The Intel row's min_version is 32.0.101.6733 (Windows' inbox driver, on which the worker mines at 11.0 MH/s with the Intel rotate rewrite); 9034 stays the version on offer for a card with no driver. Test known-failed first (the shipped table demanded 9034 of an Arc on 6733; red on build-2, then green). The mock's offer scenario shows an older 6600 so the Install row still renders. Plan 3c: the install path moves onto the app's Power Helper task next; the minidump waits for the same path.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
drivertable::install_hold_keys holds every enabled card of the vendor being installed (the other vendors' cards keep mining; a card already off has nothing to stop); the row says so before the click and while it runs, with the enclosure warning kept on an external row; the toast and the engine's event name the vendor. The x1 gen1 link signature is not on the card state and was not added. Test known-failed first (the 5090 inside the case was not held for an NVIDIA install; red on build-2, then green); box gate 259 + 33 + 8; UI 51.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
PC 2, 7 October 2026 18:34 BST: two rows sat on "exporting this hour's program" for 18 minutes while a third card
mined. The export ran on a thread per card under one lock with no bound on the wait, and a result that never came
left the slot in the building state for ever.
- watchdog::export_wait: inside one interval (the ladder rung, at least 30 s) the row keeps its word; past it the row
names what blocks the export; past two intervals the wait ends and the worker retries on its own interval.
- engine: EXPORT_HOLDER names the card whose export holds the pack lock and for how long, EXPORT_LAST_ERROR the last
failure (the node not at the epoch, no seeds.txt); export_blocker() reads them for the row; build_seq ignores a
late result after the wait ended; a failed export retries on the ladder, never a flat five minutes.
- docs/plans/miner-faults.md: MF-14 with rule, test and gate line.
Test known-failed first: an export that never returns is named at one interval and given up at two.
Gates: app tests 260 + 33 + 8 green on igneum-build-2; the tree gate GREEN, 56 checks.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The rule (drivertable::install_hold_keys): a driver install on a card the app reads as external (the eGPU kind from update-return-21b) stops that card's worker before the installer starts, through the cards path with the restore choice kept (the --cards-off shape); the vendor's cards inside the case and every other vendor's card keep mining. The row says so before the click and while it runs (the display reset can take the machine for a minute and may need a restart; save your work first). When the installer ends the held card goes back as it was; a card the install took away comes back when the card does (driver_release_held on the detection that lists it again). The app never restarts the machine.
Tests: the known-failed rule test first (an install on an eGPU card with the worker still running held nothing: red on build-2, then green), the view test for the two sentences; box gate 258 + 33 + 8 (test --release on build-2). Mock scenarios drivers-egpu and drivers-egpu-running; captures 13 to 16 (light and dark). Branch rebased onto release-0.3.21, which already carries the driver-check commits; the earlier tip is kept as driver-check-0320.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
the project lead's /live screenshot: a fixed tall box with the lanes in its top third and dead space under them. The renderer now knows
the height it wants: top padding + axis padding + lanes shown (at most maxLanes 7, narrowLanes 4 on a phone, never under 2)
times laneHeight (46 px, 40 on a phone, 26 compact; the mount option laneHeight overrides). It reports it through
onSize(heightPx, {lanes, laneHeight, narrow, compact}) whenever it changes and through getWantedHeight(); stats() carries
laneHeight, padT, padB, capacity, wantedHeight and autoHeight, so a page that sizes its own box reads no constants. With
autoHeight (on by default when the host set no CSS height on the canvas, which is read before the first size(); forced either
way by the option; never in compact mode, the app's card keeps its own height) the renderer sets canvas.style.height itself
and lets every lane through instead of fitting the lanes to the box. Every current host sets a height (/live 420, the hero
500, the app's card 220 and its Inspect view 420), so the frames are unchanged: the parity test on build-2 stayed equal on
every comparison. The site lane switches /live and the home fold to the hook.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
(cherry picked from commit 1546b61fa1)
The drift the app carried against the site's home fold and /live, and what moved:
- renderer: both were 2.0.2 byte for byte; the app now takes the shared scene/live-dag.js 2.0.3 (paint on push whatever the
document's visibility says: the blank /live; the phone rule on the viewport width) and proof-core.js through tools/scene/sync.mjs,
and the gate refuses a drifted copy.
- palette: the dark tokens were equal; the light theme's --ember was #E04A14 and --ember-hi #F2541B against the brand package's
#D0420D / #E04A14. The fourteen scene tokens now sit in the scene-tokens block app.css takes from scene/tokens.css (dark,
[data-theme="light"], prefers-color-scheme light) and are defined nowhere else in the file.
- phone rule: the app passed narrow: window.innerWidth < 720 at mount time and never again; the site keyed it on the canvas
width (a 640 px hero on a laptop rendered as a phone). Both now leave it to the renderer: the viewport, live on resize.
- feed window: the app asked the engine for 120 s, the site 300 s; both 300 now, so the viewer can pan the same range.
- Inspect view: 360 px tall against /live's 420 (five lanes against seven); 420 now.
- feed shape: the engine rewrote this machine's blocks to miner: "you" (a word the observer never emits) and its node-only
fallback carried now as a float of seconds, rows with timestamp_ms / is_chain_block / vote_key_hash / timestamp_source and no
number or rx, miners as {id, vote_key_hash, blocks_10m}, no proving, a finality with checkpoints alone. live.rs now passes the
observer's rows through untouched (state.you_blocks counts them; the UI's mine function marks the lane from the card ids, which
is the overlay: own blocks glow, the lane reads YOUR KEY) and node_only_reply builds the fallback in the contract's shape
(every key of scene/feed-contract.json, null where the node cannot know, partial: true, state.source "node", ISO now). The
test the_node_only_reply_has_the_contract_shape reads the contract file itself (include_str!), so the Rust side and
tools/scene/feed-contract.mjs cannot drift.
App gate on build-2 (test --release, --priority gate): 228 + 32 + 8 passed. Parity on build-2: app Inspect = /live = home fold
at T+0, T+2, T+4 s, the overlay identical when both surfaces know the key.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A node before the exec RPC bounds fix (every 0.3.17 node) dies when a method that resolves a block number or indexes the record vector is asked while its exec follower holds no record; PC 1 crash-looped on two callers in one night (eth_getBlockByNumber from the clock sample, then igneum_getAssignedShards from the prover loop: 'panicked at igneum/exec/src/rpc.rs:808:35: range start index 1 out of range for slice of length 0'). Every caller (prover.rs's evm_rpc, update.rs's clock sample, extnode's rpc for chainfacts and the external-node probe) now goes through execrpc::call: SAFE_ON_EMPTY methods go out, GATED ones wait for igneum_getExecStatus's executedTipHash, an unclassified method is refused. The test every_caller_goes_through_the_gate scans src/ for JSON-RPC requests built elsewhere and for unclassified exec method names.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
docs/plans/miner-faults.md: MF-1 to MF-7, each with its rule, test and gate line.
- MF-1/MF-2: a worker starts and is judged only when the node is READY (synced and igneum_getExecStatus reports an
executed tip; execrpc::probe every 5 s off the engine thread); the node watchdog never counts the catch-up (settled
once read synced; 30 min cap before that; any RPC answer is a sign of life); the watchdog restarts on a ladder 10 s,
30 s, 2 min, 5 min, then every 5 min for ever (watchdog::RETRY_LADDER_S); the faulted state and the one-restart
budget are gone (tools/ci/permanent-fault-check.sh in the gate); a node-caused restart resets the ladder at sync.
- MF-3: the hot-plug pass starts a recovered or revived card's worker (unchanged rule, now in the register).
- MF-4: the status clock starts at ready (program loaded), loading bounded by 300 s; a self-test failure holds the
card 30 min with the reason on its row, released on a driver change; a crash loop climbs the ladder; the pack is
exported once a minute for every card (a refused pack forces one).
- MF-5: the app reads template_wait=, template_ms=, identities_active= from the 0.3.20 miner's STATUS; waiting on
the node is never the card's fault; the row says node slow; every node-wait label clears on the first rate.
- MF-6: a miners hold belongs to the job that took it and releases when that job is gone or at its own cap.
- MF-7: the engine owns every igneum-miner it started: an untracked one on this engine's node RPC is killed at start,
after every stop and every minute, one line and one fault report per kill; a restart kills the old process first.
- Every fault line posts one FAULT line to the log intake (label fault-<id8>, app and node version, 60/h cap).
- The signed cards job kind (per card enabled, identities, power_pct; refused for a card the machine lacks; applied
through the app's own card path, persisted, read back): packaging/ota/publish-jobs.sh add --kind cards.
- LG-4 as a job: relay/playbooks/first-share.ps1 and tools/fleet/first-share-gate.mjs (no Windows box yet).
- tools/reliability: the fault injector with one step per class (catch-up, card-appears, own-restart, zero-ladder,
no-status, node-silent, one-card-fails, orphan-miner); fake-worker.mjs lists devices and fails self-tests on command.
- master's build tooling (18a086b9) and release-0.3.20's igneum-pow taken into the worktree for the box routes.
Box: app 198 + 27 + 8 tests green on igneum-build-2; the tree gate green (33 checks).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
NVIDIA GeForce Game Ready 617.42 WHQL (6 October 2026): us.download.nvidia.com/Windows/617.42/..., 990,853,168
bytes, sha256 f115c927..., subject CN=NVIDIA Corporation (DigiCert G4). AMD Software Adrenalin 26.9.2 WHQL
(29 September 2026, the win11-b build): drivers.amd.com/drivers/whql-amd-software-adrenalin-edition-26.9.2-win11-b.exe
(an amd.com Referer required), 1,000,800,840 bytes, sha256 593c1d73..., subject CN=Advanced Micro Devices (Sectigo).
Read on the box: curl, sha256sum, the PKCS7 out of the PE security directory through openssl pkcs7 -print_certs.
The table's placeholders are gone: every row installs. The drivertable test sample, the mock and the view test name
the real NVIDIA release. detect.rs:990 carried a #[test] above the doc comment of the Intel test from the cherry-pick,
the test build's one warning ("duplicated attribute"): removed.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
(cherry picked from commit afafbcd52e)
