re: build the fast probe -- and it refutes the diagnosis that motivated it

Last iteration I blamed four failed runs on the probe sampling every
~41 s, slower than the title screen lasts, and withdrew three earlier
conclusions on that basis. Building the fix tested the claim and killed
it.

The speedup is real and control-verified. One long-lived ffmpeg x11grab
stream, raw RGB, glyph counted in numpy -- no per-sample process startup,
no PNG encode, no convert -crop:

  wrapper `screenshot`            3.98 s per sample (emulator running)
  import -window root -> PPM      1.20 s
  long-lived x11grab stream       0.29 s          13.7x

The counter is byte-identical to is_title.py: 753 on the committed title
capture, 327 on the main menu.

Pointed at a running game it says the opposite of what I expected:

  332 frames in  85.3 s = 3.89 fps; max glyph 0
  1674 frames in 420.0 s = 3.99 fps; max glyph 0

1674 consecutive samples over seven unbroken minutes, four per second,
zero green-(A) pixels. Sampling rate was a real defect that happened not
to be the cause.

So "neither locale reaches the interactive title without a pad press" --
withdrawn last iteration for want of evidence -- is reinstated, now as a
dense measurement, with its reach stated: a MID-RUN window only, silent
about the boot title.

Leading hypothesis, unconfirmed: the PRESS (A) plate appears only in the
boot title window and the attract loop's title carries none, which is
exactly what title_states_capture.sh was written to test. The experiment
is to start the fast probe from t=0 rather than attach to a run already
in progress.

METHOD: fixing the instrument is how you test the explanation that blamed
it -- a plausible mechanism is a hypothesis, and the fix is its
experiment, not its proof.

docs/port/MISSION.md

@@ -120,8 +120,18 @@ afterwards is using canary's own persistence, not patching its code.
 two-grab polling loop therefore samples every **~41 s**, against a title screen
 this corpus documents as lasting *a few seconds*. Four runs — two locales, two
 launch paths, two gamma settings — were all blinking slower than the event.
-Everything below is the trail that led there; the conclusions it draws about the
+🔴 **…and that diagnosis is itself refuted (same day).** The faster probe was
-*game* are withdrawn. See [capture-harness-status](../re/capture-harness-status.md).
+built — 3.98 s → **0.29 s** per sample, 13.7×, control-verified at 753/327 — and
+at 3.99 fps for **420 continuous seconds (1 674 samples)** the interactive title
+*still* never appeared. Sampling rate was a real defect and not the cause.
+✅ So "the interactive title does not appear mid-run without a pad press" is
+reinstated, now as a dense measurement. ⚠️ Its reach is a **mid-run** window; it
+says nothing about the boot title.
+🟡 Leading hypothesis, unconfirmed: the plate appears only in the **boot** title
+window and the attract loop's title has none — which is precisely what
+`title_states_capture.sh` was written to test. The experiment is to start the
+fast probe from t=0, not attach to a run already in progress.
+See [capture-harness-status](../re/capture-harness-status.md).
 
 **🟡 Still not settled — two runs, and the reason moved again.**
 

docs/re/METHOD.md

@@ -510,3 +510,11 @@ agent's loop prompt, i.e. nowhere durable. See [`README.md`](README.md) for the
   classifies movie-vs-static on an *absolute* rmse threshold — so the gamma
   change biased the very classifier the run depended on. Harness thresholds tuned
   on one output configuration are not portable to another.
+* **Fixing the instrument is how you test the explanation that blamed it.** I
+  diagnosed four failed runs as "the probe samples slower than the event",
+  wrote it up, and withdrew three earlier conclusions on that basis. Building the
+  faster probe refuted it: at 100× the sampling density, over 420 unbroken
+  seconds, the event still never occurred. The diagnosis was a real defect that
+  happened not to be the cause — the most seductive kind, because fixing it feels
+  like confirmation. A plausible mechanism that explains a failure is a
+  hypothesis, and the fix is its experiment, not its proof.

docs/re/REFUTED.md

@@ -436,3 +436,13 @@ neighbourhood, not just the line.
   costs 10.8 s while the emulator runs (0.117 s idle, 92×). A title lasting a few
   seconds would be missed. The observations stand; the conclusions drawn from
   them do not. [`capture-harness-status.md`](capture-harness-status.md)
+* "the boot harness fails because its polling loop samples every ~41 s, slower
+  than the title screen lasts" → **mine, and refuted by my own fix.** The
+  sampling defect was real (3.98 s → 0.29 s per sample, 13.7×, control-verified
+  at 753/327), but a probe running at 3.99 fps for **420 continuous seconds —
+  1 674 samples — still saw zero green-Ⓐ pixels.** Sampling rate was not the
+  cause. [`capture-harness-status.md`](capture-harness-status.md)
+* ~~"neither locale reaches the interactive title without a pad press"~~ →
+  withdrawn last iteration for want of evidence, now **reinstated as a
+  measurement**: 1 674 dense samples over 420 s, English, zero glyph frames.
+  ⚠️ Reach: a mid-run window only; it says nothing about the boot title.

docs/re/capture-harness-status.md

@@ -57,7 +57,61 @@ inflates that difference and biases every frame toward "movie". **The capture
 harness's tuning is coupled to the display settings** — changing gamma and
 capture behaviour in one run confounds both.
 
-## The fix, not applied here
+## 🔴 The fix works — and it REFUTES the diagnosis above
+
+Built and measured (`tools/re-capture/fast_title_probe.py`): one long-lived
+`ffmpeg` x11grab stream, raw RGB frames, glyph counted in numpy. No per-sample
+process startup, no PNG encode, no `convert -crop`.
+
+| probe | seconds per sample, emulator running |
+|---|---|
+| the wrapper `screenshot` | **3.98** |
+| `import -window root` → PPM | 1.20 |
+| **long-lived x11grab stream** | **0.29** |
+
+**13.7× faster**, and the counter is control-verified against the committed
+frames — it returns **753** on `live-title-press-a.png` and **327** on
+`live-main-menu.png`, byte-identical to `is_title.py`.
+
+Then it was pointed at a running game:
+
+```
+332 frames in  85.3 s = 3.89 fps; max glyph 0
+1674 frames in 420.0 s = 3.99 fps; max glyph 0
+```
+
+**1 674 consecutive samples over seven unbroken minutes, four per second, and the
+interactive title never appeared.** So the sampling rate was a real defect and
+*not* the cause. The hypothesis on this page — that the harness was blinking
+slower than the event — is **mine, and refuted by my own fix**.
+
+## What that restores
+
+Last iteration I withdrew three conclusions on the strength of that hypothesis.
+The withdrawal was right at the time (15 samples at 41 s intervals cannot support
+them) and is now **superseded by better evidence**: dense sampling says the
+interactive title genuinely does not appear in a mid-run window. Reinstated as a
+**measurement**, with its reach:
+
+* ✅ over **420 continuous seconds**, English, `gamma_type = 2`, ~13 minutes into
+  a run with no pad input, **zero** frames carried the green Ⓐ glyph.
+* ⚠️ Reach: this covers a **mid-run** window only. It says nothing about the
+  first minutes of boot.
+
+## 🟡 The leading hypothesis, not confirmed
+
+The corpus already suspects the answer. `title_states_capture.sh` exists to test
+*"whether the interactive one draws `ptbtn00` (the PRESS Ⓐ plate) and the other
+does not"* — i.e. the title appears **twice**: once at the end of the boot
+sequence, and again from the attract loop, and only the first may carry the
+plate. If so, the plate's window is early and one-shot, and no amount of
+mid-run sampling will ever find it.
+
+That is consistent with everything measured, and it is **not confirmed**. The
+test is to start the fast probe *before* the boot title — from t=0 rather than
+attaching to a run already in progress.
+
+## The original fix note, kept
 
 Make the probe cheap enough to sample faster than the title window: grab a small
 region rather than the full surface, drop the ImageMagick `convert` round trip,

tools/re-capture/fast_title_probe.py

@@ -0,0 +1,34 @@
+"""Fast title probe: one long-lived x11grab stream, ~3.5 fps, glyph count per frame.
+
+Replaces the two-`screenshot` polling loop, which costs ~4-11 s per grab while
+xenia runs and so samples every ~41 s -- slower than the title screen lasts.
+Counter is byte-identical to is_title.py (controlled: 753 on the committed title
+capture, 327 on the main menu).
+"""
+import subprocess, sys, time
+import numpy as np
+W,H = 1280,720
+LIMIT = float(sys.argv[1]) if len(sys.argv)>1 else 120
+NEED  = int(sys.argv[2]) if len(sys.argv)>2 else 400
+p = subprocess.Popen(["ffmpeg","-loglevel","error","-f","x11grab","-draw_mouse","0",
+    "-video_size",f"{W}x{H}","-i",":98","-r","4","-f","rawvideo","-pix_fmt","rgb24","-"],
+    stdout=subprocess.PIPE, bufsize=W*H*3*2)
+n=W*H*3; t0=time.time(); frames=0; best=0; hits=0
+while time.time()-t0 < LIMIT:
+    buf=p.stdout.read(n)
+    if len(buf)<n: break
+    a=np.frombuffer(buf,np.uint8).reshape(H,W,3).astype(int)
+    r,g,b=a[:,:,0],a[:,:,1],a[:,:,2]
+    c=int(((g>130)&(g-r>45)&(g-b>45)).sum())
+    frames+=1
+    if c>best:
+        best=c
+        if c>=NEED:
+            from PIL import Image
+            Image.fromarray(a.astype(np.uint8)).save("/sylph-home/re/shots/fast-title.png")
+    if c>=NEED:
+        hits+=1
+        if hits==1: print(f"TITLE at t={time.time()-t0:.1f}s glyph={c} (frame {frames})", flush=True)
+p.kill()
+dt=time.time()-t0
+print(f"{frames} frames in {dt:.1f}s = {frames/dt:.2f} fps; max glyph {best}; hits {hits}")