Pin the active Bifrost candidate to v0.10.6 - #169
Merged
Conversation
v0.10.6 carries the PHP static-property usage fix from bifrost-dev#2535, the usage-side half of the declaration work in #2416. Confirmed against the public mirror: push_member_hit reads `let start = node.start_byte();` at v0.10.6, so the sigil is no longer skipped and php-parity-static-property-access should report both usage ranges in full. The revision moves in the same five places as the previous bump, for the same reasons: reference-image.sh refuses to build when the candidate registry and the reference-environment manifest disagree, the reproduction contract and the freeze-shard test assert the exact pair, and the historical v1 page names the active candidate so readers know its own tables are older evidence. The v0.10.2 pin for candidates-v0.3.0.json is deliberately untouched; that registry belongs to a frozen release. The development page's caveat said every Bifrost measured since the PHP audit failed at least one of those cases. That stops being true here, so it now says no version contemporary with that run satisfied them and both pass from v0.10.6. Local behavioural verification did not complete. Runs against a staged corpus on this machine now report the analyzer unavailable for every language, and the same corpus scored 98 of 110 four days ago. The cause is local: v0.10.5 fails identically on the same staged corpus, the repository tree passes 142 cases under v0.10.6, and the corpora are byte-identical apart from caches a run writes itself. Work-dir reuse, stale caches, and disk pressure were each ruled out by reproduction. CI stages its own corpus per run and has produced a scored legacy report that way, so this does not implicate the pin -- but it does mean the numbers behind this bump are static checks, not a measured run. Co-Authored-By: Claude Opus 5 <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.
Unblocks the last prerequisite on #155.
v0.10.6 carries the PHP static-property usage fix from BrokkAi/bifrost-dev#2535 —
the usage-side half of the declaration work in #2416. Verified against the
public mirror:
push_member_hitreadslet start = node.start_byte();atv0.10.6, so the sigil is no longer skipped.
v0.10.6is6624e883. Note the tag sits on a projection commit whose parentis
79ea68cf, the commit the release-readiness run built.Five pin sites, unchanged reasoning
adapters/candidates.json,containers/reference/v1/manifest.json(the imageis addressed by the analyzer identity),
validate-reproduction-contract.sh,tests/test_freeze_shards.py, and the historical v1 page that names the activecandidate. The
v0.10.2pin forcandidates-v0.3.0.jsonis deliberatelyuntouched — that registry belongs to a frozen release.
Also updates the #157 caveat, which claimed every Bifrost measured since the
PHP audit failed at least one of those cases. That stops being true here.
Testing — and an honest gap
Static checks pass:
validate-reproduction-contract.shexits 0, andunittest discoverruns 39 tests green.Local behavioural verification did not complete. Runs against a staged
corpus on my machine now report
cpp_analyzer_unavailableandsymbol_resolution_failedfor every language. The cause is local, not this pin:.bifrostcaches, and disk pressure were each ruledout by reproduction, not assumption
CI stages a fresh corpus per run and has produced a scored legacy report that
way (32177552346
gave 97/12), so the staged path works there. But the numbers behind this bump
are static checks, not a measured run — the expected set for #155 is still the
10 + 1 predicted from the v0.10.5 measurement in #160, not re-measured under
v0.10.6.
🤖 Generated with Claude Code