20215f0274 · Merge scrub-3 ff0d0dde into master (gate: green on ff0d0dde, recorded by tools/ci/pre-push.sh; landed on the box mirror under the exception declared by main: main's ruling, 7 Oct 2026 19:5x UK: the GitHub account is suspended, lanes land on the box mirror's master, the box gate stamp is the verdict; GitHub gets the fast-forward when it answers) · Updated 2026-10-07 22:46:40 +02:00
83b8c8bd74 · publish-manifest.sh: a carried app entry keeps its own ui_version or none; only a new entry is stamped from the tree (a carried 0.3.21 entry had taken the tree's 1.0.2 and passed the pair check falsely) · Updated 2026-10-07 20:42:05 +02:00igneum-labs
4bc5974a92 · Intel Arc plan: the 9034 retry on PC 2 cancelled for tonight (the project lead's rule: no PC job needs a click or a UAC prompt); 6733 stays and is acceptable for the B580 until an unattended install path exists · Updated 2026-10-07 20:41:52 +02:00igneum-labs
0974737822 · The opening window size, one rule for both hosts (window-22, 7 October 2026, the project lead on PC 2: "why did the app open super wide?"): scaling-21's 80 percent of the work area scaled above 1920 wide opened 2752 by 1152 on a 3440-wide area and 3072 by 864 on a two-monitor span, and a remembered WindowW/WindowH could carry that into later starts; now app/windows/opening_size.h: the PRIMARY monitor's work area (MonitorFromPoint primary, never the virtual desktop), width 80 percent capped at 1440 up to a 1920-wide display and 1920 beyond, height from the width at 16:10 bounded by the work area, never wider than 1.6 times the height, never under 900 by 600, never over the work area; a remembered frame kept only when it fits the work area, the minimum and the aspect cap, and never saved when it would not; the Mac window mirrors the rule on NSScreen.screens.first and discards an autosave frame that fails it; opening_size_test.c compiles on the gate with cc (known-failed first: it could not compile before the header, and 0.3.21's rule fails the 3440 by 1440 and the 3840 by 1080 cases); host.cpp touched (the shipper's named line) · Updated 2026-10-07 20:34:47 +02:00igneum-labs
bf6dbdf355 · Merge pool-finish-21 4d7521b3 into release-0.3.22 (the pool daemon keeps jobs 60 s or 256 per member, stale grace 30 s: the Devnet 3 read of 25,477 unknown_job shares at one template a second) · Updated 2026-10-07 20:34:37 +02:00igneum-labs
4d7521b3c6 · Pool daemon: jobs kept 60 s or 256 per member (never fewer than 12), the open pool's stale grace 30 s unless set (the Devnet 3 pair, 7 October 2026, 18:16Z: a template every second and a GPU batch outliving the 12-job window, 25,477 unknown_job and 19,119 stale shares, 0 accepted) · Updated 2026-10-07 20:19:52 +02:00igneum-labs
5d021c5e34 · MF-14: a worker never waits on the program export longer than one retry interval without the reason shown · Updated 2026-10-07 20:18:43 +02:00igneum-labs
769ac58b79 · The shard sentence from the block's facts (7 October 2026, the site's animation said "no shard plan yet" on every planless block): IgneumDag.shardWords(block, nowMs) in scene/live-dag.js (2.0.6) is the one source for the tooltip, the /live inspector and the app's inspector: no shards reads shard plan loading (a chain block under 10 s), excluded, nothing to prove, not yet ordered (pending), off the selected chain, no shards planned, or no shard plan after N s; a planned shard with no prover reads awaiting a prover with the block's age, assigned to <prover> when one holds it, proving, verified, paid as before; the app's inspector prints these per shard and as the count line; the copies synced; tools/scene/shard-words.test.mjs on the gate's scene line, known-failed first on 2.0.5 · Updated 2026-10-07 20:13:23 +02:00igneum-labs
4f24f81535 · The miner's chain key from the shared legend export (key-22, 7 October 2026): IgneumDag.legend({mine}) and renderLegend(el, {mine}) in scene/live-dag.js (2.0.5) carry the one list of entries with each state's token and shape (Pending, Included, Excluded, Selected chain, Proven, Locked checkpoint; the app adds Your blocks, ringed, first); the app's chain card renders its key from it at mount with the source line kept (the Included swatch had been ember while the scene draws included blocks in --included, and "pending" was lower-case), its CSS colours a swatch only from the entry's token; the copies synced; tools/scene/legend.test.mjs on the gate's scene line: the export's entries, order, tokens and shapes, every colour the scene draws has an entry, the app key is the export, and the /live key equals the export minus the ruled exceptions (no Pending, no Your blocks) with the same wording, token and shape on every shared entry, binding on master where the site deploys from and a diagnostic on a release branch whose site tree is the frozen cut; the site is not edited (main's correction: the project lead's ask was the miner's key) · Updated 2026-10-07 20:02:56 +02:00igneum-labs
9e782ee9a3 · Signing state on every card row and the node line (signing-22, the project lead's signing-bonus ruling, 7 October 2026): the miner signs the finality section by default (Settings.vote true, an old settings file without the field reads true, the one --no-vote flag comes from config::vote_flag and exists only when the user turned voting off; the Rust test vote_defaults_on pins all three); View.signingWords says "signing" with the checkpoint signed and when, or "silent" with the reason (voting off in Settings, no vote key, the clock behind or ahead of the network, no template while the node syncs, or a key that stopped for N min while the chain kept locking), "first checkpoint pending" in the first two minutes, "paused" when the network itself has not locked; the Overview's node line ends in the word and turns bad with the reason as its sub line; each Cards row shows its own keys' state under the state line; two view tests known-failed first · Updated 2026-10-07 19:52:08 +02:00igneum-labs
46140ce374 · Driver check plan: PC 2's crash named from its event log (bugcheck 0x3B during the silent Intel install over a mining eGPU; the app off until the next logon) · Updated 2026-10-07 19:49:44 +02:00igneum-labs
11aeaf1d68 · Merge driver-check f03b0e74 into release-0.3.22 (main: a 0.3.22 app item; the eGPU driver-install hold and warning) · Updated 2026-10-07 19:32:40 +02:00igneum-labs
826dbf7629 · 0.3.21 over the field node: the Windows node-source pin back to c4459193 (main's shape: the 0.3.21 app over c4459193, no node gate; the 96161037 line folds into 0.3.22) · Updated 2026-10-07 19:04:28 +02:00igneum-labs
826dbf7629 · 0.3.21 over the field node: the Windows node-source pin back to c4459193 (main's shape: the 0.3.21 app over c4459193, no node gate; the 96161037 line folds into 0.3.22) · Updated 2026-10-07 19:04:28 +02:00igneum-labs
cb61192960 · miner-faults: MF-11 in the update-return lane's words (the helper owns the return, the relay service, the ping, the tuner's ceiling) and MF-13 (the Power Helper's stale-command count); the external card kind noted · Updated 2026-10-07 17:09:50 +02:00igneum-labs
1ea13655d7 · MF-11 follow-ups on release-0.3.21 (the app half of update-return 2d3d211a and 0b423697): the external card kind from a USB4 or Thunderbolt router in the device's parent chain (PC 2's RTX 5060 Ti in a Razer Core X V2 reads eGPU, not discrete); the Power Helper's registered probe wants the task enabled and its action exe on disk; a helper that gives no heartbeat or no line inside its window is a FAULT power-helper line with the registration re-read; a same-length rewrite of the command file is a new command (effective_skip; PC 2, 7 October 2026: every tune refused since the 14:35Z boot) · Updated 2026-10-07 17:07:17 +02:00igneum-labs
e68ba08ccc · MF-11 update-return (0.3.21), the app half on release-0.3.21: the Windows update helper owns the return (the exe set kept beside the app, the app launched by the helper, api/state polled 120 s, the kept set restored when nothing answers, one intake line either way; the installer stays silent under /IGNOTA=2), the window host restarts a dead engine and answers the Restart Manager, the first act after an update is the read-back line and a FAULT pc-restart line when the PC came up from a power loss; the tuner never asks above a card's measured efficient point (the 5090 at 308 W, floored at the vendor's 400 W) unless Power control is on and the user raised the cap, and a refused cap is asked again at 2 and 10 min then reported as FAULT power-cap; the job-channel ping on every wake request (7 October 2026, PC 2 lost power at 10:46Z; the relay half stays on update-return) · Updated 2026-10-07 16:29:41 +02:00igneum-labs
d817bc1ecc · Pool daemon: --template-parallel (default 2) bounds the template fetches in flight against the node (the ten-member window on pool-1, 7 October 2026: ten parallel getBlockTemplate calls on one gRPC connection queued past the client's request timeout from 12:54Z, no job for 36 minutes); pool.md 10.5a records the window's two stacked faults · Updated 2026-10-07 15:40:58 +02:00igneum-labs
a071b7ff9a · Miner fault register: MF-11, a pool member without a node stalls at an epoch boundary (the ten-member window of 7 October, the prepare never answered; the pool lane's row, rule and gate owed) · Updated 2026-10-07 15:19:57 +02:00igneum-labs
edac467dbd · Driver table: a referer per row, sent as curl -e (drivers.amd.com answers 403 without an amd.com one; the AMD row carries it); a plain user agent on the download · Updated 2026-10-07 14:17:14 +02:00igneum-labs
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
