Skip to content

docs(release): record the LoopX 1.0.5 readiness entry - #4471

Merged
huangruiteng merged 1 commit into
mainfrom
codex/release-readiness-v1.0.5
Sep 15, 2026
Merged

huangruiteng merged 1 commit into
mainfrom
codex/release-readiness-v1.0.5

Conversation

@huangruiteng

Copy link
Copy Markdown
Collaborator

Record the promoted v1.0.5 release in the readiness ledger.

The entry is added after the matching tag (0c3971fea), the GitHub release note, the stable fast-forward, and the focused release canary all agreed. It cites only PRs merged inside the v1.0.4..v1.0.5 range and describes the shipped surfaces: the registration-only first-connect contract, explicit managed-host and steward-executor selection, and the update/runtime reliability fixes.

Validation: examples/release/release-readiness-doc-smoke.py — ok (docs-only change; no runtime behavior, adapter, permission, or public-evidence boundary change).

Add the promoted v1.0.5 record to the release readiness ledger, after the
matching tag, release note, stable ref, and focused release canary agreed.

The entry cites only PRs merged inside the v1.0.4..v1.0.5 range and describes
the shipped surfaces: the registration-only first-connect contract, explicit
managed host and steward executor selection, and the update/runtime
reliability fixes.

Signed-off-by: huangruiteng <14976749+huangruiteng@users.noreply.github.com>

@huangruiteng huangruiteng left a comment

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approval conclusion (author-owned PR; GitHub blocks formal self-approval)

Exact head reviewed: e69479ee0 (re-read immediately before this verdict; codex/release-readiness-v1.0.5 at fa8a0b35b). Review policy revision 3.

动机

v1.0.5 已经发布:tag 0c3971fea、GitHub release body、stable ref、wheel/sdist/checksum 与 PyPI 制品都已落到同一个 release commit 上。仓库的 release readiness ledger 里还停在 v1.0.4 条目,读者无法从这份权威记录看到 1.0.5 发布了什么、依据哪次验证。受影响的是依赖 ledger 判断「哪个版本可信、为什么」的 operator 与后续维护者;compounding cost 是公开历史里出现一个已发布但无记录的版本,下一轮发布无法引用上一轮的验证口径。before/after 很直接:改前 ledger 最后一节是 1.0.4,改后增加 1.0.5 条目并引用 tag SHA 与区间内 PR。更小的修法不成立:不写就等于让 ledger 与实际发布状态不一致;写在 tag 之前则违反仓库既定顺序(tag、release note、stable ref、canary 全部一致后才补记录),也拿不到准确的 tag SHA。

改动思路

唯一被改的权威表面是 docs/product/release-readiness.md 的 promoted-release 列表;条目位置、句式与字段沿用 1.0.4 条目(v1.0.4 on <时间> … at the matching tag (b6d877b0c) … artifacts were verified against the exact release source before stable fast-forwarded)。决策边界是「记录事实」而不是「创造事实」:条目只陈述已经发生并可被公开验证的状态——tag 名与 SHA、区间内 PR 编号、制品校验与 stable 快进。复用判断:没有第二处需要同步的发布记录,docs/update-notes/ 是双周摘要而不是按版本条目(1.0.4 也未新增),因此不新增文件、不改模板、不动 smoke。正面路径是读者从 ledger 追到 tag、PR 与制品;负面路径见下节(写入尚未发布的版本会被 gate 拒绝)。

具体改动

1 个文件、+22 行,纯文档新增,没有控制流、schema、契约或默认值变化;changed_line_classification.behavior_role 记为 documentation-only。

关键内容讲解

  • docs/product/release-readiness.md 新增 v1.0.5 条目:时间、主题(首连登记式契约 + 显式 host 选择 hot-fix)、v1.0.5 tag 与短 SHA 0c3971fea,并把主体改动归因到区间内 PR:首连契约与 harness(#4465)、托管 Turn host(#4443#4451)、steward executor(#4446#4419)、PR review CI 等待(#4452#4434)、不可变 source commit 资格与 runtime 身份(#4460#4423)、quota window clamp(#4384)、held fence(#4429)、frontier 诊断(#4341),以及社区贡献的 #4396#4398#4428#4372#4433#4453#4442
  • 归因只使用 v1.0.4..v1.0.5 区间内已合并的 PR:区间由 git log v1.0.4..origin/main 的 merge commit 与 squash 形式的 (#N) 后缀合并得到,共 43 个 PR,其中社区贡献 22 个(12 位作者)。未列出 @huangruiteng,符合仓库对社区贡献小节的边界要求。
  • 末尾沿用既有句式记录制品验证:wheel、source distribution、checksum 与 PyPI 制品在 stable 快进前已按精确 release source 验证。本次实际制品为 loopx-1.0.5-py3-none-any.whlloopx-1.0.5.tar.gzSHA256SUMS 与桌面产物(dmg/msi/exe/app zip/sig/updater),PyPI 1.0.5 的 wheel/sdist 已可读取。

对主干的风险

最强回归场景是 ledger 里出现一条与真实发布不一致的记录(错误 tag、错误 SHA、或引用未合并 PR),它不会破坏运行时,但会污染公开历史并误导后续发布判断。防止它的路径是条目在 tag、release note、stable 与 canary 全部一致之后写入,并且只引用区间内 PR;可观测点是 examples/release/release-readiness-doc-smoke.py 与公开 tag/PR 元数据;回退方式是 revert 这一个文档 commit。影响半径仅限文档可信度,不触及运行时行为、状态、权限或制品。负面走查由一个独立事实核对完成:条目的 tag 0c3971feaorigin/stable 当前值一致、与 release 的 targetCommitish 一致;列出的 43 个 PR 全部来自 v1.0.4..v1.0.5 区间。比例性:问题的代价是公开记录失真,最小可行修法就是这一条记录,新增机制、状态契约或 CLI 表面为零,判定 proportionate。权限语义不适用(不新增 actor/authority/coordination 语义);default-off 隔离不适用(没有可选能力或开关)。

验证结果:examples/release/release-readiness-doc-smoke.py ok;loopx canary premerge --from-git-diff --timeout-seconds 400 0 failures(8 个 catalog canary、8 个 risk-profile smoke、public-boundary loopx check --scan-path docs/product/release-readiness.md、direct diff checks 全过);同一 head 上 CI 的必需检查全部通过(文档变更下大多数 job 按路径跳过,merge-gate 与 Sign-off、build、dependency-review 通过)。

我的整体评价

这是一条只记录已发生事实的账本条目,与 1.0.4 条目同形同位置,不引入新机制。baseline/head 对比中唯一差异就是新增条目本身;代码量 1 文件/+22,verdict necessary。残余风险只有一条:条目描述的是发布事实,若后续对 1.0.5 做补丁发布,需要新的 tag 与新条目,而不是改写这一条。结论:APPROVE。

English verdict: APPROVE for e69479ee0 on codex/release-readiness-v1.0.5 — a one-file, +22-line documentation-only ledger entry that records the already-published v1.0.5 (tag 0c3971fea, matching stable, matching release target commit) with PR attribution restricted to the v1.0.4..v1.0.5 range and the same grouping, ordering and artifact-verification sentence as the v1.0.4 entry. No runtime, schema, default, permission or artifact change; the strongest risk (a record that disagrees with the real release) is bounded by the tag/stable/target-commit agreement check and by examples/release/release-readiness-doc-smoke.py, and the same head passes a clean loopx canary premerge --from-git-diff (0 failures across 8 catalog canaries, 8 risk-profile smokes and the public-boundary scan) plus the required CI checks. Residual risk: any later 1.0.5 patch needs its own tag and its own ledger entry rather than an edit to this one.

@huangruiteng huangruiteng left a comment

Copy link
Copy Markdown
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approval conclusion (author-owned PR; GitHub blocks formal self-approval)

Exact head reviewed: e69479ee0c9466bf80c662a4f7d5fdfd4e11def3 on codex/release-readiness-v1.0.5 (base fa8a0b35b, re-read immediately before this verdict). Review policy revision 3.

动机

v1.0.5 已经发布:tag 0c3971fea、GitHub release body、stable ref、wheel/sdist/checksum 与 PyPI 制品都已落到同一个 release commit 上。仓库的 release readiness ledger 里还停在 v1.0.4 条目,读者无法从这份权威记录看到 1.0.5 发布了什么、依据哪次验证。受影响的是依赖 ledger 判断「哪个版本可信、为什么」的 operator 与后续维护者;compounding cost 是公开历史里出现一个已发布但无记录的版本,下一轮发布无法引用上一轮的验证口径。before/after 很直接:改前 ledger 最后一节是 1.0.4,改后增加 1.0.5 条目并引用 tag SHA 与区间内 PR。更小的修法不成立:不写就等于让 ledger 与实际发布状态不一致;写在 tag 之前则违反仓库既定顺序(tag、release note、stable ref、canary 全部一致后才补记录),也拿不到准确的 tag SHA。

改动思路

唯一被改的权威表面是 docs/product/release-readiness.md 的 promoted-release 列表;条目位置、句式与字段沿用 1.0.4 条目(v1.0.4 on <时间> … at the matching tag (b6d877b0c) … artifacts were verified against the exact release source before stable fast-forwarded)。决策边界是「记录事实」而不是「创造事实」:条目只陈述已经发生并可被公开验证的状态——tag 名与 SHA、区间内 PR 编号、制品校验与 stable 快进。复用判断:没有第二处需要同步的发布记录,docs/update-notes/ 是双周摘要而不是按版本条目(1.0.4 也未新增),因此不新增文件、不改模板、不动 smoke。正面路径是读者从 ledger 追到 tag、PR 与制品;负面路径见下节(写入尚未发布的版本会被 gate 拒绝)。

具体改动

1 个文件、+22 行,纯文档新增,没有控制流、schema、契约或默认值变化;changed_line_classification.behavior_role 记为 documentation-only。

关键内容讲解

  • docs/product/release-readiness.md 新增 v1.0.5 条目:时间、主题(首连登记式契约 + 显式 host 选择 hot-fix)、v1.0.5 tag 与短 SHA 0c3971fea,并把主体改动归因到区间内 PR:首连契约与 harness(#4465)、托管 Turn host(#4443#4451)、steward executor(#4446#4419)、PR review CI 等待(#4452#4434)、不可变 source commit 资格与 runtime 身份(#4460#4423)、quota window clamp(#4384)、held fence(#4429)、frontier 诊断(#4341),以及社区贡献的 #4396#4398#4428#4372#4433#4453#4442
  • 归因只使用 v1.0.4..v1.0.5 区间内已合并的 PR:区间由 git log v1.0.4..origin/main 的 merge commit 与 squash 形式的 (#N) 后缀合并得到,共 43 个 PR,其中社区贡献 22 个(12 位作者)。未列出 @huangruiteng,符合仓库对社区贡献小节的边界要求。
  • 末尾沿用既有句式记录制品验证:wheel、source distribution、checksum 与 PyPI 制品在 stable 快进前已按精确 release source 验证。本次实际制品为 loopx-1.0.5-py3-none-any.whlloopx-1.0.5.tar.gzSHA256SUMS 与桌面产物(dmg/msi/exe/app zip/sig/updater),PyPI 1.0.5 的 wheel/sdist 已可读取。

对主干的风险

最强回归场景是 ledger 里出现一条与真实发布不一致的记录(错误 tag、错误 SHA、或引用未合并 PR),它不会破坏运行时,但会污染公开历史并误导后续发布判断。防止它的路径是条目在 tag、release note、stable 与 canary 全部一致之后写入,并且只引用区间内 PR;可观测点是 examples/release/release-readiness-doc-smoke.py 与公开 tag/PR 元数据;回退方式是 revert 这一个文档 commit。影响半径仅限文档可信度,不触及运行时行为、状态、权限或制品。负面走查由一个独立事实核对完成:条目的 tag 0c3971feaorigin/stable 当前值一致、与 release 的 targetCommitish 一致;列出的 43 个 PR 全部来自 v1.0.4..v1.0.5 区间。比例性:问题的代价是公开记录失真,最小可行修法就是这一条记录,新增机制、状态契约或 CLI 表面为零,判定 proportionate。权限语义不适用(不新增 actor/authority/coordination 语义);default-off 隔离不适用(没有可选能力或开关)。

验证结果:examples/release/release-readiness-doc-smoke.py ok;loopx canary premerge --from-git-diff --timeout-seconds 400 0 failures(8 个 catalog canary、8 个 risk-profile smoke、public-boundary loopx check --scan-path docs/product/release-readiness.md、direct diff checks 全过);同一 head 上 CI 的必需检查全部通过(文档变更下大多数 job 按路径跳过,merge-gate 与 Sign-off、build、dependency-review 通过)。

我的整体评价

这是一条只记录已发生事实的账本条目,与 1.0.4 条目同形同位置,不引入新机制。baseline/head 对比中唯一差异就是新增条目本身;代码量 1 文件/+22,verdict necessary。残余风险只有一条:条目描述的是发布事实,若后续对 1.0.5 做补丁发布,需要新的 tag 与新条目,而不是改写这一条。结论:APPROVE。

English verdict: APPROVE for e69479ee0c9466bf80c662a4f7d5fdfd4e11def3 on codex/release-readiness-v1.0.5 — a one-file, +22-line documentation-only ledger entry that records the already-published v1.0.5 (tag 0c3971fea, matching stable, matching release target commit) with PR attribution restricted to the v1.0.4..v1.0.5 range and the same grouping, ordering and artifact-verification sentence as the v1.0.4 entry. No runtime, schema, default, permission or artifact change; the strongest risk (a record that disagrees with the real release) is bounded by the tag/stable/target-commit agreement check and by examples/release/release-readiness-doc-smoke.py, and the same head passes a clean loopx canary premerge --from-git-diff (0 failures across 8 catalog canaries, 8 risk-profile smokes and the public-boundary scan) plus the required CI checks. Residual risk: any later 1.0.5 patch needs its own tag and its own ledger entry rather than an edit to this one.

@huangruiteng
huangruiteng merged commit c3cee40 into main Sep 15, 2026
18 checks passed
@huangruiteng
huangruiteng deleted the codex/release-readiness-v1.0.5 branch September 15, 2026 19:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant