rafatahmed · Oct 20, 2026 · Oct 20, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,6 +1,10 @@
 # Changelog
 
-## 0.5.0 — Unreleased
+## Unreleased
+
+- Record the 0.5.0 publication. The published 0.5.0 manual and changelog still carry the release-candidate wording ('not yet published', 'Unreleased'); they are corrected here, and the publishing procedure now has an explicit finalize step before tagging (#71).
+
+## 0.5.0 — 2026-10-20
 
 - Publish Benchmark v1 snapshot 2026-10-20-linux-df0d30b (hosted run 39515183621 on `main`, after #62, #65 and #69), the first with every correctness gate passing: 7/7 controlled NNLS cases, 21/21 stored chemistry comparisons, 17/17 timing workloads; complete recipes 34/52 (was 29/52). The results page no longer calls a fully passing scorecard 'not fully passing' (#66).
 - Fix `recommend()` slowness and start-dependent recipes under incidental-ion ceilings (#68). Two root causes, both measured against SciPy on every fit: the NNLS optimality test used one tolerance scaled by the whole matrix (max |A| times |b|), so a heavily weighted row hid real descent directions in the other columns and NNLS stopped early (doses off by up to 4.2e-3 g/L); it now bounds each column's dual weight by the rounding error of its own terms. And a binding ceiling was enforced by a 1024-weight penalty, which left the problem so flat in double precision that doses were fixed only to ~2e-3 g/L and depended on the solver's starting point. A binding ceiling is now a constraint: a weight-8 row whose aim is shifted by the Lagrange multiplier, found by a bracketed secant search, with one Gram cache per search. Every ceiling fit now matches SciPy to 1.4e-10 g/L, warm and cold starts agree to 7e-9 g/L, and binding ions land on their aim. Over the 52 benchmark recipes `recommend()` is 34% faster with the default ammonium limit (4.20 -> 2.78 s) and unchanged with a 3 mmol/L Cl ceiling (3.77 -> 3.87 s); the Benchmark v1 formulation workload falls from about 30 to 16 ms. Pass counts are unchanged by default (34/52 recipes, 571/636 targets) and rise with the Cl ceiling (41 -> 42 recipes, 598 -> 599 targets), with no limit violations; the 10 recipes that change have equal or better target fit.

diff --git a/docs/development-status.md b/docs/development-status.md
@@ -1,6 +1,6 @@
 # Water-to-delivery development status
 
-Updated 2026-10-21. **0.5.0 is prepared, not published** (see the [0.5.0 preparation record](site/evidence/release-evidence.md#050-preparation---2026-10-21)). **0.4.0 is published on PyPI** (tag `v0.4.0`; Apache-2.0, measured-pH water balancing) through the [hardened release workflow](publishing.md); see the [0.4.0 publication evidence](site/evidence/release-evidence.md#040-publication---2026-10-17). Before it, **0.3.1** was published (tag `v0.3.1`); see the
+Updated 2026-10-21. **0.5.0 is published on PyPI** (tag `v0.5.0`; see the [0.5.0 publication evidence](site/evidence/release-evidence.md#050-publication---2026-10-20)). **0.4.0 was published on PyPI** (tag `v0.4.0`; Apache-2.0, measured-pH water balancing) through the [hardened release workflow](publishing.md); see the [0.4.0 publication evidence](site/evidence/release-evidence.md#040-publication---2026-10-17). Before it, **0.3.1** was published (tag `v0.3.1`); see the
 [publication evidence](site/evidence/release-evidence.md#031-publication---2026-10-17).
 The matrix below describes the 0.3.1 capabilities and their boundaries.
 

diff --git a/docs/publishing.md b/docs/publishing.md
@@ -44,12 +44,17 @@ pushing a tag.
    reference; write the changelog section; record readiness and evidence.
    Merge it by pull request after CI passes.
 2. Confirm the publication decision with the owner. Merging is not approval.
-3. Create the GitHub Release for tag `v<version>` on the merged `main`
+3. **Finalize the wording before tagging** (#71): on a short branch, date the
+   changelog section and change the manual's version line from "release
+   candidate ... not yet published" to "released <date>"; merge it after CI
+   passes. The tag must point at this commit, because the packaged manual and
+   the sdist changelog are fixed at upload.
+4. Create the GitHub Release for tag `v<version>` on that merged `main`
    commit, with the changelog section as its notes. Publishing it starts the
    workflow.
-4. Review the run: the guard notice, all test jobs, the live replay and the
+5. Review the run: the guard notice, all test jobs, the live replay and the
    artifact hashes in the build summary. Then approve the `pypi` deployment.
-5. After upload: check PyPI metadata, install `flahax==<version>` in a fresh
+6. After upload: check PyPI metadata, install `flahax==<version>` in a fresh
    environment, run the manual's examples and both CLI entry points, and
    record the date, run URL and hashes in [release evidence](site/evidence/release-evidence.md).
 
@@ -76,18 +81,21 @@ Dashboard settings, recorded here because they live outside the repository:
 - The default version is `latest` until that release, then `stable`.
 - Never put a Read the Docs token in the repository.
 
-## Checklist (in progress: 0.5.0)
+## Checklist (last run: 0.5.0)
 
 - [x] Choose the version and align metadata, runtime version, citation, manual and changelog.
 - [x] Review compatibility, scientific provenance and distribution contents.
 - [x] Pass local tests with live PHREEQC and isolated wheel/sdist verification.
-- [ ] Observe successful hosted Python 3.11–3.13 checks on the merged revision.
-- [ ] Obtain explicit publication authorization.
-- [ ] Publish the `v0.5.0` GitHub Release; review the run; approve the deployment.
-- [ ] Verify PyPI metadata and a fresh installed-package check; record the date, run and hashes.
-- [ ] Confirm the Read the Docs `stable` version builds `v0.5.0` and serves the benchmark pages (#48).
+- [x] Observe successful hosted Python 3.11–3.13 checks on the merged revision.
+- [ ] Date the changelog and the manual's version line, and merge, before tagging. **Missed for 0.5.0** (#71); do it for every release.
+- [x] Obtain explicit publication authorization.
+- [x] Publish the `v0.5.0` GitHub Release; review the run; approve the deployment.
+- [x] Verify PyPI metadata and a fresh installed-package check; record the date, run and hashes.
+- [x] Confirm the Read the Docs `stable` version builds `v0.5.0` and serves the benchmark pages (#48).
+
+Completed 2026-10-20 apart from the finalize step. Copy this checklist, with that step, for the next release.
 
-## Checklist (last run: 0.4.0)
+## Checklist (0.4.0)
 
 - [x] Choose the version and align metadata, runtime version, citation, manual and changelog.
 - [x] Review compatibility, scientific provenance and distribution contents.
@@ -125,6 +133,11 @@ package was uploaded.
 
 ##
8000
 Release history
 
+- **0.5.0**, 2026-10-20: published through the hardened workflow (run
+  `39517327774`) from `692108b`; the deployment was approved at the owner's
+  instruction after artifact review, and verified by a fresh installation from
+  PyPI. The packaged manual and changelog still say "not yet published" (#71).
+  See the [0.5.0 publication evidence](site/evidence/release-evidence.md#050-publication---2026-10-20).
 - **0.4.0**, 2026-10-17: first Apache-2.0 release and first upload through the
   hardened workflow with Trusted Publishing (run `39043172800`), approved by
   the owner, verified by a fresh installation from PyPI. See the

diff --git a/docs/release-readiness.md b/docs/release-readiness.md
@@ -2,8 +2,10 @@
 
 The current preparation comes first; dated audits below it are history.
 
-## 0.5.0 - preparation 2026-10-21
+## 0.5.0 - published 2026-10-20
 
+Published from `692108b` (tag `v0.5.0`); see the
+[publication evidence](site/evidence/release-evidence.md#050-publication---2026-10-20).
 Prepared on `release/0.5.0` from `main` at `9bea54a`. The evidence is the
 [0.5.0 preparation record](site/evidence/release-evidence.md#050-preparation---2026-10-21).
 

diff --git a/docs/site/evidence/release-evidence.md b/docs/site/evidence/release-evidence.md
@@ -26,6 +26,39 @@ They are kept here because the records cite them.
 | 0.3.1 | 2026-10-17 | Published | [Publication](#031-publication-2026-10-17), [release audit](#031-release-audit-2026-10-17), [documentation audit](#031-documentation-audit-2026-10-02), [preparation](#031-preparation-2026-10-01) |
 | 0.3.0 | 2026-09-27 | Yanked 2026-10-17 | [Publication exception](#030-authorized-publication-exception-2026-09-27), [automated evidence](#030-automated-evidence), [P2/P3 acceptance](#p2p3-end-to-end-numerical-acceptance), [continuation](#package-and-delivery-continuation-2026-09-27), [cleanup](#030-release-readiness-cleanup), [preparation](#030-preparation) |
 
+## 0.5.0 publication (2026-10-20)
+
+0.5.0 was published to PyPI on **2026-10-20** (23:51 UTC) by the hardened release
+workflow (run [`39517327774`](https://github.com/rafatahmed/Flahax/actions/runs/39517327774)),
+triggered by the GitHub Release for tag `v0.5.0` on `main` at `692108b` (merge of
+[PR #70](https://github.com/rafatahmed/Flahax/pull/70)). Hosted CI had passed on `692108b`
+([run 39517146974](https://github.com/rafatahmed/Flahax/actions/runs/39517146974): Python
+3.11, 3.12 and 3.13, live PHREEQC replay, docs). The owner authorized publication
+and instructed approval of the `pypi` deployment, which was then approved through
+the GitHub API after the review below.
+
+| Check | Result |
+|---|---|
+| Release guard | Tag `v0.5.0` matched `pyproject.toml`; "flahax 0.5.0 is not on PyPI; continuing to verification" |
+| Workflow gates on `692108b` | Tests on Python 3.11, 3.12 and 3.13 with clean wheel/sdist verification, live PHREEQC replay, one build with `twine check --strict` (both PASSED): all passed before the approval gate |
+| 
8000
Pre-approval artifact review | The workflow's artifacts were downloaded and inspected before approval. `USER_GUIDE.md`, `API_REFERENCE.md`, `chemistry_25c.json` and `library.json` in both the wheel and the sdist match the hashes in the 0.5.0 preparation record. Metadata: version 0.5.0, `License-Expression: Apache-2.0`, `Requires-Python: >=3.11`, no runtime dependency. Sdist 468 files with no PDF and no `docs/publication/` file |
+| PyPI upload | Wheel 2026-10-20 23:51:03 UTC (172,854 bytes); sdist 23:51:05 UTC (5,117,131 bytes). PyPI reports 0.5.0 as the latest version, licence `Apache-2.0`, `Requires-Python >=3.11`, not yanked |
+| Hash identity | The SHA-256 values on PyPI equal those of the reviewed workflow artifacts, and a fresh `pip download` of the wheel matches |
+| Fresh install from PyPI | New Python 3.14.0 virtual environment outside the checkout, `pip install flahax==0.5.0`. The release smoke test from `tools/verify_distribution.py` passed: version, all public exports, nine manual examples, 30 catalogue products, both CLI entry points, planning EC 1.15 mS/cm; installed resource hashes as below |
+| Documentation | Read the Docs `stable` serves 0.5.0, including [Benchmark v1 results](https://flahax.readthedocs.io/en/stable/evidence/benchmark-results.html) (snapshot `2026-10-20-linux-df0d30b`) and the [competitor comparison](https://flahax.readthedocs.io/en/stable/evidence/competitor-comparison.html) (#48) |
+| Known wording defect | The published manual (line 5) says "release candidate prepared 2026-10-21; not yet published" and the sdist changelog says "0.5.0 — Unreleased": the finalize step that 0.4.0 performed before tagging was not done. Code and data are unaffected; corrected on `main` (#71) |
+
+| Artifact | SHA-256 |
+|---|---|
+| `flahax-0.5.0-py3-none-any.whl` | `c209476ae768f79da807068073aacbd55d3ffb6f59024f88e7cff83cb4359141` |
+| `flahax-0.5.0.tar.gz` | `08ca45bc53645a482a57d01a73c6b028f42ce7e7af288b930202d02887e5ab3c` |
+
+Installed resources: `USER_GUIDE.md` `92f5d5806b1f0931df05fe490b738b8b277486d70a21d6ea53e2839f36c5ffb9`;
+`API_REFERENCE.md` `db350573e41e407468d0ddb7bd7b86304c80562baf8d6b77d146d4f8df1ab9c8`;
+`chemistry_25c.json` `ad87363818ccc2471c1343a0791f9ae63798bc0b3530b5b710344aff3077188f`;
+`library.json` `dfc65ac7fc0060ebe760870b832e6c2c4febce2babf8cfa9defa6a858e28a05a`.
+The thermodynamic data are unchanged since 0.3.1.
+
 ## 0.5.0 preparation (2026-10-21)
 
 Prepared revision: `release/0.5.0`, cut from `main` at `9bea54a` (merge of

diff --git a/src/flahax/data/USER_GUIDE.md b/src/flahax/data/USER_GUIDE.md
@@ -2,7 +2,7 @@
 
 Fertilizer formulation • bounded chemistry • reviewed delivery planning
 
-**Version:** 0.5.0, release candidate prepared 2026-10-21; not yet published. See the repository's release evidence.
+**Version:** 0.5.0, released 2026-10-20. The release workflow ran the Python 3.11–3.13 checks and the live PHREEQC replay on the tagged commit; see the repository's release evidence.
 
 ## Contents