Release v0.20.0 - #27
Merged
Merged
Conversation
…ensus/gates (activation registry row 21i, the indexer twin's bytes)
…istry rows (activation registry row 28-code)
…retired, 7 logic twins and 3 carriers moved under src/consensus from the indexer's bytes, the hub-owned zero-conf gate under src/attestation (activation registry row 21h) Callers of the deleted shims read their rows through ./consensus/gate_registry by literal key and the bridge engine spells its three gate keys literally (a miss still throws at construction naming the key). Every twin is byte-identical to the indexer's copy at the same tail (D101); no registry key, row or value changes and the hub identity holds (digest 26ba9cce, gates_field_hash f19f21f0, 33 keys). The sibling reference map scanner follows the digest loader's bound-variable shape, the reachability dynamic edges name the W5 gate paths, and the identity tool's exit code on an uncomputable digest is pinned by a test. Four test files whose only assertions were a deleted shim's table are deleted. Tier 6493 passing; the pin test and the frozen-set manifest red until row 28-pin re-pins.
… consumers (a W5 comment reword in the canonical copy reverted; activation registry row 21h)
… 3 ids; frozen set, identity digest, suite titles and sibling map re-pinned; the pin membership rule follows moved paths in all three twins (activation registry row 28-pin, D102)
…ion-length limit (activation registry row 23)
…n registry The governance snapshot-lock table was the one flag-day literal outside the hub's registry. It becomes a row of a new hub-only part, consensus/gate_registry/hub_rows.js (no SHARED-GATES markers, twinned nowhere, queued after the block so the block's order is untouched), keyed validators/governance/rules.GOV_SNAPSHOT_ACTIVATION with its comment. rules.js reads the frozen table back by key, so electorate.js indexes it by network exactly as before and a build without the row throws at load naming the key. Values unchanged: mainnet 963000, testnet 0, regtest 0. The layout suite grades the hub-only part apart from the block parts and counts its keys.
The 2026-09-16 fleet roll wedged all three testnet indexers. Recreating the five validator hubs made every mirror re-drain each table from id 0 over /hub-db/snapshot, three mirrors behind ONE public address spent the shared 100 req/min budget that ordinary fleet polling was already drawing on, and the drain 429ed part way, retried every 30s, re-spent the budget and never finished until HUB_RATE_LIMIT_RPM was raised to 60000 by hand on all five hubs. Nothing in the train carried that, so a recreate from clean compose or a sixth validator reproduced the wedge exactly. The snapshot routes now carry their own per-IP budget, HUB_SNAPSHOT_RATE_LIMIT_RPM, default 600: about 32 page reads for one mirror's full ten-table bootstrap, about 96 for three mirrors behind one address, about 192 req/min while any of them is still retrying. The general limiter skips that family and stays at 100 req/min, so the public JSON-RPC surface keeps exactly the throttle it had; raising one shared limit to fit the drain would have taken that throttle off every open read instead. Driven at the shipped defaults through the real snapshot routes: 192 authenticated page reads from one public IP all answer 200, the JSON-RPC surface still 429s at call 101 with the -32029 envelope, the snapshot family still 429s at request 601 in the REST shape its routes already use, and an unauthenticated page read is still 401.
…and name each marker's age hydrateMarkers loaded every intent row this hub had ever written and reported only a count, so a marker from four days ago read exactly like one from the last hour. pendingWindows only ever asks about the MAX_CATCHUP_WINDOWS windows below the current one, so a marker below that floor names a window no sweep can reach: it is quarantined no longer, reported at info as history rather than as an operator action item, and the action-item line now carries each window's age in windows closed and in wall clock.
…rier family Carries the hub's registry rows for the mirror admission producer/consumer heights and the anchor-attest barrier, rolls HUB_SCHEMA_VERSION to v7 for the admission_height column, and re-pins the gate-registry and consensus-identity guards the new rows move. Platform buildout row 15 (child rows 7, 8, 9b, 10, 11 and three seam-fix passes).
…ge engine api schema version 7)
…gistry P11)
A changed-hash move requires a reason and writes one atomic {from, to, path, date, reason} record. Verified 17/0 byte-identical across indexer, sync and hub.
…ghts and re-pin the hub identity
…espace arming, carrier re-pin
… the BTC-arms-last dependency
… block The hub pinned snapshot_block on its pushed BTC tip, so an escrow proof could read the checkpoint one block before the lock and refuse the mint every block. Proposers now hold until the tip covers the leg; followers refuse a lower snapshot.
…nchor rewards carry the admission-height columns the live stream carries
…ghts from the v0.20.0 train (dq3 (a)) This reverts commit f39437f; the token bridge and policy gates stay dark on v0.20.0 and arm on the next train. The identity pin stays at HEAD's value, because every row f39437f re-pinned (digest, hub_schema_version, the barrier and mirror admission rows) belongs to later barrier work and the hub identity reads no token or policy row.
…ying and the BTC-arms-last dependency" This reverts commit b06036b.
This was referenced Sep 17, 2026
This was referenced Sep 17, 2026
… the BTC instant The v0.20.0 family was sized 2026-09-16 20:41Z from last-99-block cadences. LTC testnet then ran at about 82 s per block against the 146.6 s that sizing assumed, pulling its producer boundary to 3.2 h out while BTC's stayed 53.0 h out, and DOGE's drifted 7.5 h early. Two legs of one cross-chain match would have crossed the flag day about two days apart, which is what the same-wall-clock-instant rule exists to prevent. Re-measured 2026-09-17 22:45Z over a trailing window as long as the lead being sized: TBTC 152,891 at 576.7 s per block, TLTC 4,889,190 at 82.5 s, TDOGE 67,904,912 at 27.7 s. BTC is unchanged, still keyed to epoch close 153,216 plus 6 buried; LTC and DOGE are converted onto that same instant, and each consumer is its own producer plus six hours at its own measured cadence rather than a block count carried over from the first sizing. Arms the TRAIN_ACTIVATION 0.20.0 row at testnet 153,116, 106 blocks and about 17 h below the BTC producer. resolveRuleSet reads only the local map, so with no row every block above the boundary keeps resolving under 0.19.0, and a manifest naming 0.20.0 halts a fleet in which no build implements it. Records the cadence-window rule the re-cut used and makes the re-size rule per chain instead of BTC-keyed, so an LTC or DOGE drift past the six-hour ordering margin forces a re-cut the way an overrun BTC height already does. (cherry picked from commit 82ecaf4382655af714ed27b92573b5b9af18c160)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Title: Release v0.20.0
Summary
Changed
carrier_logic_pin_ops --movecan re-pin changed carrier logic and record path and hash transitions atomically.Fixed
Sibling release PRs
Merge order
Merge order: merge after xchain-documentation and before xchain-vm.
CI note
GitHub CI on this branch clones sibling repos at develop, not at this release branch, unless a workflow call passes siblings-ref for release branches. A cross-repo guard failure here can reflect that develop-only comparison rather than a real disagreement with a sibling still on this release branch; check the venue pre-push gate (which ships each companion at its own release branch) before treating a red guard as a defect.