The previous issue (#22) closed like this: “If it is resubmitted, the question will be how to assemble 51% on the SPO side.”
The answer arrived at 5:27 in the morning of the day this issue was written. The action to reduce minPoolCost to 75 ada has been resubmitted. But the contents are not the same as last time.
The Plutus memory limits are gone from the action. And because they are gone, this action has no SPO vote among its ratification requirements.
Last week, the only side that stopped this action was the SPO side. This week, that column no longer exists. That is where we start.
Executive Summary
- The
minPoolCostaction was resubmitted: at 5:27:12 a.m. JST on 12 September. That is 1 hour, 17 minutes and 39 seconds before the epoch 655 boundary (6:44:51 a.m. the same day). It expires in epoch 661, 6:44:51 a.m. JST on 12 October. - “Plutus Memory Limits” is gone from the title: the old action was a bundle that moved
maxTxExecutionUnitsandmaxBlockExecutionUnitsalongsideminPoolCost. The new action’sprotocol_param_updatecontains a single entry:min_pool_cost. - As a result, the SPO vote drops out of the ratification requirements: CIP-1694’s security-relevant parameter group includes
maxBlockExecutionUnits, but it does not includeminPoolCost. Our canonical data path also returns an SPO threshold of 0% for this action. - Three votes so far: DRep yes 0.54% (3 votes, threshold 67%). All seven committee members are unvoted. Half a day after submission, this is only the beginning.
- The committee changed over on the morning of 7 September:
enacted_epochis 654, i.e. 6:44:51 a.m. JST on 7 September. There are seven members: four with terms to epoch 799, three to epoch 726. Departing: KtorZ and the Cardano Japan Council. - One vote that had already been cast stopped counting: the committee “no” on
Governance Incentives Framework 2026appears 3 times in the itemized record and 2 times in the aggregate. The member who cast the difference is not among the current seven. - The new committee has cast no votes in five days: across all four live actions, committee votes recorded since the start of epoch 654 total zero. SPO votes in the same window are also zero.
- Two actions expire in five days:
Reimburse Ikigai Info Governance Action Deposit.sits at DRep 58.37% (threshold 67%) with 2 of 7 committee members in favour.Governance Incentives Framework 2026sits at DRep 1.62%. Both expire at 6:44:51 a.m. JST on 17 September. - The constitutional amendment did not arrive by its stated target: cardano.org wrote on 10 July that submission was “targeted before epoch 655”. Epoch 655 began this morning. Only two actions were submitted in epochs 654 and 655, and neither is a constitutional amendment.
- Hydra’s critical came from a fast path that skipped verification: when processing a snapshot request, hydra-node had a path that skipped signature verification and Plutus script evaluation, and unvalidated transactions could reach it. 2.4.1 restores full validation.
- cardano-node 11.1.1 is not for production: the release notes state plainly that it is a pre-release not yet recommended for use on mainnet. Intersect expects 11.2 within a month, with the full Dijkstra feature set.
- The market split from last week’s direction: ADA −2.46% and NIGHT −1.16% on the week. BTC is −3.11%, and only ETH is up, at +2.85%.
1. Market Pulse
Fixed-time morning readings from 5 September to 12 September.
| Asset | 9/5 | 9/12 | Week |
|---|---|---|---|
| ADA | $0.210854 | $0.205667 | −2.46% |
| BTC | $79,675 | $77,199 | −3.11% |
| ETH | $2,452.01 | $2,521.90 | +2.85% |
| NIGHT | $0.02111984 | $0.02087497 | −1.16% |
Last issue reported “all four assets up”. That shape broke this week: only ETH rose. ADA and NIGHT fell modestly, and BTC fell further than either.
Taken exactly at the epoch boundaries, ADA moved more. The epoch 653 boundary (morning of 7 September) was $0.22181616; the epoch 654 boundary (this morning) was $0.20642490, a move of −6.94%. In yen, ¥34.63 to ¥31.70 — note that USD/JPY also moved from 156.12 to 153.58 over the same span, so part of the yen-denominated fall is that.
Some network figures as well. The staking rate went from 56.54% to 56.49%. Pools that produced a block rose from 921 to 953, and delegator count fell by 975, from 1,338,967 to 1,337,992. The treasury stands at 1,356,415,272.91 ada.
2. The Action Came Back at 5:27 This Morning
Koios’s proposal_list contains one ParameterChange submitted in epoch 654.
Its title is Reduce minPoolCost to 75 ada.
The recorded time (block_time) is 1789158432, which converts to 20:27:12 UTC on 11 September 2026, or 5:27:12 a.m. JST on 12 September. Epoch 655 began at 1789163091, so the gap is 4,659 seconds — 1 hour, 17 minutes and 39 seconds.
The previous issue described a final vote landing 69 seconds before a deadline. This week, the next action was placed 1 hour and 18 minutes before an epoch changed. Both are the same shape: activity concentrating just before a boundary.
It expires in epoch 661, which begins at 6:44:51 a.m. JST on 12 October. There is a month.
3. What Came Out Was the Memory Limits
The previous action was titled:
Reduce minPoolCost to 75 ada and increase Plutus Memory Limits (Part 2)
This one is titled:
Reduce minPoolCost to 75 ada
The second half is gone. And this is not only a matter of the title.
Here are the on-chain parameter updates side by side.
| Old action (expired) | New action | |
|---|---|---|
min_pool_cost | 75,000,000 | 75,000,000 |
max_tx_ex_mem | 17,500,000 | — |
max_tx_ex_steps | 10,000,000,000 | — |
max_block_ex_mem | 77,500,000 | — |
max_block_ex_steps | 20,000,000,000 | — |
The new action’s protocol_param_update holds a single entry: min_pool_cost. Its abstract says so explicitly:
“This governance action proposes reducing
minPoolCostfrom 170,000,000 Lovelace (170 ada) to 75,000,000 Lovelace (75 ada), a decrease of approximately 55.9%. No other protocol parameters are changed by this action.“
The size of the reduction and the reasoning are the same as before. The action cites PCP-006 (authored by Cerkoryn, published on the Cardano Forum on 30 March 2026) and records that Intersect’s TSC ratified the reduction on 9 July 2026.
What changed is only that the memory limits that had been in the same bundle came out.
4. minPoolCost Has No SPO Column
This is the most important point of the week.
CIP-1694 marks a subset of the protocol parameters as security-relevant. An action that changes one of them requires an additional vote of the SPOs, on top of DReps and the committee.
The list reads:
maxBlockBodySize,maxTxSize,maxBlockHeaderSize,maxValueSize,maxBlockExecutionUnits,txFeePerByte,txFeeFixed,utxoCostPerByte,govActionDeposit,minFeeRefScriptCostPerByte
And CIP-1694 states:
“any proposal attempting to change such a parameter requires an additional vote of the SPOs.”
Two things follow.
maxBlockExecutionUnitsis on that list. The old action’smax_block_ex_memandmax_block_ex_stepsare exactly that parameter.minPoolCostis not on that list.
In other words, the old action required an SPO vote not because of minPoolCost, but because of the memory limits bundled with it. With the memory limits removed, the SPO vote drops out of the new action’s ratification requirements.
The measurements agree. Our canonical data path (AdaStat as primary, Koios as cross-check) returned an SPO threshold of 51% for the old action. For the new one it returns 0%.
Let us be explicit about what can and cannot be written here. We do not write that the proposer removed the bundle in order to avoid the SPO vote. The primary sources do not state a reason. What can be written is the consequence: with the bundle removed, that is how the requirement stands.
The consequence is clear enough. To use the previous issue’s own framing:
Last week the action fell because DReps were at 68.57%, over the 67% threshold, while SPOs were at 34.5%, short of 51%. This week, that SPO column is gone.
The same 75 ada now passes only through the gate it was already passing.
5. Three Votes So Far
Koios’s proposal_voting_summary for the new action, retrieved at 08:20 JST on 12 September 2026.
| Category | Current | Threshold |
|---|---|---|
| DRep yes | 0.54% (3 votes) | 67% |
| DRep voted no | 0.0% (0 votes) | — |
| DRep always-no-confidence | 2.68% | — |
| DRep not voted | 96.78% | — |
| SPO yes | 0.0% (0 votes) | 0% (not required) |
| Committee | 0 yes / 0 no / 7 unvoted | 2/3 |
Half a day has passed since submission. All three DRep votes are yes votes recorded around the boundary, between 21:13 and 22:30 UTC on 11 September.
No conclusion follows from numbers at this stage. As the previous issue showed, half the votes on actions of this kind land in the final epoch. With a month to go, the only readable fact is that no opposing vote has arrived yet.
What to watch is whether the SPO side stays silent, or votes anyway. SPOs can still cast votes even when those votes do not gate ratification. Whether the side that was excluded chooses to register an opinion will become visible over the next month.
6. The Committee Changed Over on the Morning of 7 September
This follows up on the previous issue’s line: “enactment is at 6:44:51 on the morning of 7 September.”
Update Constitutional Committee 2026 now has enacted_epoch 654. Epoch 654 began at 1788731091 = 21:44:51 UTC on 6 September 2026, which is 6:44:51 a.m. JST on 7 September. It enacted exactly as forecast.
Intersect’s weekly update #128 (11 September) records the same: “The new Constitutional Committee was enacted at Epoch 654.”
The current composition, from Koios’s committee_info:
| Item | Value |
|---|---|
| Members | 7 (all authorized) |
| Quorum | 2/3 |
| Term to epoch 799 | 4 |
| Term to epoch 726 | 3 |
The previous issue noted that only key hashes could be confirmed. The departing side is now named, in #128:
“A sincere thanks to KtorZ and Cardano Japan Council for their service”
The Cardano Japan Council‘s term ended with this changeover. The seat that had carried a Japanese-language presence on the committee is, for now, not there. For the four new or renewed members, Koios returns key hashes only, and we have not yet confirmed a primary source tying those to names.
7. One Vote That Had Been Cast Stopped Counting
What happens to committee votes already cast when the committee changes? This week gives a concrete case.
We counted the committee “no” votes on Governance Incentives Framework 2026 two ways.
| Method | Committee “no” |
|---|---|
Koios proposal_votes (itemized, 123 votes total) | 3 |
Koios proposal_voting_summary (aggregate) | 2 |
| AdaStat (aggregate) | 2 |
The itemized record still holds all three, each with a timestamp: 16:33:22 UTC on 21 August, 16:07:21 UTC on 28 August, and 10:31:25 UTC on 30 August. As an on-chain record, nothing has disappeared.
The vote accounting for the difference came from a member whose hot key begins cc_hot1qde96n2y…. That key is not among the seven returned by the current committee_info. It left the committee in the 7 September changeover.
The same member also cast an Abstain on Reimburse Ikigai Info Governance Action Deposit. on 30 August, and that action likewise shows 3 itemized votes against 2 in the aggregate.
So: when the committee changes, votes cast by the departing members stop counting on the ratification side. They remain as a record; they do not remain as a judgement. The two need to be read separately.
One technical note. Our canonical data path halted on this action specifically, having detected the mismatch (committee “no” at 2 on AdaStat, 3 in the Koios itemized list). That halt is by design, to prevent a figure from being misread. This issue treats the mismatch itself as the fact.
8. The New Committee Has Cast No Votes in Five Days
What did the new committee do in its first five days?
For each of the four live actions, we pulled the complete Koios vote record and counted only votes recorded at or after the start of epoch 654 (1788731091).
| Action | Votes in window | DRep | SPO | Committee |
|---|---|---|---|---|
| Governance Incentives Framework 2026 | 11 | 11 (all no) | 0 | 0 |
| Reimburse Ikigai … | 11 | 11 (10 yes, 1 no) | 0 | 0 |
| Reduce minPoolCost to 75 ada | 3 | 3 (all yes) | 0 | 0 |
| Withdraw 11,787,063 ada … (OpenZeppelin) | 10 | 10 (8 no, 2 yes) | 0 | 0 |
Across all four actions, committee votes are zero. SPO votes over the same window are zero as well.
Only DReps moved: 35 votes in five days.
The record does not say why. Post-appointment procedure (hot key authorisation is complete), or a judgement about the time remaining, are both conceivable — but neither appears in any primary source we could confirm. What can be placed as fact is that the figure is zero over five days.
That zero, however, has a deadline attached. See the next section.
9. Two Actions Expiring Soon, Two Expiring Later
The four live actions, in order of deadline (retrieved 08:2x JST on 12 September 2026).
Expiring in epoch 656 — 6:44:51 a.m. JST on 17 September
Reimburse Ikigai Info Governance Action Deposit. (treasury withdrawal, 103,000 ada requested)
| Category | Current | Threshold |
|---|---|---|
| DRep yes | 58.37% (118 votes) | 67% |
| DRep voted no | 2.51% (7 votes) | — |
| Committee | 2 of 7 in favour | 2/3 |
DReps are 8.63 points short. The committee stands at 2 of 7, where 2/3 needs close to five. Both of those yes votes came from the old committee’s side in August. The remaining five are unvoted under the new composition.
Governance Incentives Framework 2026 (treasury withdrawal, 4,207,967 ada requested)
| Category | Current | Threshold |
|---|---|---|
| DRep yes | 1.62% (11 votes) | 67% |
| DRep voted no | 91 votes | — |
| Committee | 2 of 7 against | 2/3 |
Here the direction is settled: 11 yes votes against 91 recorded no votes.
These figures match Intersect’s weekly update #128 (as of 18:30 UTC on 11 September: GIF 1.62%, Ikigai 58.26%, committee 2-0-0-5 on both). The spread on Ikigai — 58.26 / 58.30 / 58.37 — is a matter of retrieval time.
Expiring in epoch 661 — 6:44:51 a.m. JST on 12 October
Reduce minPoolCost to 75 ada — see §5.
Withdraw 11,787,063 ada for the OpenZeppelin Stack administered by Intersect (treasury withdrawal)
| Category | Current | Threshold |
|---|---|---|
| DRep yes | 0.02% (2 votes) | 67% |
| DRep voted no | 2.06% (8 votes) | — |
| Committee | 7 unvoted | 2/3 |
Submitted at 1:27:04 a.m. JST on 9 September. It requests funding to bring an audited smart contract library to Cardano, for 11,787,063 ada. By vote count it currently stands at 2 in favour against 8 opposed.
10. The Constitutional Amendment Did Not Arrive Before Epoch 655
In its 10 July 2026 article “Preparing the Constitution for Dijkstra”, cardano.org wrote that the constitutional amendment action had its “submission targeted before epoch 655”.
Epoch 655 began at 6:44:51 a.m. JST this morning.
We scanned all 157 entries in Koios’s proposal_list by type. The most recent NewConstitution is CARDANO BLOCKCHAIN ECOSYSTEM CONSTITUTION v2.4, submitted in epoch 601 and enacted in 609. Only two actions were submitted in epochs 654 and 655, and neither is a constitutional amendment (the new minPoolCost, and OpenZeppelin).
By the stated target, it has not been submitted.
No primary source we could confirm explains the delay. We set out the target and the current state, and leave it there. The new committee that took its seat this week is the body that will assess constitutionality when an amendment does arrive.
11. What Hydra’s Critical Was Skipping
The previous issue covered GHSA-cg83-6w6r-6hx3. This week we cover the substance. The release itself is dated 2 September and is not a new event this week, but the content of the advisory circulated across the week boundary.
The cause was a fast path that skipped verification.
When processing a snapshot request, hydra-node had a path that assumed a transaction had already been validated at an earlier stage, and therefore skipped signature verification and Plutus script evaluation. Unvalidated transactions could reach that path.
The consequence is that a malicious head participant could place an L2 transaction with an invalid signature, or with a script that should have failed, into a confirmed snapshot — and from there, take funds.
The fix in 2.4.1 removes that assumption:
“transactions are now always fully validated, including signature and script checks, when processing a snapshot request.”
The upgrade path is as reported last week. From 2.4.0 it is a drop-in replacement (no script or database changes). From 2.3.0 or earlier, heads must be closed first.
This class of vulnerability has a property worth remembering. An optimisation made for speed had removed one precondition of correctness. If the precondition always holds, nothing goes wrong; if there is even one entrance where it does not, that entrance is the hole.
12. The Node and the Tools
cardano-node 11.1.1 Is Not for Production
Published at 19:13:04 UTC on 8 September. This is the patch for the resident memory increase in v11.1.0 mentioned last week.
But the first line of the release notes says:
“Node version
11.1.1is a pre-release version ofcardano-nodethat is not yet recommended for use on mainnet.”
GitHub also flags it as a prerelease. Please do not read this as “the memory issue is fixed, so upgrade.”
What it contains:
- Complete removal of the legacy tracing system (iohk-monitoring-framework). Configurations still using the old keys must be migrated.
- Removal of the V1 LedgerDB and the LMDB storage backend. LMDB users must switch backends.
- Ledger snapshots made predictable and Mithril-compatible.
- Initial support for Peras (off by default, behind the experimental
NodeToNodeV_16). - Continued groundwork on the Dijkstra era.
A known issue remains: memory use while syncing is higher than 11.0.1, though the notes state it is significantly less than the 11.1.0 regression.
Intersect’s #128 says 11.2 is expected within a month, carrying the full Dijkstra feature set apart from Leios. For operational purposes, the natural reading is to treat 11.1.1 as a holding release and 11.2 as the target for production.
Lace 2.3.0 — The DRep Ordering Changed
Published at 15:19:47 UTC on 10 September. One change here bears directly on governance:
“Better search in the DRep browser, and a default ordering based on the quality and completeness of each DRep’s profile rather than raw voting power“
The screen where a delegator chooses a DRep no longer sorts by size first. A structure in which DReps that have already accumulated delegation stay at the top reproduces that concentration. Changing the ordering decides nothing by itself, but it changes which way the default option points.
Also in this release: delegating to a script-based (multisig) DRep now works correctly, support for the Trezor Safe 7, and a “Reset sync state” option in Midnight account settings.
Leios — The Testnet Was Rebuilt
ouroboros-leios Prototype 2026w36 was published at 11:29:24 UTC on 7 September.
The largest change is an implementation that selects the Leios committee at epoch boundaries. Voting keys now expire and require rotation, like KES keys. A key stays valid for ceil(maxKESEvolutions × slotsPerKESPeriod / epochLength) + 2 epochs after registration.
Committee size, quorum thresholds, endorser block limits and certification gaps are now configurable through governance parameters — adjustable without shipping a new node.
And this is what lands directly on operators:
- The domain separation tag used for the BLS proof-of-possession was corrected to follow the draft IETF standard. Re-registration with the new
cardano-cliis required. - From the release notes: “Requires respin: Delete your local state and re-sync from genesis or sideload from the IOG relays, using the
musashinetwork config… The respin is expected to happen in 1-2 days.”
For testnet participants this means rebuilding the database, updating node and configuration, and re-registering the pool.
13. Intersect — Eleven Candidates for Two Seats
Applications for the board election have closed. Eleven candidates are standing for two member-elected seats.
| Item | Schedule (UTC) |
|---|---|
| Voting opens | 12:00 on 14 September |
| Voting closes | 12:00 on 25 September |
| Results | 28 September |
Three conditions apply to voters: at least 180 days of membership as of the moment voting opens, membership in good standing, and completed identity verification with Intersect. Joining now will not qualify anyone for this round.
The candidates are Adam Rusch, Alexa K., Anuj Chaudhary, Christian Taylor, Eric Helms, Ian McCullough, Jose Miguel De Gamboa, Kavinda Kariyapperuma, Ken Erik Ølmheim, Nana Safo and Ridwan Itopa Musa.
The same update carries two more operational items.
- Maintenance of GovTool has moved to Sireto, led by Sandip Pandey, one of GovTool’s original builders. The update states explicitly that “there are no practical changes for users at this stage.”
- On the 2026 budget: 16 of 25 contracts are funded, with 9 remaining. A total of 74m ada and 22m USDCx has been transferred to vendors through the Treasury Reserve Smart Contract framework.
CAP Review sessions begin on 17 September, alternating between 07:00 and 14:00 UTC.
14. Midnight Watch
Shielded Assets and Phishing
One post went up on the official blog this week: “How shielded assets thwart phishing attacks” (10 September, Sebastien Guillemot).
The argument: public chains hand attackers the data they need to choose whom to target. Shielded assets make that selection difficult. It frames the starting point of phishing as the selection step, not the sending step.
MIP-0017’s Document Arrived Before Its Number
A proposal to use a verifiable random function for shielded spend authorisation landed as MIP-0017 shielded-spend-auth. PR #289 was merged at 01:16:59 UTC on 11 September. Immediately before it, a commit renamed mip-xxxx.md to mip-0017-shielded-spend-auth-vrf.md.
However, NUMBERS_INDEX.md, the canonical register of assigned numbers, still lists only up to MIP-0016 as of today’s retrieval. The document is in the repository; the index has not caught up. On the MPS side the index runs to 0039 (lightweight-contract-interaction).
As of the previous issue, there were 16 MIPs and 38 MPSs. Counted by the index, MPS gained one this week; counted by documents, MIP gained one as well.
The DevRel Team Was Dissolved
The Midnight Foundation announced on X (11 September) that it has dissolved its dedicated DevRel (developer relations) team. There is no post about this on the official blog.
What replaces it — documentation, grants, hackathons, ambassadors — is not stated in any primary source we could confirm, so we leave it unwritten.
In the Japanese-speaking region, an ambassador tour was announced in the same week: Fukuoka on 11 September, Nagoya on the 12th, Tokyo on the 13th.
15. Risk Dimensions
- Governance (structure). What this week’s resubmission demonstrates is that the same substantive change can require a different set of voters depending on how it is bundled. This is behaviour as specified, not a loophole. But the fact that a resubmission can be shaped so the SPO vote falls outside the ratification requirements changes how actions should be read. Read the
protocol_param_update, not the title. - Governance (time). Two actions expire in five days, and the committee has cast no votes in five days.
Reimburse Ikigaisits at DRep 58.37% with 2 of 7 on the committee. If nothing moves before the deadline, it expires even where DRep support is short by only a little. - Committee (continuity). When members change, votes already cast stop counting on the judgement side. It was one vote this time; if the same thing happens on an action close to its deadline, both the days remaining and the votes required change at once.
- Hydra (operations). If you are running 2.3.0 or 2.4.0, upgrading to 2.4.1 remains outstanding from last week. From 2.3.0 or earlier, heads must be closed first.
- Node (operations). 11.1.1 is a pre-release and not for mainnet. Because legacy tracing and LMDB have been removed, migrating when 11.2 arrives will involve configuration changes. That is work that can be prepared now.
- Leios (participation). A respin including BLS key re-registration is required. For testnet participants this is hands-on work for this week.
- Information (Midnight). The DevRel restructuring was announced only on X, with nothing on the blog. While the party running the front door is changing, the sources of information scatter for a time.
16. What to Watch Next Week
- The two epoch 656 actions — deadline 6:44:51 a.m. JST on 17 September. Whether committee votes arrive on
Reimburse Ikigai, orGovernance Incentives Framework 2026simply expires. - The new committee’s first vote — whether five days at zero breaks in front of an action with a deadline.
- DRep votes on
Reduce minPoolCost to 75 ada— when opposition arrives, and how much. Last time it climbed from 61.73% to 68.57%. - Whether SPOs vote anyway — whether the side now outside the requirement registers an opinion. This week it was zero.
- The constitutional amendment — the epoch 655 target has passed. The new committee is the body that will assess constitutionality.
- cardano-node 11.2 — Intersect expects it within a month, with the full Dijkstra feature set.
- Turnout in the Intersect board election — voting opens 12:00 UTC on 14 September. Eleven candidates, two seats, under a 180-day membership and identity-verification requirement.
- Uptake after the Leios respin — whether participating pools complete BLS key re-registration.
- MIP-0017 in the index — whether the document-only state is resolved.
- What follows Midnight’s DevRel — what takes over the front door. The Japan tour of 11–13 September is the first place an answer may appear.
Summary
The previous issue closed like this: “If it is resubmitted, the question will be how to assemble 51% on the SPO side.”
The answer came in the form of not assembling it.
The action resubmitted at 5:27 this morning did not change a single thing about its substance: reduce minPoolCost to 75 ada. What changed is the bundling. The Plutus memory limits that had been in the same bundle came out. And maxBlockExecutionUnits is a security-relevant parameter under CIP-1694, while minPoolCost is not. So the SPO vote falls outside the ratification requirements.
This is behaviour as specified. It is not a loophole. But the fact remains: the side that stopped this at 34.5% last week is not in a position to stop it this week.
What follows from this is a lesson about reading actions. Read the protocol_param_update, not the title. What is bundled together determines who decides. This week made that concrete.
Something of the same shape happened on the committee. Four seats changed over on 7 September, and one vote cast by a departing member stopped counting on the judgement side. It remains in the record. It does not remain in the judgement. And the new committee has cast no votes in five days — with two actions expiring in five days.
The people who decide changed; the requirements for deciding changed. Both landed in the same week. Both happen before any vote is counted.
And 170 ada is still 170 ada.
References
- Koios REST API —
tip/epoch_info/epoch_params/totals/proposal_list/proposal_votes/proposal_voting_summary/committee_info(epoch boundary times; action submission, state and parameter contents; vote timestamps; current parameters; committee composition): https://api.koios.rest/api/v1 - AdaStat (stake-weighted yes percentages, ratification denominators, SPO thresholds): https://adastat.net
- CIP-1694 — Cardano governance (the security-relevant protocol parameter list and the additional SPO vote requirement): https://cips.cardano.org/cip/CIP-1694
- Cardano governance actions: https://gov.tools/governance_actions
- Intersect MBO — Intersect weekly update #128 (2026-09-11: board election candidates and schedule, GovTool maintenance transfer, budget execution, committee enactment and departures, live action figures): https://www.intersectmbo.org/news/intersect-weekly-update-128-sep-11-2026
- Intersect MBO — Intersect weekly update #127 (2026-09-05: the committee update passing, minPoolCost expiring, MusashiNet): https://www.intersectmbo.org/news/intersect-weekly-update-127-sep-5-2026
- Cardano — Preparing the Constitution for Dijkstra (2026-07-10: “submission targeted before epoch 655”): https://cardano.org/news/2026-07-10-preparing-the-constitution-for-dijkstra/
- Hydra — security advisory
GHSA-cg83-6w6r-6hx3(severity critical): https://github.com/input-output-hk/hydra/security/advisories/GHSA-cg83-6w6r-6hx3 - Hydra — release 2.4.1 (2026-09-02: cause, fix and upgrade procedure): https://github.com/input-output-hk/hydra/releases/tag/2.4.1
- cardano-node 11.1.1 (2026-09-08: pre-release status, legacy tracing and LMDB removal, initial Peras support, known memory increase): https://github.com/IntersectMBO/cardano-node/releases/tag/11.1.1
- Lace extension 2.3.0 (2026-09-10: DRep browser default ordering, script-based DRep delegation fix, Trezor Safe 7, Midnight sync reset): https://github.com/input-output-hk/lace/releases/tag/lace-extension%402.3.0
- Ouroboros Leios — Prototype 2026w36 (2026-09-07: committee selection at epoch boundaries, voting key rotation, BLS domain separation tag fix, respin procedure): https://github.com/input-output-hk/ouroboros-leios/releases/tag/prototype-2026w36
- Essential Cardano — Weekly development report as of 2026-09-11 (Leios prototype improvements, 2.4.1 superseding 2.4.0, Plu-stan’s first stable release): https://www.essentialcardano.io/development-update/weekly-development-report-as-of-2026-09-11
- Midnight — How shielded assets thwart phishing attacks (2026-09-10, Sebastien Guillemot): https://midnight.network/blog
- Midnight Improvement Proposals repository (MIP-0017 PR #289 merge,
NUMBERS_INDEX.mdstate, MPS-0039): https://github.com/midnightntwrk/midnight-improvement-proposals - SITION
market_indicators(fixed-time morning readings for ADA / BTC / ETH / NIGHT on 5 and 12 September, CoinGecko series): https://www.coingecko.com
Transparency Notes
Where judgement was required this week, and what was left out.
- Grounds for “the SPO vote is no longer required”: we confirmed from the primary source that CIP-1694’s security-relevant list excludes
minPoolCostand includesmaxBlockExecutionUnits, and we compared theprotocol_param_updatecontents of the old and new actions against Koios measurements. We further confirmed that our canonical data path returns an SPO threshold of 51% for the old action and 0% for the new one. Three lines of evidence point the same way, which is why this is stated without hedging. - The proposer’s intent: we do not write that the bundle was removed in order to avoid the SPO vote. No primary source states a reason. What is written stops at the consequence.
- The committee vote that “disappeared”: we confirmed by measurement the mismatch between the Koios itemized record (3) and the Koios aggregate and AdaStat (2 each), and confirmed that the key casting the difference is absent from the current
committee_info. Our data path halted fail-closed on this action, by design. We did not bypass the halt to take a figure; we describe the mismatch instead. - Names of the four new or renewed members: Koios returns key hashes only, and again this week we could not confirm a primary source tying them to names. Only the two departing parties (KtorZ, Cardano Japan Council) are named, on the authority of Intersect’s weekly update.
- Why the committee did not vote for five days: not readable from the voting record. We place the fact only.
- MusashiNet Earth phase figures: day counts, block counts, participating pool counts and throughput multiples relative to Praos circulate as secondary summaries, but no such passage appears in the primary sources we could retrieve (this week’s development reports and the Leios release notes). We therefore do not give those numbers. What we do give is the respin and BLS key procedure quoted verbatim from the release notes.
- Midnight’s DevRel dissolution: the primary source is a Midnight Foundation post on X. There is nothing on the official blog. No primary source describes the replacement, so we do not describe one.
- The number of MIPs: MIP-0017 exists as a document, but
NUMBERS_INDEX.md, the canonical register, runs only to MIP-0016. Writing only one of the two would make the count inconsistent, so we write both. - Why the constitutional amendment did not arrive: no primary source explains it. We write only that it has not been submitted by epoch 655. We do not write that it was cancelled.
- Aligning the price sources: this issue’s market table takes both endpoints — 5 and 12 September — from the same canonical source (SITION’s
market_indicatorsfixed-time morning capture). As a result the 5 September values differ in the decimals from the previous issue’s table, which used a one-off capture at 07:07 JST that day (ADA $0.210854 here versus $0.211514 there). Direction and magnitude are the same. - The ADA figure at the epoch boundaries: the −6.94% in the body comes from CoinGecko values taken exactly at the epoch boundary times. Because the capture times differ from the morning fixed-time readings, it does not match the −2.46% for the week. It is set out as a separate measure.
- How cardano-node 11.1.1 is characterised: because the release notes state it is a pre-release, we do not summarise it as “the memory problem is fixed.”
- Prices and governance actions: not connected to each other.
This is not investment advice. For information and research purposes only.