Renumber release as v0.1.6 - #49
Conversation
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 040db67e82
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). | ||
|
|
||
| ## [0.1.7] - 2026-09-10 | ||
| ## [0.1.6] - 2026-09-10 |
There was a problem hiding this comment.
Merge the duplicate 0.1.6 changelog sections
Renumbering this heading creates two separate 0.1.6 releases—one dated 2026-09-10 and another dated 2026-06-24—so readers and changelog tooling cannot unambiguously determine which changes belong to the public release. Since the earlier version was reportedly never published, consolidate both sets of notes beneath one 0.1.6 heading (or otherwise mark the earlier section as unreleased) before releasing.
Useful? React with 👍 / 👎.
v0.1.4 (2026-06-22) and v0.1.5 (2026-06-24) were both tagged and registered while their entries still sat under `## [Unreleased]`, so neither ever got a version heading. Commit 7c10ea9 then retitled that whole accumulated block `## [0.1.6] - 2026-06-24` and bumped to 0.1.6 -- but that 0.1.6 was never registered. When #49 later renumbered the September release from 0.1.7 back down to 0.1.6, it renamed only the new section, leaving two `## [0.1.6]` headings and no 0.1.4/0.1.5 at all. Split the June block back along the real tag boundaries, determined by diffing CHANGELOG.md as it stood at each tag: - `[0.1.5] - 2026-06-24` -- DuckDB Parquet migration, ZSTD default, and the OHLCV/STATUS/IMBALANCE/DEFINITION, empty-record-set and create_metadata_from_dataframe fixes. - `[0.1.4] - 2026-06-22` -- symbol resolution helpers, plus the MBP-1/MBP-10 FieldError, read_stat_msg sizing, stat_to_dataframe, symbol-mapping-interval and invalid-enum-byte fixes (#32-#35, #40). The replay feature moves up into `[0.1.6] - 2026-09-10`: it was written in June but first reached the registry in that release. Also adds the missing [0.1.7]/[0.1.5]/[0.1.4] compare links. No entry text was changed -- only the headings it sits under. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The General registry currently ends at v0.1.5; the repository's earlier v0.1.6 value was never tagged or registered. This renumbers the corrected, fully tested release from v0.1.7 to v0.1.6 so public versions remain sequential.
No implementation code changes.
Validation