specmod 0.5.0__tar.gz → 0.7.0__tar.gz

{specmod-0.5.0 → specmod-0.7.0}/.github/workflows/build.yml

@@ -18,6 +18,8 @@ jobs:
       # fails every job at "Getting action download info", before a step runs.
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
+        with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
       - run: uv python install 3.11
 
       - name: Build sdist and wheel

{specmod-0.5.0 → specmod-0.7.0}/.github/workflows/docs.yml

@@ -32,6 +32,8 @@ jobs:
       # fails every job at "Getting action download info", before a step runs.
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
+        with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
       - run: uv python install 3.11
       - run: uv venv # create a venv to install into
 

specmod-0.7.0/.github/workflows/pr-title.yml

@@ -0,0 +1,30 @@
+name: pr-title
+
+# A squash merge replaces the branch's commit subjects with this title, so the
+# title is what release-please parses for the version bump and the changelog.
+# `edited` is in the list because a title is most often fixed after the check
+# has already failed once.
+on:
+  pull_request:
+    types: [opened, edited, reopened]
+
+# `pull_request`, not `pull_request_target`: this reads a title an outside
+# contributor controls, and must never do so with a write token.
+permissions: {}
+
+concurrency:
+  group: pr-title-${{ github.event.pull_request.number }}
+  cancel-in-progress: true
+
+jobs:
+  pr-title:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v7
+      - name: The title must be a Conventional Commit
+        # Through the environment rather than as an argument: the title is
+        # attacker-chosen text, and `${{ }}` interpolation into a shell command
+        # would execute it.
+        env:
+          PR_TITLE: ${{ github.event.pull_request.title }}
+        run: python tools/check_pr_title.py

{specmod-0.5.0 → specmod-0.7.0}/.github/workflows/release.yml

@@ -90,6 +90,8 @@ jobs:
       # fails every job at "Getting action download info", before a step runs.
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
+        with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
       - run: uv python install 3.11
       - run: uv build
 

{specmod-0.5.0 → specmod-0.7.0}/.github/workflows/test.yml

@@ -21,6 +21,7 @@ jobs:
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
         with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
           enable-cache: true
       - run: uv python install 3.11
       - run: uv venv
@@ -41,6 +42,7 @@ jobs:
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
         with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
           enable-cache: true
       - run: uv python install 3.11
       - run: uv venv
@@ -65,6 +67,7 @@ jobs:
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
         with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
           enable-cache: true
       - run: uv python install ${{ matrix.python-version }}
       - run: uv venv
@@ -93,6 +96,8 @@ jobs:
       # fails every job at "Getting action download info", before a step runs.
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
+        with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
       - run: uv python install 3.11
       - run: uv venv
       # Exercise the declared minimums. CI otherwise installs the newest of
@@ -117,6 +122,7 @@ jobs:
       # Bump this deliberately; do not "tidy" it to `@v10`.
       - uses: astral-sh/setup-uv@v10.0.1
         with:
+          version: "0.13.7"  # see UV_VERSION in tests/test_release_config.py
           enable-cache: true
       - run: uv python install 3.11
       - run: uv venv

specmod-0.7.0/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+  ".": "0.7.0"
+}

{specmod-0.5.0 → specmod-0.7.0}/AGENTS.md

@@ -41,6 +41,13 @@ to generate `CHANGELOG.md`. See
 - Say *why*, with the measurement if there was one. The history is the record
   of what was checked; a message that only restates the diff wastes it.
 
+**The pull request title is one of them.** A squash merge discards every
+subject on the branch and uses the title instead, composed by GitHub after all
+local hooks have run, so the `commit-msg` hook never sees it. `pr-title.yml`
+checks it against the same types `release-please-config.json` declares. Merging
+with a merge commit keeps the individual subjects and is the safer default; the
+check is what makes a squash safe when one is used anyway.
+
 ## Workflow files need a permission you may not have
 
 `.github/workflows/` is editable directly, but only when the session's GitHub

{specmod-0.5.0 → specmod-0.7.0}/CHANGELOG.md

@@ -9,6 +9,51 @@ release could have this problem — every later one has a real predecessor tag
 so leave subsequent headings as generated. release-please only prepends to this
 file, so this edit survives. -->
 
+## [0.7.0](https://github.com/sgjholt/SpecMod/compare/v0.6.0...v0.7.0) (2026-11-08)
+
+
+### ⚠ BREAKING CHANGES
+
+* **datasets:** event directories, waveform files and pick files are named in ISO 8601 basic format (`20190826T073047.000000Z`) instead of extended format (`2019-08-26T07:30:47.000000Z`), and `specmod fetch` writes the new names. Run `specmod migrate-names <dir>` on data fetched before 0.7; `Event.directory()` reads the old names with a DeprecationWarning until 0.8. Code that built a path from `Event.origin` or `SpectrumSet.event` should call `specmod.datasets.path_name` on it. See docs/upgrading.md, From 0.6 to 0.7.
+
+### Features
+
+* **datasets:** name times on disk without colons, and migrate existing data ([52168de](https://github.com/sgjholt/SpecMod/commit/52168dea5cd9ffd8ff16dd9dd70996824c0bbb93))
+
+
+### Code Refactoring
+
+* **config:** narrow the default_factory sentinel instead of suppressing mypy ([961da50](https://github.com/sgjholt/SpecMod/commit/961da50c28aa8ba785879a2b2287c219f64ec216))
+
+
+### Documentation
+
+* **roadmap:** file the default change and the config pin as shipped in v0.6.0 ([92eff1a](https://github.com/sgjholt/SpecMod/commit/92eff1af91dceea5ee65da5a55b460c026cb75de))
+
+## [0.6.0](https://github.com/sgjholt/SpecMod/compare/v0.5.0...v0.6.0) (2026-09-27)
+
+<!-- One Documentation line below was removed by hand, the third edit in this
+file. release-please listed the roadmap entry twice: once for 9f35727, the
+commit, and again for ed4e10c, the merge commit of #48, whose body repeated the
+commit's subject and so parsed as a second commit. The entry kept points at the
+commit itself. release-please only prepends to this file, so the edit
+survives. -->
+
+### ⚠ BREAKING CHANGES
+
+* **config:** default to FFT with Konno-Ohmachi smoothing
+
+### Features
+
+* **config:** default to FFT with Konno-Ohmachi smoothing ([1081364](https://github.com/sgjholt/SpecMod/commit/108136484a55ea5cc6c3490ced0fc8cdaf0052b5))
+* **config:** hold a configuration across an ambient read, and pin the goldens ([71a4e9c](https://github.com/sgjholt/SpecMod/commit/71a4e9c68c60c7ded332b6e2cdecd8da7c0b7a99))
+
+
+### Documentation
+
+* **fitting:** say what the fit reads, and document it as a modelling choice ([33975ed](https://github.com/sgjholt/SpecMod/commit/33975ed1042e7374e7750c46aa879e666cf0296a))
+* **roadmap:** file the station selection as shipped in v0.5.0 ([9f35727](https://github.com/sgjholt/SpecMod/commit/9f357276a6e6a7fd4694cd5107405b4d53f12f44))
+
 ## [0.5.0](https://github.com/sgjholt/SpecMod/compare/v0.4.0...v0.5.0) (2026-09-19)
 
 <!-- The `acquire` entries below are edited by hand, and are the second entry

{specmod-0.5.0 → specmod-0.7.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: specmod
-Version: 0.5.0
+Version: 0.7.0
 Summary: A Python toolbox for processing and modelling seismic spectra
 Project-URL: Homepage, https://github.com/sgjholt/SpecMod
 Project-URL: Repository, https://github.com/sgjholt/SpecMod

{specmod-0.5.0 → specmod-0.7.0}/docs/REFACTOR_PLAN.md

@@ -3477,11 +3477,15 @@ bumping `smoothing.n_bins` from 151 to 158 and running the suite — and **9 of
 25 golden tests fail**, loudly. The golden values live in a committed JSON
 file, so any behaviour change breaks them whatever moved it.
 
-The residual gap is narrower than the claim suggests but real: the golden file
-records no config, so a *regenerated* reference silently adopts whatever
-defaults were current. Pinning a study file, as this plan proposes, is what
-would close it. Until then the protection is "the numbers are frozen", not
-"the settings are frozen".
+**Closed in 0.6.** `specmod.config.using()` holds a configuration across the
+ambient reads the pipeline makes — a `ContextVar`, so it does not leak between
+threads, and explicit arguments still win — and `tests/golden/reference.toml`
+carries every setting the committed numbers were captured under, not an
+overlay that would inherit future default changes for whatever it omitted.
+`tools/make_golden.py` captures inside the pin and the golden suite compares
+inside it, so the property the claim asserted now holds: changing a shipped
+default leaves every committed number untouched. Measured by doing it — the
+0.6 change of default moved 11 golden tests before the pin and none after.
 
 **Claims that were checked and hold:**
 
@@ -3717,33 +3721,57 @@ changes anything shipped.
 
 ### Still open
 
-1. **Default estimator.** Multitaper (matching current behaviour) or FFT +
-   Konno–Ohmachi (more conventional in engineering seismology)? **No longer
-   blocked** — §6.6's dead `[smoothing] method` is wired up, so the second
-   option is selectable — and now measured, on the 28 PNR windows with each
-   station fitted freely:
-
-   | Against the shipped multitaper + `log_bins` | Median ΔMw | Per-station range | `fc` ratio |
-   |---|---|---|---|
-   | `fft` + `log_bins` | −0.039 | −0.85 to +1.90 | 0.003 to 5.0 |
-   | `fft` + `konno_ohmachi` | +0.027 | −0.52 to +0.51 | 0.038 to 3.2 |
-
-   **The smoothing is what makes the switch defensible, not the estimator.**
-   FFT with the shipped binner moves a station by up to 1.9 magnitude units,
-   because `log_bins` reduces the axis without reducing the variance — 151 log
-   bins over a few hundred Fourier samples is roughly one sample per bin.
-   Pairing it with a real smoother halves the worst excursion and pulls the
-   `fc` range in by a third. The median station barely moves either way.
-
-   What this event cannot settle: the *event* corner is unconstrained under all
-   three (stage-one spread 895% to 1042% of the event value), so the ensemble
-   number is not evidence. Deciding on the ensemble needs an event where it is
-   constrained.
+1. ~~**Default estimator.**~~ **Resolved: FFT + Konno-Ohmachi**, from 0.6. The
+   measurement below is what decided it, on the 28 PNR channels, each fitted
+   independently (stage one) on velocity, which is the motion the pipeline fits
+   by default and from which `llpsp` is the displacement plateau:
+
+   | Against the shipped multitaper + `log_bins` | Median ΔMw | Per-station ΔMw | `fc` ratio, median (range) | Median band (decades) | Median points fitted |
+   |---|---|---|---|---|---|
+   | `multitaper` + `log_bins` | — | — | — | 1.73 | 113 |
+   | `fft` + `log_bins` | −0.079 | −0.71 to +0.23 | 0.91 (0.18 to 3.9) | 1.55 | 107 |
+   | `fft` + `konno_ohmachi` | −0.068 | −0.25 to +0.19 | 0.82 (0.17 to 3.5) | 1.68 | 120 |
+   | `fft` + `konno_ohmachi`, `fit_bins = true` | −0.073 | −0.25 to +0.15 | 0.90 (0.18 to 2.2) | 1.68 | 120 |
+
+   **What each setting reaches.** The estimator sets the spectrum that is
+   fitted. The smoother sets the band: the signal-to-noise ratio and the gate
+   are computed on smoothed spectra. With `[fitting] fit_bins = false`, the
+   default in 0.1.1 and since, the fit and the initial guess both read the
+   *unsmoothed* amplitudes inside that band, so the smoother reaches `Omega`
+   only through the band. The two default-path FFT rows therefore fit identical
+   amplitudes, and the whole difference between them — the per-station range
+   narrowing from −0.71/+0.23 to −0.25/+0.19 — comes from Konno–Ohmachi
+   selecting a wider band. Fitting the smoothed spectrum instead
+   (`fit_bins = true`) changes little further on this event.
+
+   Measured on displacement instead, the same comparison gives −0.85/+0.70 for
+   `fft` + `log_bins` and −0.56/+0.52 for `fft` + `konno_ohmachi`: the same
+   ordering, a wider spread.
+
+   What this event cannot settle, and what the decision therefore does not
+   rest on: the *event* corner is unconstrained under all three (stage-one
+   spread 895% to 1042% of the event value), so the ensemble number is not
+   evidence. On the ensemble, PNR's Mw moves −0.1615 under both constant sets
+   — a pure Omega shift, since the constants are an offset applied afterwards.
+   Against the catalogue's 2.9 the two sets move in opposite directions, the
+   default constants from 0.194 to 0.032 away and §4.7's from 0.160 to 0.321,
+   so they no longer both sit within 0.25. Recorded, not offered as evidence.
 
    Speed, measured rather than assumed and far too small to decide on: the
    transform is 0.209 s against 0.323 s for the 28 windows, and the smoothing
    adds 0.14 s (`log_bins`) to 0.83 s (`konno_ohmachi`) on top.
 
+   **What had to be built first.** Changing a default moved 11 golden tests,
+   because §6.6's config pin was a claim with no mechanism: the references were
+   captured under whatever the defaults were. `config.using()` and
+   `tests/golden/reference.toml` close that, so the references now hold their
+   own settings and a default change leaves every committed number untouched.
+
+   **The fitting choice is documented where it is made.** Whether the fit
+   reads the unsmoothed or the smoothed spectrum is a modelling decision a user
+   should make knowingly, so the tutorial's *What the fit reads* states the
+   default and fits the event both ways.
+
 2. **Python floor.** 3.11 is proposed. Any users stuck on 3.9/3.10?
 3. **History rewrite.** Deleting the 9.9 MB catalog and stripping the notebook
    (§5.1) shrinks the *working tree* but leaves both in history, so a fresh clone

{specmod-0.5.0 → specmod-0.7.0}/docs/_builders/specmod_tutorial.py

@@ -91,13 +91,15 @@ from pandas.core.series import Series
 import specmod.preprocess as pre
 import specmod.utils as ut
 from specmod.core.collection import SpectrumPair, SpectrumSet
+from specmod.datasets import path_name
 from specmod.pipeline import spectrum_set_from_streams
 from specmod.staged import StagedFit
 
 # Relative to this notebook, so it runs from wherever it is opened.
 EVENTS = Path("data/events")
 EVENT = "2019-08-26T07:30:47.000000Z" # Event origin time
-DATA: Path = EVENTS / EVENT
+# Named without colons, which Windows does not allow: 20190826T073047.000000Z.
+DATA: Path = EVENTS / path_name(EVENT)
 WAVEFORMS: Path = DATA / "waveforms"
 STATIONS: Path = DATA / "stations" / "*.xml"
 PICKS: Path = DATA / "picks"
@@ -237,12 +239,14 @@ spectrum_set_from_streams
 ```
 
 `spectrum_set_from_streams` transforms both windows, puts the noise on the
-signal's frequency axis, bins both, raises the noise to account for what sits
-*under* the signal, takes the ratio and selects the band where it passes.
+signal's frequency axis, smooths both, raises the noise to account for what
+sits *under* the signal, takes the ratio and selects the band where it passes.
+The smoothing is Konno–Ohmachi by default (`[smoothing] method`). It is used
+for the ratio and the band, not for the fit — see *What the fit reads* below.
 
 Which estimator is used is configuration, not code: `fft`, `welch`,
 `multitaper`, `quadratic` and `cwt` all satisfy the same Parseval contract, so
-they are interchangeable here. The shipped default is `multitaper`.
+they are interchangeable here. The shipped default is `fft`.
 """)
 
 
@@ -256,7 +260,7 @@ print(
 
 
 code("""
-# One station in detail: signal, noise, the binned spectra the ratio is
+# One station in detail: signal, noise, the smoothed spectra the ratio is
 # actually computed on, the selected band (red) and the resolution floor (grey).
 from specmod.plotting import plot_pair, plot_set
 
@@ -453,9 +457,10 @@ displacement = spectra.to_motion("displacement")
 print(f"velocity     {spectra[station].signal.unit}")
 print(f"displacement {displacement[station].signal.unit}")
 
-# The unbinned signal-to-noise ratio is invariant under this — both spectra are
-# divided by the same 2*pi*f — but the *binned* ratio is not, because a bin
-# holds a geometric mean. So a few bands do move.
+# The unsmoothed signal-to-noise ratio is invariant under this — both spectra
+# are divided by the same 2*pi*f — but the smoothed ratio is not: Konno-Ohmachi
+# averages amplitudes across a window over which 2*pi*f changes, so some
+# bands move.
 moved = [i for i in spectra.ids() if spectra[i].band != displacement[i].band]
 print(f"bands that moved: {len(moved)} of {len(spectra)}")
 """)
@@ -473,9 +478,10 @@ from the largest amplitude inside the band, the corner from where that
 maximum falls.
 
 `FitSpectra(spectra)` then `fit_spectra()` is the whole thing. The minimiser
-(Powell), the `t*` lower bound and whether to fit the binned or unbinned
+(Powell), the `t*` lower bound and whether to fit the smoothed or unsmoothed
 spectrum all come from `[fitting]` in the configuration, so the defaults are
-recorded rather than remembered.
+recorded rather than remembered. The last of those is a modelling choice with
+consequences; *What the fit reads* below compares the two.
 """)
 
 
@@ -551,6 +557,62 @@ fits.table[["id", "fc", "fc-stderr", "llpsp", "ts", "pass_fitting"]].head(10)
 """)
 
 
+md("""
+### What the fit reads
+
+The smoothed spectra choose the band — the signal-to-noise ratio and the gate
+are computed on them — but **the fit reads the unsmoothed spectrum inside that
+band**. Smoothing decides *where* the model is fitted, not *what* it is fitted
+to. That is `[fitting] fit_bins = false`, the default, and it is a modelling
+choice:
+
+- **Unsmoothed** (the default). Every Fourier sample in the band, each with its
+  full scatter. The samples are evenly spaced in Hz, so most of them sit in the
+  top decade of the band and dominate an unweighted misfit;
+  `[fitting] weight_method = "log"` weights residuals by `1/f` to shift that.
+- **Smoothed** (`fit_bins = true`). The spectrum the band was chosen on. Under
+  Konno–Ohmachi that is the same frequencies with the scatter averaged down,
+  but neighbouring points share samples, so they are not independent and the
+  misfit counts the same information more than once. Under `log_bins` it is
+  one value per occupied bin, evenly spaced in log frequency.
+
+The cell below fits the event both ways.
+""")
+
+
+code("""
+smoothed_fits = FitSpectra(spectra, fit_bins=True)
+smoothed_fits.fit_spectra()
+
+first = spectra.ids()[0]
+raw_fit, smoothed_fit = fits.models[first], smoothed_fits.models[first]
+low, high = raw_fit.mod_freq[0], raw_fit.mod_freq[-1]
+top_decade = np.mean(raw_fit.mod_freq >= high / 10)
+print(
+    f"{first}: {raw_fit.mod_freq.size} points unsmoothed, "
+    f"{smoothed_fit.mod_freq.size} smoothed, over {low:.2f}-{high:.2f} Hz"
+)
+print(f"  share of the points in the band's top decade: {top_decade:.0%}")
+print(
+    "  median |difference| in the log10 amplitudes the two fits see: "
+    f"{np.median(np.abs(raw_fit.mod_amp - smoothed_fit.mod_amp)):.3f}"
+)
+
+raw_table = fits.table.set_index("id")
+smoothed_table = smoothed_fits.table.set_index("id")
+d_plateau = (smoothed_table["llpsp"] - raw_table["llpsp"]).abs()
+fc_ratio = smoothed_table["fc"] / raw_table["fc"]
+print(f"\\nacross {len(raw_table)} channels, smoothed against unsmoothed:")
+print(
+    f"  |d log10 plateau|  median {d_plateau.median():.3f}, max {d_plateau.max():.3f}"
+)
+print(
+    f"  fc ratio           median {fc_ratio.median():.2f}, "
+    f"range {fc_ratio.min():.2f} to {fc_ratio.max():.2f}"
+)
+""")
+
+
 md("""
 ### The fit is not unique, and that is not a detail
 
@@ -944,7 +1006,7 @@ it, because journal supplements want one.
 code("""
 from specmod.io import load, save
 
-path: Path = save(path=OUTPUT / spectra.event, spectra=spectra)
+path: Path = save(path=OUTPUT / path_name(spectra.event), spectra=spectra)
 print(f"{path}  ({path.stat().st_size // 1024} KB)")
 
 back: SpectrumSet = load(path)
@@ -956,8 +1018,8 @@ print(
 
 
 code("""
-FitSpectra.write_flatfile(path=OUTPUT / "flatfiles" / f"{spectra.event}.parquet", fits=fits)
-FitSpectra.write_flatfile(path=OUTPUT / "flatfiles" / f"{spectra.event}.csv", fits=fits)
+FitSpectra.write_flatfile(path=OUTPUT / "flatfiles" / f"{path_name(spectra.event)}.parquet", fits=fits)
+FitSpectra.write_flatfile(path=OUTPUT / "flatfiles" / f"{path_name(spectra.event)}.csv", fits=fits)
 sorted(p.name for p in (OUTPUT / "flatfiles").iterdir())
 """)
 

{specmod-0.5.0 → specmod-0.7.0}/docs/releasing-data.md

@@ -30,6 +30,13 @@ station wildcards against the data centre, cuts each channel to the window
 relative to origin, and writes an `EventDirectory` with a `manifest.json`
 beside it.
 
+The event directory and each waveform file carry the origin in ISO 8601 basic
+format — `20200318T130931.000000Z`, not `2020-03-18T13:09:31.000000Z` — because
+the colons are not legal in a Windows file name, and a source distribution or
+archive containing one is refused by uv 0.13. A fetch made before 0.7 has the
+old names: run `specmod migrate-names build/magna_2020` before archiving it,
+which renames them and updates the manifest to match.
+
 Expect it to take a few minutes — Magna is 88 stations across seven networks.
 
 **Check the manifest before going further.** It records what came back, not

{specmod-0.5.0 → specmod-0.7.0}/docs/roadmap.md

@@ -171,9 +171,7 @@ Fourier grid is uniform in Hz. Weighting samples equally tilts the result by
 axis brings that to 3.2e-8 dex. The latter is the default. Konno–Ohmachi
 applies the former, by definition, and keeps it.
 
-## In progress — not yet released
-
-Merged on `main`, and it will name its version when a release goes out.
+## Shipped in v0.5.0
 
 ### The station selection means what the config says
 
@@ -193,6 +191,56 @@ found by a live fetch rather than by the suite:
   as two independent records of one ground motion. `channel_priorities` and
   `location_priorities` rank them, first match wins per station.
 
+## Shipped in v0.6.0
+
+### The default is FFT with Konno-Ohmachi smoothing
+
+The shipped pair was multitaper with log-binning. Measured on the 28 PNR
+channels, each fitted independently on velocity, against that pair:
+
+| | Median ΔMw | Per-station ΔMw | `fc` ratio, median (range) |
+|---|---|---|---|
+| `fft` + `log_bins` | −0.079 | −0.71 to +0.23 | 0.91 (0.18 to 3.9) |
+| `fft` + `konno_ohmachi` | −0.068 | −0.25 to +0.19 | 0.82 (0.17 to 3.5) |
+
+By default the fit reads the unsmoothed spectrum inside the band the smoothed
+signal-to-noise ratio selects, so the two FFT rows fit identical amplitudes
+and differ only in the band. Konno–Ohmachi's is wider, and that is what
+narrows the per-station spread. This event's corner is unconstrained under
+every pair, so the event-level number is not evidence either way.
+
+**Every result moves** unless a configuration already pinned both settings.
+[Upgrading](upgrading.md#from-05-to-06) has what to pin to keep the old
+numbers, and what `[fitting] fit_bins = true` changes: it fits the smoothed
+spectrum instead of the unsmoothed one.
+
+### A configuration can be held across an ambient read
+
+`config.using()` pins one configuration for everything inside the block,
+including the many pipeline functions that read their settings from call sites
+taking no configuration argument. That was the prerequisite for the change
+above: the golden references were captured under whatever the defaults were,
+so moving a default moved 11 committed numbers, and a deliberate change was
+indistinguishable from a regression. They now carry their own settings in
+`tests/golden/reference.toml`, and a default change leaves every one of them
+untouched.
+
+## In progress — not yet released
+
+Merged on `main`, and it will name its version when a release goes out.
+
+### Names on disk carry no colons
+
+Event directories and the files in them were named by the origin as ObsPy
+prints it, `2019-08-26T07:30:47.000000Z`. Windows does not allow the colons,
+and uv 0.13 refuses to build a source distribution containing them, which
+stopped the release. Names are now ISO 8601 basic format,
+`20190826T073047.000000Z`, from `specmod.datasets.path_name`, and
+`specmod fetch` writes them. `specmod migrate-names` renames data written
+earlier and keeps its manifest verifying; until 0.8, data that has not been
+migrated is still found, with a warning.
+[Upgrading](upgrading.md#from-06-to-07) has the steps.
+
 ## Planned
 
 ### 1.0 — the API stops moving
@@ -223,6 +271,27 @@ the same code path. Nothing breaks at import or at the call; results move. That
 is a weaker promise than the one 1.0 makes and a real one to have broken, so
 the count starts again rather than standing at one.
 
+**v0.5.0 is not clean either, in the same way.** Nothing was renamed and no
+signature moved. But `max_radius_km` now cuts against true WGS84 distance
+rather than a fixed kilometres-per-degree, so the same config can select a
+different set of stations than it did before, and three radius values that used
+to be accepted — zero, negative, and a minimum outside the maximum — are now
+refused when the config is read. A config is as much a caller as a call site.
+
+Twice in a row makes it worth naming: what keeps moving is not the API, it is
+what the API *does* with a setting that was already there. A deprecation cycle
+does not protect against that, since there is nothing to deprecate; the
+[upgrading guide](upgrading.md) is the only thing that does, and 1.0 will not
+change that.
+
+**v0.6.0 is not clean either, and is the first to break on purpose.** Its
+default estimator and smoother changed, so every result moves unless a
+configuration pinned both; the change is marked breaking, and
+[Upgrading](upgrading.md#from-05-to-06) says what to pin. One signature moved
+as well: the first argument of `load_config` is positional-only, so
+`load_config(start=path)` is now read as a `[start]` configuration section and
+refused. The count stands at zero.
+
 ## After 1.0
 
 Designed but deliberately not on the path to 1.0, because each is blocked on