From 3633c91a9b073392ef6a158d88f658131c4aed90 Mon Sep 17 00:00:00 2001 From: Valentin Popov Date: Sat, 18 Jul 2026 09:59:27 +0400 Subject: [PATCH] docs(render): record part preview probes --- docs/tomes/05-render.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/docs/tomes/05-render.md b/docs/tomes/05-render.md index cc65092..e7b122c 100644 --- a/docs/tomes/05-render.md +++ b/docs/tomes/05-render.md @@ -955,6 +955,17 @@ synchronized readback (FNV-1a `12332214764918703197`) and validation warnings/errors `0/0`. This is not a full-scene or original-renderer pixel-parity claim. +The same bounded command was then run against the licensed installed Part 1 +(`C:\\Program Files (x86)\\Nikita\\IS`) and Part 2 +(`C:\\Program Files (x86)\\Nikita\\IS2`) corpora supplied for testing. Both +`MISSIONS/Autodemo.00/data.tma` runs completed with the same one-material, +1280×720 two-image report, the same `12332214764918703197` readback hash and +validation warnings/errors `0/0`; Part 1 took 28.5 seconds and Part 2 took +94.5 seconds. Part 2's synchronous checkpoint was `Graph` while it was still +loading, so the longer startup is not attributed to Vulkan. This is only a +cross-corpus confirmation of the first-root static-preview bridge, not a claim +that the two games' full mission renderers are compatible. + The preview now asks `load_mission_static_preview` for both graph and assets. Normal mission loading remains full and transactional; preview graph traversal and asset preparation are restricted to the first TMA root. It is therefore not