erik/acdream
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Packages Wiki Activity Actions

feat(render) Campaign FW1: doorway-still conformant - deg_mul is dynamic

Browse source 
Render::deg_mul is auto-tuned by frame load and SWUNG between oracle
captures: doorway-still ran right after the heavy terrace-edge capture
with the multiplier depressed (mul <= 0 puts thresholds at/below ideal,
so 001e/0026/002f select the portless level 1 - retail's zero look-in
floods, reproduced exactly at mul = 0), while every other fixture pins
~ +0.99 (thresholds at max). The recon session's "-0.99" live dump was
real - taken under the same cdb load. RetailFrameWalk now exposes the
multiplier; the doorway test pins 0, the rest use the default.

foundry-entry: frames 1-66 reproduce exactly; the F67-F79 standing
segment diverges ONLY in building 0036's intra-building DC order
(retail 116,118,11d vs replay 11d,116,118). The BSP-traversal microscope
pins the flip to 0036's ROOT plane (N=(0,0,-1) D=2.8): replay eye z
2.33 (d=+0.47, NEG-first) vs retail behaving as d<0 (POS-first) - a
structural ~0.5 m frame question (positionPush(2)/part-scale), only
adjudicable live. Turnkey probe:
tools/walk-oracle/fw1-f67-viewpoint-probe.cdb.

Fixture status: nine of ten fully conformant; foundry-entry exact
through F66 with the 13-frame order segment parked on the probe.
Suites: Walk 127/1 skip; hermetic 6,687/0.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Erik 2026-08-30 11:49:27 +02:00
parent c1a029edf9
commit 77f5342b62
4 changed files with 148 additions and 25 deletions
Download patch file Download diff file Expand all files Collapse all files

28
tools/walk-oracle/fw1-f67-viewpoint-probe.cdb Normal file
View file

@ -0,0 +1,28 @@
* FW1 F67 adjudication probe (2026-08-30): the foundry-entry fixture's
* F67-F79 segment floods building a9b40036's portals in the order
* 116,118,11d while the offline replay orders 11d,116,118. The microscope
* pins the flip to the ROOT plane of 0036's drawing BSP (N=(0,0,-1)
* D=2.8): the replay's building-local eye z is 2.33 (d=+0.47), retail
* behaves as d < 0. This probe dumps the viewpoint retail's walk actually
* uses, adjudicating the ~0.5 m frame question (positionPush(2) /
* part-scale semantics).
*
* USE: stand OUTSIDE the Holtburg foundry (building a9b40036) roughly
* where foundry-entry F67+ was captured: near the entrance, close enough
* that the interior is visible through the door. Attach:
* & "C:\Program Files (x86)\Windows Kits\10\Debuggers\x86\cdb.exe" `
* -pn acclient.exe -cf tools\walk-oracle\fw1-f67-viewpoint-probe.cdb `
* *>&1 | Tee-Object -FilePath fw1-f67-probe.log
* Auto-detaches after 40 hits.
.logopen C:\Users\erikn\source\repos\acdream\fw1-f67-probe.log
.sympath C:\Users\erikn\source\repos\acdream\refs
.symopt+ 0x40
.reload /f acclient.exe
r $t0 = 0
* BSPTREE::build_draw_portals_only — dump the frame's viewpoint (the
* value the plane-side tests read) plus the root node's plane. ecx = the
* BSPTREE. FrameCurrent->viewer.viewpoint: resolve the frameContext
* layout live with `dt acclient!frameContext` first if offsets drift.
bp acclient!BSPTREE::build_draw_portals_only "r $t0 = @$t0 + 1; .printf \"WALK tree=%08x pass=%d\\n\", @ecx, poi(@esp+4); .printf \" viewpoint: \"; dd poi(acclient!Render::FrameCurrent)+0x0 L0; df poi(acclient!Render::FrameCurrent)+0x40 L3; .if (@$t0 >= 40) { qd } .else { gc }"
g