docs: the summary goes day by day, from the commit record

The weeks were wrong — grouped from memory of the rounds rather than
from what actually landed when, which put the repository split, the soak
tier and the CI work in the wrong buckets.  Days are checked against
`git log`: 20 active days between 2026-07-22 and 2026-08-12, all 20
listed, none invented (25–26 July and 1–2 August weekends have no
commits and no entries).

Each day is one point with its headline changes nested under it and what
they buy: 22 July stood the whole stack up, 27 July took visible-behaviour
parity in one sitting, 4 August made v4 the package, 10 August put the
expensive algorithms on the GPU, 11 August took the load path apart.

Rendered: three tables, 95 points, zero prose blocks.

Claude-Session: https://claude.ai/code/session_01CZuBgFBGSDy6QtAYFL5sE3

Author: maxkfranz

EXECUTIVE_SUMMARY.md

@@ -14,15 +14,14 @@ The v4 rewrite: a columnar model and a WebGPU renderer, per
 - **Rewrite when a round closes** — by re-reading the record and restating it,
   never by appending.
 - **Point form only.** Short points, high level, readable in five minutes.
-- **Week by week, and headline changes only.** Each week is a list of what
-  changed and **what it buys** — the benefit, not the implementation. A change
-  earns a line only if it changed what the library *is* or what it can do; most
-  do not.
+- **Day by day, headline changes only.** Each day is a list of what changed and
+  **what it buys** — the benefit, not the implementation. A change earns a line
+  only if it changed what the library *is* or what it can do; most do not.
 - **No round narrative.** Rounds, file names, sub-round numbering and
   implementation detail belong in `PLAN.md`. Round numbers appear here only
   where one is the sole handle on an open question.
-- **Restate, don't append.** Later rounds routinely change what an earlier
-  decision meant, so earlier weeks need correcting too.
+- **Restate, don't append.** Later work routinely changes what an earlier
+  decision meant, so earlier days need correcting too.
 - **Facts, not judgement.** The numbers table below stays current; assessments
   of how close a release is, what is "left", or whether the work is on track do
   not belong here — they need information this file does not carry.
@@ -49,74 +48,110 @@ The v4 rewrite: a columnar model and a WebGPU renderer, per
 
 ---
 
-## Week 1 — 22–24 July
-
-- **A columnar, CPU-canonical model** — typed-array columns, stable slots,
-  coalesced dirty spans, a CSR adjacency index. Buys the whole performance
-  case: the graph is already in the shape the GPU wants.
-- **Reads stay synchronous.** Buys a public API that did not have to become
-  async to get the model.
-- **A WebGPU renderer** — SDF shapes, compute culling, indirect draws, GPU
-  picking, GPU text. Buys interaction on graphs that stall v3: **33 ms frames
-  where v3 takes 4,460**.
-- **A co-signed model↔renderer contract in one file.** Buys a layout change
-  that cannot happen by accident on one side only.
-- **Structured queries instead of a selector language.** Buys queries that are
-  data — composable, inspectable, no parser.
-- **A serializable mapper DSL for style.** Buys paint channels evaluated on the
-  GPU, and style that can cross a wire.
-
-## Week 2 — 27 July – 2 August
-
-- **Visible-behaviour parity with v3** — curves, compounds, labels, images,
-  animation. Buys an existing app the same picture.
-- **Goldens *plus* live v3-vs-v4 parity diffs.** Buys the answer to *is it
-  right*, which a golden alone cannot give: goldens compare v4 against its own
-  past.
-- **Benchmarks with an HTML report.** Buys claims that are numbers.
-- **Design sittings, decisions recorded when taken.** Buys a migration guide
-  that could later be compiled rather than reconstructed.
-
-## Week 3 — 3–8 August
-
-- **Error and documentation contracts gated at zero tolerance.** Buys a guard
-  no test has ever fired failing the build, and a public member without docs
-  failing it too.
-- **The repository split: v4 is the package, v3 lives whole in `v3/`.** Buys a
-  v4 that ships alone, and a real v3 to keep measuring parity against.
-- **Packaging gate, migration guide, soak tier, status site.** Buys release
-  mechanics that are testable before there is a release.
-- **CI made honest.** Buys a green run that means something: every defect found
-  was in something never executed on a fresh checkout, on a runner, or in a
-  browser nobody could launch locally.
-
-## Week 4 — 9–11 August
-
-- **Both long-standing API decisions taken, each by declining the surface its
-  own homework had priced.** Buys a smaller library and a recorded reason.
-- **The force layout rebuilt on published methods.** Buys a layout that stays
-  numerically stable past node degree ~20, and lays out the 465k-edge network
-  live in **1.3 s**.
-- **The nine expensive whole-graph algorithms went async with GPU executors.**
-  Buys Markov clustering at up to **642×** its CPU reference, k-medoids 146×,
-  Floyd–Warshall 28×; the traversal tier stays synchronous.
-- **Per-element bypasses returned as a first-class stylesheet section.** Buys
-  v3's ergonomics without v3's cost — id-keyed, surviving re-add, carried by
-  `cy.json()` (v3 loses them), twice as fast.
-- **Cross-commit benchmark comparison pages.** Buys regressions caught by the
-  archive rather than by a user: one four-day-old regression on first use.
-- **The load path measured, then taken apart.** Buys **1.9×** on the 465k-edge
-  load with a byte-identical rendered frame, and initialisation of that network
-  in **~0.95 s against v3's ~19 s**.
-
-## Week 5 — 12 August
-
-- **The benchmark runner runs jobs concurrently.** Buys a publishing run of the
-  `all` profile in **8.4 minutes instead of ~55**, so re-measuring stops being
-  a reason not to measure.
-- **Concurrency folded into the harness fingerprint, and the published archive
-  kept serial.** Buys the guarantee that a faster *instrument* can never be
-  read as a faster *library*.
+## Day by day
+
+- **22 Jul** — the whole stack, stood up in one day
+  - Columnar store, model↔renderer contract, core facade, collections, events,
+    viewport, compiled style engine, grid layout, WebGPU pipeline, GPU picking,
+    pointer interaction, SDF labels, debug harness.
+  - Buys the architecture proved end to end before any of it was polished.
+- **23 Jul** — the model earns its shape
+  - Packed numeric keys for set operations, columnar flag scans, id-index
+    selector resolution, slot-native traversal.
+  - Buys v3's own API measuring faster on v4, not just its drawing.
+- **24 Jul** — style becomes data
+  - Mapper spec compile, scale programs, OKLab colour and named schemes,
+    ordinal dictionaries, GPU evaluation of scalar and colour channels.
+  - Buys paint that never round-trips through the CPU, and style that can cross
+    a wire.
+- **27 Jul** — visible-behaviour parity, in one sitting
+  - Search, path, structure and centrality algorithms; polygon shapes; line
+    styles; label visuals; edge labels; one easing layer; the gesture family
+    (cxttap, taphold, dbltap, drag-all-selected); mount/unmount; device-loss
+    recovery.
+  - Buys an existing app the same picture and the same interactions.
+- **28 Jul** — measurement infrastructure
+  - A single-page benchmark report, and browser renderer benchmarks putting v3
+    canvas against v4 WebGPU.
+  - Buys claims that are numbers.
+- **29 Jul** — memory comes back
+  - String-dictionary and id-blob compaction, CSR adjacency rebuild.
+  - Buys a long-running session that does not keep the graph it removed.
+- **30 Jul** — curved edges
+  - Curve parameters, route pipeline, arrows on end tangents, labels at route
+    midpoints, exact accessors, haystack and manual endpoints.
+  - Buys v3's whole edge vocabulary.
+- **31 Jul** — compound graphs
+  - Hierarchy model, lifecycle, auto-bounds, ancestor-gated visibility, event
+    bubbling, parent draw/cull/pick, compound layouts and loop edges; end
+    labels on v3's 3×3 grid.
+  - Buys nested graphs, which most real applications use.
+- **1 Aug** — the layout leaves the main thread
+  - The force layout on the extension contract with a GPU integrator; touch
+    gestures; the display/visibility split; style transitions; slot compaction.
+  - Buys layout that does not freeze the page, and a hide that costs only
+    paint.
+- **2 Aug** — the visual vocabulary completed
+  - Every remaining v3 node shape and arrowhead, v3's nonlinear arrow sizing,
+    size/width/font/padding tweens, shipped TypeScript declarations.
+  - Buys a stylesheet that ports without a lookup table of what is missing.
+- **3 Aug** — contracts gated, hot paths fixed
+  - Documentation and error contracts gated at zero tolerance; the style read
+    path's 150-case switch became a dispatch table; memoized collections;
+    O(1) `indexOf`.
+  - Buys a build that fails on an undocumented member or an untested guard, and
+    a worst-case style read 2.6× faster.
+- **4 Aug** — v4 becomes the package
+  - The repository split so v4 is the package and v3 lives whole in `v3/`; v4's
+    own event emitter; packaging gates; the docs generator; the debug harness
+    rebuilt.
+  - Buys a v4 that ships alone, and a real v3 to keep measuring against.
+- **5 Aug** — robustness and publication
+  - The soak tier (leaks, churn, wire fuzzing, multi-instance isolation); the
+    status site; machine provenance and a tracked benchmark archive; fixtures
+    shipped in v4's binary wire format.
+  - Buys a deployable preview of the branch, and a benchmark history that
+    survives the machine that produced it.
+- **6 Aug** — CI made honest
+  - Fresh-checkout failures fixed, bundles built before the suite; v3's arrow
+    gap and spacing constants ported; feature pipelines built on first draw.
+  - Buys a green run that means something, and a first frame of 2.7 s where it
+    had been 4.6.
+- **7 Aug** — parity found by measuring, not by looking
+  - Numeric routing parity (geometry, not pixels); the arrow gap; element state
+    (`:selected`, `:active`, `:grabbed`) as ordinary style condition rather than
+    a shader constant; v3's selection colours.
+  - Buys differences 400× below the pixel bound being caught, and every
+    affordance being restyleable or disableable by an application.
+- **8 Aug** — the things you click
+  - v3's hit halos, arrowheads as hit targets, `:active` on edge press;
+    border-style and outline-style on every shape; minified WGSL; directional
+    fit bounds.
+  - Buys edges that are clickable again, and smaller shipped shaders.
+- **9 Aug** — every benchmark pair reads v4-faster
+  - State flips write the diff rather than the element; the force layout made
+    stable by construction; cross-commit benchmark comparison pages.
+  - Buys performance visible across commits — the pages caught a four-day-old
+    regression on first use.
+- **10 Aug** — the expensive algorithms go async, onto the GPU
+  - Nine whole-graph algorithms return promises and take
+    `executor: 'cpu' | 'gpu' | 'auto'`, with WGSL kernels and a performance
+    pass; per-element bypasses returned as a stylesheet section;
+    `cy.collection()` throws on an argument, `cy.$()` and `cy.byId()` return.
+  - Buys Markov clustering at up to **642×** its CPU reference, and v3's bypass
+    ergonomics at twice v3's speed.
+- **11 Aug** — the load path taken apart
+  - Definition-form payloads convert to columns before ingest; mapped style
+    costs per distinct value; a run of identically styled edges is written once;
+    the benchmark comparison learned harness epochs, repeat medians and
+    screened movers.
+  - Buys **1.9×** on a 465k-edge load with a byte-identical frame, and a
+    comparison page that can tell a change from its own noise.
+- **12 Aug** — the instrument
+  - The benchmark runner runs jobs concurrently; concurrency is folded into the
+    harness fingerprint and the published archive stays serial.
+  - Buys a publishing run in **8.4 minutes instead of ~55**, with no way to
+    mistake a faster instrument for a faster library.
 
 ---
 
@@ -134,9 +169,6 @@ The v4 rewrite: a columnar model and a WebGPU renderer, per
   `executor: 'cpu' | 'gpu' | 'auto'`.
 - **`cy.collection()` throws if passed an argument**; **`cy.$()` and
   `cy.byId()`** restored as aliases.
-- **Element state** (`:selected`, `:active`, `:grabbed`, …) is ordinary style
-  condition rather than a shader constant, so an application can restyle or
-  disable every affordance.
 
 ## Open decisions