8000 release(rc.92): version stamp, CHANGELOG, ROADMAP roll, landing stamps by xerj-org · Pull Request #1256 · xerj-org/xerj · GitHub
[go: up one dir, main page]

Skip to content

release(rc.92): version stamp, CHANGELOG, ROADMAP roll, landing stamps - #1256

Merged
xerj-org merged 5 commits into
mainfrom
release/rc92
Oct 8, 2026
Merged

xerj-org merged 5 commits into
mainfrom
release/rc92

Conversation

@xerj-org
@xerj-org xerj-org commented Oct 8, 2026
Copy link
Copy Markdown
Owner

Window: #1246 (xerj code own-field recall legs + quarter-union gate, #1244) and #1251 (multi-family family-route coverage + coverage-ranked cap, #1244).

release-notes-gate.sh passes locally (exit 0). After merge: tag from the fetched merge ref; the release workflow owns the release.

Window: #1246 (xerj code own-field recall legs + quarter-union gate,
#1244), #1251 (multi-family family-route coverage + coverage-ranked cap,
#1244).

- engine/Cargo.toml 1.0.0-rc.91 -> 1.0.0-rc.92, Cargo.lock refreshed
  offline (58 members re-pinned).
- CHANGELOG: rc.92 section. #1244 stays open — completion is the
  pre-registered G7 suites passing on rebuilt corpora; both rebuilds
  (cve-records on :19285, vuln-fix-commits on :19295) were in flight at
  cut time.
- ROADMAP: review-line roll to rc.92; twelve open (ten carried + #1250
  breaker-cap + #1254 unassigned-file drop, both filed inside the
  window). #1250 carries the same-day attribution correction (parent
  RSS breaker, boot-time tier cap, not max_query_memory_mb) and the
  measured idle flush loop (644/858 GB writes on two parked A/B nodes).
  #1183 gained the 2026-10-08 standing-rebuild outcome: abort on a
  store-lost segment, read-back half now working, success footer still
  prints after ok=false (348-doc overstatement, comment on the issue).
- Landing stamps RC.91 -> RC.92 across the 54 stamped pages (both
  patterns: data-latest-tag span + footer version); sitemap regenerated,
  185 URLs, no drift (lastmod rolls only).
@cla-bot cla-bot Bot added the cla-signed label Oct 8, 2026
- Both ### Fixed entries cite #1244, which is open; each now says what
  remains open (the pre-registered G7 suites on the rebuilt corpora).
- #1252 is the changelog-only carry PR; exempted with a reason inside
  the rc.92 section per the gate's exemption format.
The ROADMAP carried the retracted diagnosis (silently unindexed) into
both the review line and the open-defects shortlist. The corrected
mechanism, measured on the live index today: all 11 .proto files ARE
indexed (79 txt-lines chunk records in opentelemetry-proto-docs); the
G7 FAIL 1/5 is a ranking outcome — the needle chunk ranks 19/50 at
17.99 vs cutoff 23.14, outcompeted by whole-document semconv prose.
Issue #1254 retitled and corrected (issuecomment-6061349808); graded
file fixed on #1253 (47065c2).
…ect the ROADMAP mechanism line

#1258 enters the rc.92 window: the corpus-declared query.text_weight
fix for #1254, measured on otel-proto (four G7 needles 19/absent/39/23
-> rank 1 on unchanged index bytes, suite re-graded 5/5, zero drift on
no-hint corpora). #1254 stays open for the code_files=0 labeling gap,
stated in both the CHANGELOG entry and the shortlist line.

The ROADMAP's two #1254 spots still carried the SECOND retracted
mechanism (txt-lines chunk-splitting: fragments vs whole-document
prose). That mechanism was disproven by simulation before the fix was
built — a whole-document route ranked the needle 25th, worse than the
real chunks' 19th, and a definition-sectioned route 25/39/23/19 — so
rolling the tag with it in the record would repeat the honesty error
the first retraction corrected. Both spots now state the measured
variable (the text^0.5 recall-leg weight) and the two-step retraction.

release-notes-gate.sh: PASS (3 issue citations checked, #1254 entry
names what remains open; #1258 cited ahead of its merge).
@xerj-org
xerj-org merged commit f767cdb into main Oct 8, 2026
23 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant

0