Time
On August 24, S.BLOX K.K. began handling NIGHT and ADA. The press release states plainly that NIGHT is being handled for the first time domestically. The “first in Japan” applies to NIGHT; ADA started at the same time on this venue but has been handled by other domestic exchanges before.
S.BLOX, formerly Amber Japan, is a Sony group crypto asset exchange operator (Kanto Local Finance Bureau registration No. 00016). To mark the listing, campaigns worth up to ¥14,000 in NIGHT and up to ¥10,000 in ADA run through September 30.
One piece of framing. When crypto assets are used in Japan, whether a registered operator handles the asset is a channel separate from the technology. For a token belonging to a Cardano partner chain, one entrance to that channel has now opened in Japan.
The release introduces Midnight as a “privacy-focused blockchain.” More precisely, Midnight uses zero-knowledge proofs to let applications design how much they disclose, and NIGHT is the capital asset that generates DUST on top of that. The official description is that 1 NIGHT yields up to 5 DUST, and that DUST is a non-transferable resource used to execute contracts. That is a different thing from NIGHT being a coin whose purpose is to hide transactions.
State of the Network, August 26
Midnight published its monthly state report on August 26. Picking out what carries numbers:
- VIA Labs’ universal cross-chain messaging is live on Midnight, connecting to over 130 public and private networks and letting contracts talk to contracts directly without wrapped assets.
- Midnight Node 2.1.0-beta.1 has been released to Preprod and Mainnet.
spec_versionmoves from 1_000_003 to 2_001_000 andtransaction_versionfrom 3 to 4. Midnight writes that enacting it is a ledger 8 → 9 hard fork with an on-chain multi-block migration. - DUST is described as running a four-phase cycle — generation, constant capacity, decay and empty — with holders able to update or revoke registration at any time.
- The rationale for sponsorship is stated as most decentralized applications losing up to 80% of users during onboarding; separating NIGHT from DUST lets developers cover users’ transaction costs.
- The Midnight Buildathon on AKINDO carries $12,500 in grant funding across three waves over three months.
For operators, the weight is less in the node version rising than in the fact that it is designed as a ledger hard fork. With the migration described as running on-chain across multiple blocks, the timing of enactment is something to follow.
Correction — the MIP Process Was Running
In #19 and #20 we wrote that we had not been able to confirm a first proposal following the MIP (Midnight Improvement Proposal) process.
That was our own confirmation failing to reach, and we are correcting it.
Reading the public repository directly (measured August 29), mips/ holds 15 documents, MIP-0001 through MIP-0015, plus a template, and mps/ holds 38 files. The structure runs on two tracks: MPS (Midnight Problem Statements) describes what needs solving in a solution-agnostic way, and MIPs propose how to change it. The stages are defined as Proposed → Review → Accepted → Implemented.
It moved this week too. On August 25, MIP-0015 (Wallet-Derived Deterministic Secrets), MPS-0036 (Security-Review Evidence for Compact Contract Releases) and MPS-0037 (Heritable Rights for Self-Replicating Off-Chain Assets) were merged. On August 28 a new MIP for a Midnight CLI tool was opened.
Two lines in the repository’s rules are worth recording. One is that signed commits are mandatory — commits whose signature cannot be verified cannot be merged. The other concerns AI: drafting and editing with AI tools is welcomed, but the commit must be authored by the responsible human contributor. Writing the rule as a question of where accountability sits, rather than whether the tool may be used, is the notable part.
We were able to correct the earlier statement because we changed where we looked. Official blogs and announcements alone leave you at “cannot find it,” while the development record was sitting in the repository. Get the primary location wrong and things that are moving stay invisible. We are recording that against ourselves as well.
12. Risk Dimensions
Five points from this week’s material worth holding onto.
1. “Moved a lot” and “will make it” are different claims. The committee renewal rose 26.66 points on the DRep side in a week. But with under four days left it needs 2.76 more from DReps and 15.53 more from SPOs. If this week’s rate repeated it would clear, but votes do not arrive at a constant rate. There is no guarantee that whatever produced the late-week climb repeats over the next four days.
2. Both actions have to pass. Each requires both DReps and SPOs to clear their thresholds. Reading “almost there” off one action misreads the board. minPoolCost in particular is 24.89 points short on SPOs — the further of the two.
3. The stated deadline varies between write-ups. The primary source (Intersect #126) gives 21:44:51 UTC on September 1; other write-ups give September 6. That is a five-day difference for anyone voting. Check where the date comes from before you act on it.
4. This week’s price cannot be explained with Cardano-specific material. ADA −10.52% against BTC −0.97% and ETH −2.90%. Unlike last week, only ADA and NIGHT fell hard. But the size of the fall does not let us reason backwards to a cause. In a week that also produced a first domestic listing, simple attributions will miss.
5. Midnight’s node update carries a ledger hard fork. More than the version number rising, the point is that enactment brings a ledger 8 → 9 migration. It is described as running on-chain across multiple blocks, so the timing and procedure need to be followed in the next update.
13. What to Watch Next Week
- The outcome of both actions — the deadline is 06:44:51 JST on September 2. Ratified or expired. Issue #22 starts there.
- Operating with a three-member committee — if the action expires, only info actions and update-committee actions remain. How quickly a resubmission appears decides how long the practical gap lasts.
- Dijkstra’s timeline — the stated reason for risk moved to governance this week. How it reads in the next weekly update once the two actions resolve.
- Amaru’s pre-release — a block-producing pre-release is targeted for late September. Whether it lands.
- The Leios Water phase — what gets varied on the fresh chain and how far it is measured, against the year-end mainnet-ready objective.
- CAP #9 — whether it moves past author-ready, and where it and CIS #4 stall or do not.
- Midnight’s ledger 8 → 9 — the timing of enactment, and what SPOs and node operators are asked to do.
- Intersect Board Elections 2026 — applications open August 31 at 12:00 UTC for two member-elected seats. How many stand.
Summary
Issue #20 ended like this: “The deadline is September 1. There are roughly ten days left.”
Across seven of those ten days, the votes came in. Pools voting yes on the Constitutional Committee renewal went from 32 to 179, and DReps from 37.58% to 64.24%. As one week’s movement, it is large even against the previous week’s 30.12% to 37.58%.
And still: 64.24%.
The requirement is 67%. The gap is 2.76 points — about 141,204,790 ada of delegation, or 8.46% of the DRep stake that has not voted. On the SPO side it stands at 35.47% against 51%, and a quarter of the non-voting delegation would carry it.
The clearest thing this week is that these two actions are not divided. Pools voting no on the committee renewal still number one. DReps voting no number four. Opposition is almost absent, and the threshold is still not met. The reason is not disagreement; it is the size of the delegation left unvoted inside the denominator.
And this week the consequence of failure was made concrete. The committee falls from seven members to three, below the minimum of five. What remains at that point is info actions and update-committee actions. What gets rejected is not the proposal; the machinery for bringing the next one narrows first.
In the same week, Leios carried 54% of everything reaching the chain and came through its red team without a single protocol failure. Core protocol work moved to BlockPQR with Intersect gating the milestones. In Japan, a token from a Cardano partner chain appeared on a registered exchange’s list for the first time. The people building it and the people delivering it are moving forward.
And the machinery that approves how they move forward is stuck 2.76 points short.
The deadline is 06:44:51 on the morning of September 2, Japan Standard Time.
Sources
- Intersect MBO — Intersect weekly update #126 (2026-08-28; the two deadlines and vote status, the committee falling from seven to three, the stated reason for Dijkstra timeline risk, Amaru’s targets, node diversity workshops, the SPO voting documentation update, CAP phase 2 approval, Board Elections 2026): https://www.intersectmbo.org/news/intersect-weekly-update-126-aug-28-2026
- Cardano Developer Portal — SPO governance voting, including voting on several actions in one transaction: https://developers.cardano.org/docs/operators/governance/spo-governance/
- Input Output — Leios hits 6x Cardano’s throughput on its first testnet (2026-08-27; Earth phase measurements, 54%, 18x and 3.3x, red team results, Water phase opening): https://www.iog.io/news/leios-hit-6x-cardano-s-throughput-on-its-first-testnet
- Input Output — Cardano’s core protocol work continues in the same hands, now at BlockPQR (2026-08-26; the four workstreams, the arrangement, Intersect’s gate): https://www.iog.io/news/cardano-core-protocol-work-continues-at-blockpqr
- Essential Cardano — Weekly development report as of 2026-08-28 (Leios prototype fixes, db-analyser extension, Plutus 1.68.0.0, Hydra performance work): https://www.essentialcardano.io/development-update/weekly-development-report-as-of-2026-08-28
- Intersect Constitutional Amendment Portal (the five items in consultation as of 2026-08-29 and the content of CAP #9): https://cap.intersectmbo.org
- Cardano governance actions (
minPoolCost, the Constitutional Committee renewal, von Bergen): https://gov.tools/governance_actions - Koios REST API —
tip/epoch_info/epoch_params/proposal_list/proposal_votes/proposal_voting_summary/totals(epoch boundary times, proposal state and expiry, vote timestamps): https://api.koios.rest/api/v1 - AdaStat (delegation-weighted support and the ratification denominator): https://adastat.net
- S.BLOX K.K. — press release on beginning to handle NIGHT, a first in Japan, and ADA (2026-08-24): https://prtimes.jp/main/html/rd/p/000000092.000042886.html
- Midnight — State of the Network, August 2026 (2026-08-26; DUST mechanics, VIA Labs, Node 2.1.0-beta.1 and the ledger 8 → 9 hard fork, the Buildathon): https://midnight.network/blog/state-of-the-network-august-2026
- Midnight Improvement Proposals repository (15 MIPs and 38 MPS files, this week’s merges, contribution rules): https://github.com/midnightntwrk/midnight-improvement-proposals
- CoinGecko (ADA / BTC / ETH / NIGHT values for August 22 and August 29): https://www.coingecko.com
Transparency Note
Material that surfaced this week but did not make the article, and the judgment calls behind it.
- How the deadline was established. We took the primary figure from Intersect #126 — “September 1, 2026 at 21:44:51 UTC” — and confirmed that it matches the epoch 652 end time returned by Koios to the second. We separately read every vote on von Bergen, whose deadline was recorded as epoch 651, and confirmed that voting stopped 24 minutes before epoch 651 began. We did not infer the calendar date from the epoch record alone.
- Difference from Intersect’s figures. Update #126 gives SPO 35.3% on the committee renewal as of 15:30 UTC on August 28; this article gives 35.47% as of 08:17 JST on August 29. The gap is capture time, not a contradiction.
- The Ikigai action’s support level. Of the two actions expiring in epoch 656, for
Reimburse Ikigai Info Governance Action Depositour retrieval path found a divergence between sources beyond tolerance, so we could not establish our own figure (two attempts). Intersect’s weekly update records DRep support at 44.47%; we do not publish a number of our own. Governance Incentives Framework 2026. DRep support stands at 1.42% (10 votes) while votes cast against have grown to 37.73% (58 votes), up from 25 no votes in #20. Its deadline is epoch 656, still ahead, so we left it out of the body.- Why the votes moved so much in a week. The voting record establishes that votes increased and nothing more. We placed the outreach that happened in the same week alongside it without asserting causation.
- Why 28 pools voted against
minPoolCost. Not readable from the votes themselves, so we state the fact only. - What happened to the portal’s test submissions. We measured that they are gone from the list, but we have not determined whether they were deleted or hidden. No explanation from the portal side was found.
- When Midnight’s ledger 8 → 9 takes effect. Not stated in the official report, and not stated here.
- The Cardano Foundation CTO departure. Effective August 31, but first reported on August 10, so not this week’s material.
- Hydra, Plutus and Lace improvements this week. Present in the weekly development report but distant from this issue’s subject, so omitted.
This is not investment advice. For information and research purposes only.