NiPoPoW: fail closed when V1 branch context is unavailable - #16
Open
a-shannon wants to merge 14 commits into
Open
NiPoPoW: fail closed when V1 branch context is unavailable#16a-shannon wants to merge 14 commits into
a-shannon wants to merge 14 commits into
Conversation
This was referenced Aug 6, 2026
Open
a-shannon
marked this pull request as ready for review
August 6, 2026 22:39
Author
|
Review acknowledgement: many thanks to Muadib for the thorough initial security review, the reproducible test cases, and the independent re-audit of this exact head ( |
a-shannon
pushed a commit
to a-shannon/ergo-node-rust
that referenced
this pull request
Aug 15, 2026
…ists "Parallel validation — concurrent transaction evaluation within blocks, pipelined across blocks" described deferred evaluation, which v0.8.0 removed. Intra-block parallelism is unchanged; the cross-block pipeline is gone, and what replaces it is worth stating positively — a block that reaches state.redb has already had its scripts checked. The at-tip figures (7.3 GB -> 1.35 GB, "~80% reduction") were measured in the deferred era. Remeasured today on mainnet at 1.85M blocks with the same cache_mb=1024 / synced_cache_mb=128: 1.9 GB cold sync -> 950 MB at tip. The config the numbers came from is now stated, because without it they are not reproducible. The config sample gains cache_store_pct and says that cache_mb changed meaning in v0.8.0 — it is the one breaking change an existing operator hits silently, since the old key still parses and just means something else. Both upstream PR sections were enumerated by hand and had drifted: on ergo_avltree_rust mwaddip#10 and mwaddip#11 are CLOSED while listed as open, and mwaddip#13/mwaddip#14/mwaddip#16/ #24/#27 plus six merged were missing entirely; on sigma-rust #852 has merged and #854 was closed (superseded by #876). Verified with `gh pr list`, not from memory. At 45 open PRs the list turns over faster than a README does, so it is now a live query plus the merged set and the curated avltree entries that matter to this node. Also records that upstream `develop` is frozen and `v0.30.0` is the live integration branch — every merge landed there, so GitHub reporting all 45 open PRs as MERGEABLE means only "mergeable into a branch nobody moves". Checked and left alone: "38 JVM-compatible endpoints" and "43 endpoints" are both correct (38 + 5 of our own = 43), and no operator doc or man page documents the removed script_eval / eval_backlog_* keys. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
a-shannon
pushed a commit
to a-shannon/ergo-node-rust
that referenced
this pull request
Aug 15, 2026
Every "Last Active" date was wrong, and wrong in a way that mattered: the column said Feb 2026 for the sigma-rust crates and Dec 2021 for ergo-nipopow, while the GitHub API reports the repo last pushed 2026-08-04. a-shannon is currently landing work in the nipopow files, so "Dec 2021" would have told a reader that module was abandoned. Renamed the column to "Upstream last push" and stamped the check date, because the old header invited exactly the drift it accumulated. Also noted what the version column actually is: the last crates.io release (ergo-lib-v0.28.0, published 2024-08-09), NOT what the build resolves — we consume the fork by rev. Those two numbers were silently measuring different things. The AVL row cited PR mwaddip#10 for the Resolver change. mwaddip#10 is closed; the live one is mwaddip#16, and the fork now carries persistent-prover support as #27 — the commit this release pins for. REST API said 23 endpoints in two places; it is 43 (38 JVM-compatible plus /debug/memory, three p2p-capture routes and /stats/p2p), which is what the README and openapi.yaml already said. Phase 6 said "Released as v0.4.x" with no indication four minor releases had happened since. ergoscript-compiler is marked available-but-unused: it resolves to nothing in Cargo.lock, and the table is an ecosystem inventory rather than a dependency list, so the entry stays with that stated. sigma-ser's version was a dash; it is 0.19.0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
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.
Summary
V1 boundary
Legacy V1 represents sparse difficulty context as height-selected headers, but does not bind those headers as members of the branch selected by the proof. The node therefore treats V1 bootstrap authorization as unavailable: verification returns NipopowBootstrapDisabled, synchronization returns BootstrapDisabled, and processing stops before candidate comparison or state installation.
The verified-envelope, exact-suffix, and bounded context plumbing remains as non-authorizing infrastructure. A branch-authenticated V2 format is outside this change.
Coordination
Validation
Independent maintainer review and upstream CI are pending.