Skip to content

chore(release): prepare CableDyn 0.2.0 - #30

Merged
SMI-Lab-Inha merged 4 commits into
devfrom
chore/release-0.2.0
Oct 20, 2026
Merged

SMI-Lab-Inha merged 4 commits into
devfrom
chore/release-0.2.0

Conversation

@SMI-Lab-Inha

Copy link
Copy Markdown
Owner

Summary

Prepares CableDyn 0.2.0 on dev before dev is merged into main.

  • Version 0.2.0, released 2026-10-20, in every file release/check_release_metadata.py names: CMake, the Fortran banner and OpenFAST module, the C API (0.2.0; ABI 1 minor 2 unchanged), the Python package, CITATION.cff, the README and the manual.
  • Short changelog. The [Unreleased] notes (about 850 lines) become a [0.2.0] section of about 90 lines, one entry per user-visible change. The derivations and validation figures stay in the manual and VALIDATION.md. Every figure kept was checked against the full notes.
  • README. Links go to the manual and to GitHub instead of to Markdown files. The README and the manual say that CableDyn is not affiliated with the EPSRC-funded University of Edinburgh project of the same name.
  • OpenFAST default dtM = DT. A deck without dtM now solves every glue step instead of about every 0.1 s.
    • Held loads reach the platform about dtM/2 late, a negative damping of about the mooring stiffness times dtM/2. That matters for taut or tension-leg moorings and for tension statistics.
    • The shipped OpenFAST examples already use dtM = DT.
    • A coarser deck dtM still trades accuracy for speed. The console says so, the OpenFAST page carries a note, and the changelog lists it as a known issue.
  • No compiler warnings. Five solver lines warned:
    • The Hermite node-value cache and the static seabed support compared reals with ==//=. They now use a <= b .AND. a >= b, which has the same meaning (a NaN never matches).
    • The external-load update keeps its remainder in a fixed-size array instead of an allocatable flagged as possibly uninitialised.
    • No result changes.
  • Validation records rerun with the 0.2.0 build, as test_validation_records requires after a version bump.
    • Holcombe: every tracked result is byte-identical.
    • Bergdahl: force matrix and refinement identical. The waveform RMSEs, now computed from the article supplement, move from 5.95 % to 5.93 % and from 2.54 % to 2.56 % (limit 6.5 %).

Release checks (RELEASE.md steps 1-3, local)

  • release/check_release_metadata.py, release/audit_release_tree.py --tree ., pre-commit run --all-files and git diff --check pass.
  • Fresh Release and Debug builds (gfortran, Windows): every test passes, slow tier included. The solver sources build without warnings.
  • Python and the Sphinx build with -W pass.
  • OpenFAST build gate: openfast.exe builds with this tree and prints Running CableDyn (v0.2.0, 2026-10-20).
    • The mixed IEA-15 deck terminates normally with its dtM = 0.025 and without dtM. Without it the module reports 1 x glue DT; the default.
  • validation/scripts/openfast_lin_ab.py: PASS.
    • Per-fairlead stiffness within 0.07-0.11 % of stock MoorDyn (0.29 % transverse), against a 0.5 % gate.
    • Platform 6x6 diagonals within 0.1 %.
    • Symmetry and the sign checks pass.

Checklist

  • pre-commit run --all-files passes
  • ctest --test-dir build --output-on-failure passes
  • python -m pytest tests python/tests passes (if Python or validation records changed)
  • VALIDATION.md updated if a validation result changed
  • CHANGELOG.md updated (release heading [0.2.0])
  • Documentation in doc/ updated for changed inputs, options, or outputs
  • Conventional Commit title (feat:, fix:, docs:, ...)
  • Commits are signed off (git commit -s) under the
    Developer Certificate of Origin

Set the version to 0.2.0 (C API 0.2.0, ABI 1 minor 2 unchanged) and
the release date to 2026-10-20 in every file the release metadata check
names, and move the release notes under a short [0.2.0] heading: one
entry per user-visible change, with the derivations and validation
figures left to the manual and VALIDATION.md.

The README links to the manual and to GitHub instead of to Markdown
files, and the README and the manual state that CableDyn is not
affiliated with the EPSRC-funded University of Edinburgh project of
the same name. The OpenFAST page notes that CableDyn holds its loads
between its own steps, a lag that matters for stiff moorings and
tension statistics, and the changelog lists it as a known issue.

Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
Rerun both experimental comparisons with the 0.2.0 Release build
(GNU Fortran 16.2.0, Windows 11 x86-64, cabledyn.exe sha256
55fe6256...2424a3), following the Reproduce sections of the two
experiment READMEs, and record the release, date, executable and
checksums in each provenance.json.

Holcombe: all 12 mesh and end-condition cases converged. Every committed
result file is reproduced byte for byte (marker RMSE 39.4 / 69.4 mm, peak
marker-curvature error 0.34% / 5.54%, refinement shift 0.007 / 0.12 mm).
Only the hashes of the two primary static profiles changed, together with
the executable, release and date fields.

Bergdahl: all 30 primary and 8 refinement cases completed without a
nonlinear miss. The force-maxima matrix, aggregate metrics, case manifest
and refinement summary are reproduced byte for byte (RMSE 1.58 N, MAPE
3.13%, bias +0.46 N, Pearson 0.993, Lin 0.989; refinement 0.56% and
0.19%).

The two waveform comparisons were previously left at the 2026-10-09
executable because the article supplement's expData.dat was not at hand.
The supplement archive and its expData.dat now match the checksums in the
README, so the waveforms are rerun with this build too: the
phase-averaged RMSEs move from 5.95% to 5.93% (T = 1.25 s) and from 2.54%
to 2.56% (T = 3.50 s) of the measured force range, with correlations of
0.990 and 0.999. The measured phase averages in the waveform CSVs are
unchanged. The provenance waveform note and the matching caveats in
VALIDATION.md and the experiment README are removed.

Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
The release procedure asks for a build with no compiler warnings, and
gfortran warned on five lines of the solver. The node-value cache of
the Hermite cable compared its keys with == on reals, and the static
seabed support of an end element tested a real with /= 0; both now use
a <= b .AND. a >= b, which keeps the meaning of == (a NaN never matches,
+0 matches -0) without the warning. The external-load update of an
EI = 0 line kept the distributed-load remainder in an allocatable that
gfortran flagged as possibly uninitialised; it is now an array of the
size of the external load, restored under the same condition it was
saved under. No result changes.

Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
Without a deck dtM the OpenFAST module solved about every 0.1 s and
held its loads between solves, so the platform received the mooring
load about half a step late, a negative damping of about the mooring
stiffness times dtM/2. That is small for a catenary mooring but not for
a taut or tension-leg one, and it moves tension statistics: in the
DLC-1.1 comparison the fairlead tension's standard deviation differed
by about 3 % from the solution at every glue step. A deck without dtM
now solves every glue step (dtM = DT), as the shipped OpenFAST examples
already do; a deck dtM that is a multiple of DT still trades the lag
for speed, and the console note for it now says so.

The manual (OpenFAST page, modelling guide, OpenFAST tutorial), the
default noted in the example and test decks, and the changelog follow.

Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
@SMI-Lab-Inha

Copy link
Copy Markdown
Owner Author

@codex review

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 20, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-20T09:48:36.297179Z 6132f01 Manual request
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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@read-the-docs-community

Copy link
Copy Markdown

Documentation build overview

📚 CableDyn | 🛠️ Build #35265823 | 📁 Comparing 6132f01 against latest (f27c5e3)

  🔍 Preview build  

33 files changed · ± 33 modified

± Modified

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Already looking forward to the next diff.

Reviewed commit: 6132f018ec

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".

@SMI-Lab-Inha
SMI-Lab-Inha merged commit 391a8ba into dev Oct 20, 2026
15 checks passed
@SMI-Lab-Inha
SMI-Lab-Inha deleted the chore/release-0.2.0 branch October 20, 2026 11:01
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