A corridor may be corridor-shaped: no aspect cap on circulation

Owner's ruling: "there should be no cap on the proportion of a corridor,
especially for big buildings, the crinkliness rule is there to prevent these
becoming unpleasant spaces."

proportion_circulation was [1.5, 0.5], failing a corridor above aspect 2.57 --
at the 1.97 m minimum width the width factor allows, a corridor 5.1 m long; at
the 2.4 m target width, 6.2 m. The corpus shows the consequence: median
circulation leaf 14.3 m2 at aspect 1.67, a stubby room rather than a corridor.

The justification checks out arithmetically. The unpleasant space the cap was
standing in for is a long BURIED corridor -- and crinkliness already sends that
to zero, since a buried leaf has crink == 0. A long corridor along a facade
scores 0.90. Aspect cannot tell those two apart; crinkliness can, so the cap was
duplicating a rule that already exists and does the job better.

Shipped: proportion_circulation = None (no aspect requirement).
quality_proportion returns 1.0 for circulation, and shapecurve.leaf_constraints
yields rmax = inf so the DP agrees with the fitness instead of pruning
topologies the objective would accept. A habitable room's aspect target is
untouched. The narrow side still holds -- width_circulation keeps a corridor
>= 1.97 m, and "edge too long" still caps a single wall at 8 m.

Unlike 39.14/39.18/39.19 this DOES change the fail set, which is the point.
Across the twelve baseline artefacts it removes exactly 7 corridor proportion
fails and adds none: harbor 33/43/42 -> 33/42/40, maple 54/73/55 -> 54/71/54,
health-centre 4/9/5 -> 3/9/5, programme-house unchanged.

Also recorded in 39.22, and a retraction: hxi was titled "search is rewarded
for deleting the circulation spine", which 39.8 had already measured and
refuted -- 0 of the 7 connectivity-breaking deletions sampled were rewarded. My
own earlier comment on the bead restated that retracted claim; corrected, and
the bead retitled.

What binds next: removing the cap roughly doubles a corridor leaf's reach, from
proportion at ~6.2 m to size_circulation at 12.5 m (2.4 m wide hits the 30 m2
fail edge there). size_circulation's target area is ZERO, the other half of the
double-charge -- circulation priced as overhead once in value_circulation = 50
and again in a size factor whose optimum is non-existence. Not changed: it is a
distinct parameter with its own rationale, unruled, and 39.16 is a standing
reminder about inherited constants.

415 passed.

Refs homemaker-py-hxi.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MJ84Feep79Hhm3E4zZJmnB

DESIGN.md

@@ -7814,3 +7814,88 @@ Next: `homemaker-py-hxi` — circulation at 0.07 return against a room's 0.66,
 which the owner has ruled needs fixing — and the corpus re-baseline
 (`homemaker-py-bk9`), which §39.19 made necessary and which every future number
 depends on.
+
+### 39.22 A corridor may be corridor-shaped (`homemaker-py-hxi`)
+
+Asked whether the scoring of outdoor and circulation space is likely correct,
+the answer for circulation is no — but **not** for the reason the bead carried.
+
+**First, a retraction of the bead's own premise.** `hxi` was titled "search is
+rewarded for deleting the circulation spine". §39.8 already measured that and
+found the opposite: of 64 sampled deletions across harbor and maple, 7 broke
+connectivity and **0 of those 7 were rewarded** (×0.00 to ×0.58). Severing costs
+one or two connectivity fails plus the cascade behind them, which outweighs the
+×6 value gain. The low value rate does not translate into the spine being
+deleted. (A comment on the bead restating the retracted claim has been
+corrected, and the bead retitled.)
+
+**What the parameters actually say.** Over the 111 circulation leaves in the
+twelve baseline runs, per leaf:
+
+| factor | mean | fails |
+|---|---|---|
+| perpendicular | 0.982 | 0/111 |
+| proportion | 0.710 | 7/111 |
+| size | 0.539 | 16/111 |
+| width | 0.772 | 2/111 |
+| crinkliness | 0.485 | 29/111 |
+| access | 1.000 | 0/111 |
+
+Median corridor 14.3 m², median aspect **1.67** — a stubby room, not a
+corridor. Two parameters explain that, and both are the double-charge shape
+§39.14 and §39.18 found elsewhere:
+
+1. **`proportion_circulation = [1.5, 0.5]`** caps aspect at 2.57. At the 1.97 m
+   minimum width the width factor allows, that is a corridor **5.1 m long**; at
+   the 2.4 m target width, 6.2 m. A corridor is long and thin by nature, and
+   the bigger the building the longer the spine must be.
+2. **`size_circulation = [0.0, 14.0]`** — the target area of a corridor is
+   **zero**, so a median 14.3 m² corridor scores 0.592 on size purely for
+   existing. Circulation is then priced as overhead twice: once in
+   `value_circulation` = 50 (a sixth of a room) and again in a size factor
+   whose optimum is non-existence.
+
+**Owner's ruling on the first:** *there should be no cap on the proportion of a
+corridor, especially for big buildings, the crinkliness rule is there to
+prevent these becoming unpleasant spaces.*
+
+That justification is exactly right, and it is checkable. The unpleasant space
+the aspect cap was being used to prevent is a long *buried* corridor — and
+crinkliness already sends that to zero, because a buried leaf has `crink == 0`.
+A long corridor *along a facade* is a perfectly good corridor, and crinkliness
+scores it **0.90**. The cap was standing in for a rule that already exists and
+does the job better, since it distinguishes the two cases where aspect cannot.
+
+**Shipped:** `proportion_circulation = None`, meaning no aspect requirement.
+`quality_proportion` returns 1.0 for circulation; `shapecurve.leaf_constraints`
+yields `rmax = inf` so the DP agrees with the fitness rather than pruning
+topologies the objective would accept. A habitable room's aspect target is
+untouched — the ruling is about corridors.
+
+The narrow side is still held: `width_circulation` keeps a corridor ≥ 1.97 m
+wide, and `edge too long` still caps any single wall at 8 m.
+
+**Unlike §39.14/§39.18/§39.19, this changes the fail set** — that is the point
+of it. Across the twelve baseline artefacts it removes exactly **7** corridor
+proportion fails and adds none:
+
+| programme | fails before → after |
+|---|---|
+| harbor-house | 33/43/42 → 33/42/40 |
+| maple-court | 54/73/55 → 54/71/54 |
+| health-centre | 4/9/5 → 3/9/5 |
+| programme-house | 1/1/1 → unchanged |
+
+**What binds next, and it is worth knowing before reading too much into this.**
+Removing the aspect cap roughly doubles how long a single corridor leaf may be:
+the binding constraint moves from proportion at ~6.2 m to `size_circulation` at
+**12.5 m** (a 2.4 m wide leaf reaches the 30 m² fail edge there). A longer spine
+is still representable as several corridor leaves in series, so this is a real
+relaxation rather than a cosmetic one — but the zero-area size target is still
+there, and it is the other half of the double-charge above. It has **not** been
+changed: it is a distinct parameter with its own rationale (circulation is
+overhead, minimise it), the owner has not ruled on it, and §39.16 is a standing
+reminder about how often an inherited constant has turned out to be right.
+
+Whether any of this helps the search is unmeasured and, as with everything since
+§39.19, gated on the re-baseline (`homemaker-py-bk9`).

src/homemaker_layout/fitness.py

@@ -233,7 +233,17 @@ CONF_DEFAULTS: dict = {
     "size_circulation": [0.0, 14.0],
     "size_inside": [16.0, 3.5],
     "proportion_outside": [1.5, 50],
-    "proportion_circulation": [1.5, 0.5],
+    # homemaker-py-hxi (DESIGN.md §39.22). Was [1.5, 0.5], which fails a
+    # corridor above aspect 2.57 -- at the 1.97 m minimum width the width factor
+    # allows, that is a corridor 5.1 m long. A corridor is long and thin by
+    # nature, and the longer the building the longer the spine has to be.
+    #
+    # `None` means NO proportion requirement. What stops a corridor becoming an
+    # unpleasant space is crinkliness, not aspect: a long buried corridor has
+    # crink == 0 and fails outright, while a long one along a facade scores
+    # 0.90 -- which is the right answer for both. Width still holds the
+    # narrow side (>= 1.97 m) and `edge too long` still caps a single wall.
+    "proportion_circulation": None,
     "proportion_inside": [1.5, 0.5],
     "width_outside": [3.0, 0.3],
     "width_circulation": [2.4, 0.2],
@@ -1182,6 +1192,8 @@ class Fitness:
             params = self.conf("proportion_outside")
         elif t0 == "c":
             params = self.conf("proportion_circulation")
+            if params is None:
+                return 1.0          # no aspect requirement -- §39.22
         else:
             params = self.get_space_params(leaf.type, "proportion")
             co_type = self._leaf_co_type(leaf)

src/homemaker_layout/shapecurve.py

@@ -233,6 +233,9 @@ def leaf_constraints(fit, leaf: dom_mod.Node) -> LeafBounds:
         params = fit.conf("proportion_circulation")
     else:
         params = fit.get_space_params(leaf.type, "proportion")
+    if params is None:
+        rmax = math.inf         # no aspect requirement (§39.22, circulation)
+    else:
         target, sigma = params[0], params[1]
         rmax = max(1.0 + 1e-9, target + _K * sigma)
 

tests/test_circulation_proportion.py

@@ -0,0 +1,96 @@
+"""Corridors have no aspect requirement (homemaker-py-hxi, DESIGN.md §39.22).
+
+Owner's ruling: "there should be no cap on the proportion of a corridor,
+especially for big buildings, the crinkliness rule is there to prevent these
+becoming unpleasant spaces."
+
+The tests below pin both halves of that — the cap is gone, AND crinkliness
+still does the job the cap was wrongly doing, so removing it did not leave
+long buried corridors unpunished.
+"""
+
+from __future__ import annotations
+
+import copy
+import math
+from pathlib import Path
+
+import pytest
+
+from homemaker_layout import dom as dom_mod
+from homemaker_layout.fitness import (
+    CONF_DEFAULTS, FAIL_THRESHOLD, Fitness, gaussian, load_config,
+)
+
+EXAMPLES = Path(__file__).resolve().parent.parent / "examples"
+pytestmark = pytest.mark.skipif(not (EXAMPLES / "harbor-house").is_dir(),
+                                reason="examples absent")
+
+
+def _fit(**ov):
+    conf, cost = load_config(EXAMPLES / "harbor-house", overrides=ov)
+    return Fitness(conf, cost)
+
+
+def test_there_is_no_corridor_aspect_requirement():
+    assert CONF_DEFAULTS["proportion_circulation"] is None
+    fit = _fit()
+    assert dom_mod.is_circulation(dom_mod.Node(type="C"))
+    # any aspect at all scores 1.0 -- checked through the real code path
+    for aspect in (1.0, 3.0, 12.0, 40.0):
+        fit_leaf = dom_mod.Node(
+            type="C", node=[[0.0, 0.0], [aspect, 0.0], [aspect, 1.0], [0.0, 1.0]])
+        assert fit.quality_proportion(fit_leaf) == 1.0, aspect
+
+
+def test_a_room_still_has_one():
+    """The ruling is about corridors. A habitable room's aspect target stands."""
+    fit = _fit()
+    room = dom_mod.Node(type="r", node=[[0.0, 0.0], [12.0, 0.0], [12.0, 1.0], [0.0, 1.0]])
+    assert fit.quality_proportion(room) < FAIL_THRESHOLD
+
+
+def test_crinkliness_is_what_punishes_an_unpleasant_corridor():
+    """The ruling's own justification, as arithmetic.
+
+    A long BURIED corridor is exactly the unpleasant space the aspect cap was
+    being used to prevent; crinkliness sends it to zero. A long corridor along
+    a facade is a perfectly good corridor, and crinkliness likes it.
+    """
+    target, sigma = CONF_DEFAULTS["uncrinkliness_circulation"]
+    h, width, length = 3.0, 2.4, 30.0
+    area = width * length
+
+    buried = 0.0                       # no illuminated wall at all
+    assert buried == 0.0
+
+    lit_one_side = (length * h) / area  # the long wall is a facade
+    q = gaussian(1.0 / lit_one_side, 1.0, target, sigma)
+    assert q > 0.85, "a daylit corridor along a facade should score well"
+
+
+def test_removing_the_cap_only_ever_removes_proportion_fails():
+    """It cannot introduce a failure, and it touches nothing but corridors."""
+    old = {"proportion_circulation": [1.5, 0.5]}
+    seen = 0
+    for name in ("harbor-house", "maple-court", "health-centre", "programme-house"):
+        d = EXAMPLES / name
+        for p in sorted(d.glob("coldstart-500000-s*.dom")):
+            root = dom_mod.load(str(p))
+            c_old, cost = load_config(d, overrides=old)
+            c_new, _ = load_config(d)
+            _, f_old = Fitness(c_old, cost).score_with_fails(copy.deepcopy(root))
+            _, f_new = Fitness(c_new, cost).score_with_fails(copy.deepcopy(root))
+            assert not (set(f_new) - set(f_old)), "must not add a failure"
+            assert all(f.endswith(" proportion") for f in set(f_old) - set(f_new))
+            seen += 1
+    assert seen >= 4
+
+
+def test_the_shape_curve_dp_accepts_an_unbounded_aspect():
+    """`leaf_constraints` feeds rmax into the DP; None must become inf, not crash
+    and not silently fall back to a finite cap."""
+    from homemaker_layout import shapecurve
+    fit = _fit()
+    leaf = dom_mod.Node(type="C", node=[[0.0, 0.0], [4.0, 0.0], [4.0, 4.0], [0.0, 4.0]])
+    assert math.isinf(shapecurve.leaf_constraints(fit, leaf).rmax)