Rename default_version from 1.0.0 to stable

[jnasbyupgrade]

Re-lands the content of #26, which was accidentally lost.

#26 ("Rename default_version from 1.0.0 to stable") legitimately merged into the fix-cat-tools-install branch, creating commit f08db32. That branch was later reset/rebased back to an earlier point before it itself was squash-merged into master (as part of #10), so f08db32's changes never actually reached master despite #26 showing as MERGED on GitHub. The commit object still exists in the repo but isn't an ancestor of master or any other branch.

This PR cherry-picks f08db32's content back onto current master (resolving conflicts in HISTORY.asc caused by unrelated history added since then). Same intent as #26:

• extension_drop.control: default_version = 'stable' (was 1.0.0, which was never actually tagged or published to PGXN — the only real release ever published was 0.1.1 in 2017). stable is an honest label for "current unreleased source", matching pg_count_nulls#13's convention, until a real release is cut.
• sql/extension_drop--1.0.0.sql removed — since stable is a pseudo-version, its versioned sql file (sql/extension_drop--stable.sql) is now generated at build time rather than committed (see .gitignore update).
• META.in.json / META.json: version fields updated to stable.
• HISTORY.asc: heading renamed from 1.0.0 to STABLE.

Verified with make test (all test groups pass) and make lint (zero findings) against current master.

See #26 for the original rationale/discussion.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 39b9d80f-fd81-4a2f-b802-4125ecade8e8

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Claude finished @jnasbyupgrade's task in 2m 6s —— View job

---

Claude's Review

• Gather context (PR diff, CLAUDE.md, changed files)
• Review .gitignore, HISTORY.asc, META.in.json, META.json, extension_drop.control changes
• Review deletion of sql/extension_drop--1.0.0.sql and generated-file implications
• Verify build/test process still produces sql/extension_drop--stable.sql correctly
• Post review feedback

This is a clean, mechanical re-land of #26's content, and the change is internally consistent:

• extension_drop.control: default_version = 'stable' ✓
• META.in.json / META.json: both version fields updated in lockstep (version and provides.extension_drop.version) ✓
• HISTORY.asc: STABLE heading + underline (------, 6 dashes matching 6 chars) ✓
• .gitignore: adds sql/extension_drop--stable.sql, with a comment correctly explaining that versioned SQL files are committed once actually released — matches RELEASE.md's existing "Ongoing development" workflow, which already describes renaming a STABLE-suffixed file at release time (git mv sql/<ext>--<last-released-version>--stable.sql ...), so this PR aligns with docs that were apparently already written for this convention.
• sql/extension_drop--1.0.0.sql deletion: confirmed via pgxntool/base.mk (EXTENSION__CURRENT_VERSION__FILES, generated from control.mk) that the versioned SQL file is mechanically regenerated from default_version at build time — nothing else in the repo hardcodes the 1.0.0 filename, so nothing breaks from the deletion. Searched .github/, Makefile, *.mk for other 1.0.0 / extension_drop--1.0.0 references — none found except one stale comment (below).

One pre-existing stale comment worth fixing while you're here (not introduced by this PR, but this PR is exactly what invalidates it):

Makefile:110-112 says:

# `sql/ test/`: sql/extension_drop--1.0.0.sql is a frozen, already-released
# version file (RELEASE.md's "Ongoing development" section -- once a version
# is released, its sql/<ext>--<version>.sql is never hand-edited again), so

This now references a file this PR deletes, and its premise ("already-released") was already inaccurate before this PR too — per this PR's own description, 1.0.0 "was never actually tagged or published to PGXN." Doesn't break make lint (the target scoping — LINT_TARGETS = sql/extension_drop.sql test/ — doesn't depend on the comment), but leaving a comment that names a now-deleted file with a factually wrong rationale is confusing for the next reader. Worth a one-line follow-up: e.g. "the versioned SQL file is auto-generated (see pgxntool/base.mk), so linting it would produce noise/false findings anyway."

Fix this →

No functional issues found. RELEASE.md's remaining 1.0.0 mentions (lines 71, 95, 99) are generic illustrative examples (PGXN meta-spec version, tag/zip naming convention) unrelated to this extension's actual version — no changes needed there.

[github-actions]

Claude finished @jnasbyupgrade's task in 1m 8s —— View job

---

Claude's Review

• Gather context (PR diff, CLAUDE.md, changed files)
• Review .gitignore, HISTORY.asc, META.in.json, META.json, extension_drop.control changes
• Review deletion of sql/extension_drop--1.0.0.sql and generated-file implications
• Check whether prior feedback (stale Makefile comment) was addressed
• Post review feedback

This is a clean, mechanical re-land of #26's content, and the change is internally consistent:

• extension_drop.control: default_version = 'stable' ✓
• META.in.json / META.json: both version fields updated in lockstep (version and provides.extension_drop.version) ✓
• HISTORY.asc: STABLE heading + underline (------, 6 dashes matching 6 chars) ✓
• .gitignore: adds sql/extension_drop--stable.sql with a comment explaining that it's a mechanical copy of sql/extension_drop.sql while stable remains a pseudo-version, so committing it would just duplicate that file — consistent with RELEASE.md's existing workflow.
• sql/extension_drop--1.0.0.sql deletion: confirmed via pgxntool/base.mk (EXTENSION__CURRENT_VERSION__FILES, generated from control.mk) that the versioned SQL file is mechanically regenerated from default_version at build time — nothing else in the repo hardcodes the 1.0.0 filename, so nothing breaks from the deletion.

Pre-existing stale comment still present (not introduced by this PR, flagged in the prior review pass, still unaddressed):

Makefile:110-114 still reads:

# `sql/ test/`: sql/extension_drop--1.0.0.sql is a frozen, already-released
# version file (RELEASE.md's "Ongoing development" section -- once a version
# is released, its sql/<ext>--<version>.sql is never hand-edited again), so
# linting it would produce permanent, unfixable findings and make `make
# lint` unusable as a CI gate. Lint the hand-maintained source instead.

This now references a file this PR deletes, and its premise ("already-released") was already inaccurate before this PR too — per this PR's own description, 1.0.0 "was never actually tagged or published to PGXN." It doesn't break make lint (the target scoping — LINT_TARGETS = sql/extension_drop.sql test/ — doesn't depend on the comment), but it's confusing for the next reader. Since this PR doesn't touch the Makefile, this is optional cleanup rather than a blocker for this PR.

Fix this →

No functional issues found. RELEASE.md's remaining 1.0.0 mentions are generic illustrative examples (PGXN meta-spec version, tag/zip naming convention) unrelated to this extension's actual version — no changes needed there. This matches the assessment from the prior review pass on this PR; nothing has regressed and the flagged Makefile comment remains the only open item.