igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity

Node lane, plan 6.12: the 0.3.24 node line (class v5 on Devnet 3): the orders, the commits, the gates so far, what is open

Browse source 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
igneum-labs 2026-10-07 20:14:51 +00:00
parent 69c3070e22
commit 0b7a7a3066
1 changed files with 17 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
docs/plans/counter-asic-3-node.md
View file

@ -280,6 +280,23 @@ Note (the testnet lane, 18:4xZ): the five testnet-lane rows were written against
Not in 69d1b56e and owed to 0.3.23 (the testnet lane's content check on 96161037): c7ea1e21 silent_split, d840537b subsidy_per_block_activation_daa, 420f9305 vote-or-burn removal (`vote_or_burn_activation_daa` still in two files), daa61847 W5 succession. Caveat on every testnet-lane row: its gates ran with the switch on from genesis; the mid-chain crossing on Devnet 3 is gated by the node lane (Devnet 2 or the fast-time harnesses) before any height is named. Clock conversion: DAA d after the genesis lock = lock time + d seconds at 1 block/s (7,200 = 2 h; 108,000 = 30 h).
### 6.12 The 0.3.24 node line: class v5 on Devnet 3 (opened 7 October 2026, 19:3xZ)
Main's orders through the shipper and the Counter lane: class v5 is its own release, 0.3.24, cut the minute every v5 gate is green at any hour; the publish minute 23:30 BST (22:30Z); program_class_v5_activation_daa = dn3-g1's DAA at the cut + the seconds to the publish minute + 7,200, rounded up to the next 3,600 boundary (the estimate at 1 block/s from the lock: publish DAA about 19,732, floor 28,800, about 01:01Z 8 October; the commit uses the live reading); the signal window stays 86,400 (main's (c): no new field, the v5 tally rides v4's window as the harness rehearsed); byte 6 counted exactly; the testnet and mainnet objects carry v5 at 0. The line: fork branch v5-object-0323 = release-0.3.23-node 2720d8d2 + class-v5-node-wire + the pool lane's tag fix + the chain id by height:
| Commit | What |
|---|---|
| 7737ebd9 (class-v5-node-wire) | the exec snapshot carries the two newest epoch state streams (IGSD1), last field with a legacy fallback, version 1 kept; the loader checks each against its own record and publishes the final one |
| 6d827c5d | the day-state witness beside the pruning proof's seed headers (EpochSeedHeader.stateRoot, proto field 3); a verifier without the state accepts a class v5 epoch's headers on it and holds every later stream to it |
| 40c03806 | the stateless and stale-chip rule as a kaspa-pow unit test (a block hashed on a stale dataset passes on the stale side only) |
| b1680b57 | class_signal: an epoch under the signal window takes its floor's class, never a bare v3 (the v5-fasttime lane's first exception: e0 and e1 read v3 with the v5 object set and v4 at floor 0) |
| ee26af39, 29aecfab | the IBD catch-up across the flip: a header chunk refused for want of the epoch's state is deferred with every chunk after it, the bodies validated so far are synced, the executor passes the seed block, the deferred headers validate (the restart step's first failure) |
| a3b2049d | the relay flow holds off a class v5 state-wait block: no strike, no N6 refusal memory, the flow stays up (the restart step's second failure: a relay refusal tore down the catch-up's peer and N6-banned it) |
| 7455b8d5 (pool lane) | the open pool's coinbase tags IGNH and IGNW (IGNS and IGNP were the proving sections' tags) |
| 63524e28 | Devnet 3's own EVM chain id 4464 from the class v5 floor (`evm_chain_id_at(net, daa)`: 4463 below, 4464 from it; validation, execution, statements, the mempool at the next DAA, eth_chainId at the tip); every object names its own id (mainnet 4461, testnet 4462, devnet 4463, Devnet 3 4464, suffix N 4461 + N) |
Gates on the line so far (build-2): consensus 123 to 126, core 146 to 154, exec 44 to 46, miner 25, pow 18 to 19, p2p-flows 37 to 38, p2p-lib 20, the kaspad check with tests; the matched pair for the fast-time lane: 959b57c9 (19:45:29Z, /srv/artefacts/v5-pair-959b57c9/, igneumd 519ee6c4…, miner f1a0d9cc…, igneum-pow class-v5 ab6f980b): prelude e0 and e1 class v4 at rung 0, rung 1 by signal at epoch 6, class v5 by signal at epoch 8 (3 of 3 nodes, 11 of 11 ids equal to the CLI's), the stale miner 122 of 122 refused, the digest case PASS; the restart step FAIL (the relay refusal, fixed in a3b2049d); the second pair on 63524e28 builds for the rerun. Open before the v5 object commit: the v5 lane's freeze commit (the census), the restart-step PASS on the second pair, the F8 line on v5, the kaspa-pow pairing on Devnet 3's v5 id e5a4ac5978462156 (Metal fingerprint 82b19cbde8557ea5), and dn3-g1's DAA at the cut. Lease rule (main, 19:2xZ): no box run except through the lease tool; build-remote's runs take slots and pool cores (the class on bounded suites in remote-run.sh's next commit); adversarial sweeps through `lease pool`, classes release > v5 > measure > adv.
## 7. 0.3.17 node items after the 3.0 lane (Horizon queue, 6 October 2026, evening)
Numbering (the shipper, 6 October 2026, 22:4xZ): this tree ships as 0.3.17. 0.3.16 became the corrective cut of 0.3.15's tree the same night (the Mac and PC 1 had taken early 0.3.15 builds), node f1ea7a38 unchanged. The branch keeps its name `ca3-v4-0316` and the commit messages their "0.3.16"; every "0.3.16" in the fork's code comments means this tree.