4608d7c416
backlog: the world-unit blocker was the ROUTE, not entities2.py's VA window
Sylpheed RE agent
2026-08-26 20:47:43 +00:00
eb21c977ef
tools: classify the mission briefing instead of filing it as menu
Sylpheed RE agent
2026-08-26 20:47:15 +00:00
64fe8ebdcd
re: entity objects ARE spawned -- the vtables were stale-labelled, the ROUTE was wrong
Sylpheed RE agent
2026-08-26 20:46:21 +00:00
c75c982fa0
re: the working entity captures came from a SAVE SLOT, not mission select
Sylpheed RE agent
2026-08-26 20:13:33 +00:00
f2a18e555c
re: "follow the name pointer" is structurally dead for pak strings
Sylpheed RE agent
2026-08-26 20:06:10 +00:00
3e267ad4c4
re: the mission DOES load -- DEF_VTABLE=0 was a stale constant, not an empty game
Sylpheed RE agent
2026-08-26 19:56:52 +00:00
4ee7971340
re: the post-take-off state is measurably neither a menu nor flight
Sylpheed RE agent
2026-08-26 19:43:40 +00:00
9031c2510a
re: the take-off sequence lands in the launch CUTSCENE, not a hang
Sylpheed RE agent
2026-08-26 19:30:21 +00:00
926cee6ba0
re: the sign-in fix holds end-to-end, but no units are loaded yet
Sylpheed RE agent
2026-08-26 19:20:24 +00:00
0274b29368
re: the sign-in fix must be applied at launch, not by the navigator
Sylpheed RE agent
2026-08-26 19:07:25 +00:00
c09107fa8d
re: correct the "screen id" -- they are monotonic counters, not identities
Sylpheed RE agent
2026-08-26 18:51:25 +00:00
93bc7bf061
re: SOLVED -- the mission freeze is a modal sign-in dialog
Sylpheed RE agent
2026-08-26 18:41:09 +00:00
78696338b6
re: WITHDRAW "the freeze requires the rasterizer" -- the control was not controlled
Sylpheed RE agent
2026-08-26 18:28:47 +00:00
c791b3c374
re: find the menu screen id and cursor in guest memory
Sylpheed RE agent
2026-08-26 18:21:04 +00:00
da17e05c68
re: the freeze requires the rasterizer -- --gpu=null does not freeze
Sylpheed RE agent
2026-08-26 18:12:35 +00:00
8bc471224a
re: WITHDRAW the infinite-loop conclusion -- gdb was the confound
Sylpheed RE agent
2026-08-26 18:02:54 +00:00
a1b6add50d
re: prove the guest PC -- an infinite copy loop in sub_82457780
Sylpheed RE agent
2026-08-26 17:50:57 +00:00
e5dc3e1e4d
re: audit the gdb session -- withdraw the sub_82457780 localisation
Sylpheed RE agent
2026-08-26 17:38:49 +00:00
9ea0a00d80
re: locate the stuck thread -- a grow-and-copy in sub_82457780
Sylpheed RE agent
2026-08-26 17:28:31 +00:00
36eb8c226c
re: the allocation failure is definitively not the freeze -- four runs, three outcomes
Sylpheed RE agent
2026-08-26 17:15:03 +00:00
bd639bb344
re: the 171 MB of boot allocations are the game's own pools, ~90% empty
Sylpheed RE agent
2026-08-26 16:59:26 +00:00
e71efb0845
re: MISSION SELECT is not special -- NEW GAME freezes too
Sylpheed RE agent
2026-08-26 16:51:59 +00:00
59fe3d87fe
re: it is not a stage-load freeze -- the game freezes entering MISSION SELECT
Sylpheed RE agent
2026-08-26 16:42:13 +00:00
0241ebf5b6
re: the stage-load freeze is not a regression -- an 08-17 binary reproduces it
Sylpheed RE agent
2026-08-26 16:19:57 +00:00
9350ced6ac
re: the freeze is not a kernel-object wait -- log_stuck_waits stays silent
Sylpheed RE agent
2026-08-26 16:10:07 +00:00
88ca60aa0e
re: the allocation failure is NOT the freeze -- two patched builds say so
Sylpheed RE agent
2026-08-26 15:22:32 +00:00
7c19e33a12
re: the heap ledger refutes my own leak hypothesis
Sylpheed RE agent
2026-08-26 12:22:40 +00:00
9be2eb5a93
re: the leaking release is confirmed in Xenia's source, not inferred
Sylpheed RE agent
2026-08-26 11:32:34 +00:00
66bc2462c2
re: the freeze's 512 MB is the emulated console's, not the host's
Sylpheed RE agent
2026-08-26 11:29:51 +00:00
480b699cd5
re: the freeze is independent of the mask poke -- control run settles it
Sylpheed RE agent
2026-08-26 11:26:44 +00:00
28cf52b297
re: the mission freeze is a failed 128 MB guest allocation, not an emulator hang
Sylpheed RE agent
2026-08-26 11:14:24 +00:00
b641034be7
re: three boot-path failures that look like the emulator and are not
Sylpheed RE agent
2026-08-26 11:02:15 +00:00
de1ff6ed76
docs: define the status markers, and fix six mislabelled or superseded entries
Sylpheed RE agent
2026-08-26 10:29:15 +00:00
ce0720a3e9
re: correct a mislabelled blocker -- the world unit needs a run, not a waiver
Sylpheed RE agent
2026-08-26 10:07:01 +00:00
35f67a2a37
re: retract the paint-key census and redo it over all 21 184 sprites
Sylpheed RE agent
2026-08-26 10:06:23 +00:00
e5dde6f4db
re: flag that the paint-key census may have sampled the wrong sprite population
Sylpheed RE agent
2026-08-26 09:58:32 +00:00
e0394647b2
re: census the T8aD paint-order key -- it is pak-local, not a global vocabulary
Sylpheed RE agent
2026-08-26 09:50:29 +00:00
83d019515f
re: confirm the cmesh<->model link via GameResourceID; world unit is oracle-blocked
Sylpheed RE agent
2026-08-26 09:41:46 +00:00
cc1f9e6734
re: decode CollisionSet_*.bin -- the per-object collision-mesh library
Sylpheed RE agent
2026-08-26 09:31:22 +00:00
e30881189b
re: record the MiscBin name resolution in the backlog
Sylpheed RE agent
2026-08-26 09:21:15 +00:00
3916f30ce2
re: name-resolve all 40 MiscBin entries, and demonstrate the REGN<->MCOL pairing
Sylpheed RE agent
2026-08-26 09:21:00 +00:00
e1db4d84ff
re: MCOL solved -- a closed triangle collision mesh in a uniform grid
Sylpheed RE agent
2026-08-26 09:12:36 +00:00
cda5a31601
re: MCOL's 0x5C block is bounding spheres at stride 16 -- the 0.75 was 12/16
Sylpheed RE agent
2026-08-26 09:04:30 +00:00
324389c612
re: MCOL's u16 entries do not index the point block -- the bound-check was a false positive
Sylpheed RE agent
2026-08-26 08:52:44 +00:00
ee3188bd4c
re: MCOL array B decoded -- {count, u16 index array}, and the chain closes
Sylpheed RE agent
2026-08-26 08:45:58 +00:00
fe06b26751
re: MCOL A->B is 1:1 with count 1, matching REGN's cell-index convention
Sylpheed RE agent
2026-08-26 08:38:46 +00:00
e684eb45dd
re: MCOL array A confirmed as the per-cell record -- 100% once split
Sylpheed RE agent
2026-08-26 08:32:41 +00:00
d1c89e5d79
re: MCOL's 32-byte record is a cell entry, and there are two interleaved arrays
Sylpheed RE agent
2026-08-26 08:31:46 +00:00
a2aa0a4cf8
re: MCOL's pointer layout, read from POF0
Sylpheed RE agent
2026-08-26 08:25:29 +00:00
af834b7ff4
re: open MCOL -- same container as REGN, same map parameters
Sylpheed RE agent
2026-08-26 08:18:17 +00:00
da499901b5
re: REGN is excluded as the wave scheduler -- it is a navmesh
Sylpheed RE agent
2026-08-26 08:11:36 +00:00
6f81ef93cf
re: independently confirm the REGN face decode -- 133573/133573 against a 0.69% control
Sylpheed RE agent
2026-08-26 08:06:30 +00:00
c54924d08f
re: backlog said REGN's coupling was unreachable statically -- it has been reached
Sylpheed RE agent
2026-08-26 08:05:31 +00:00
214ae8fca2
re: resolved -- the REGN base is chunk+0x10 and my section-0 offsets were 16 bytes early
Sylpheed RE agent
2026-08-26 08:00:18 +00:00
553208754e
re: withdraw "everything was read 16 bytes early" — the planes were right
Sylpheed RE agent
2026-08-26 07:58:45 +00:00
84358e61eb
re: record an open conflict with auto/regn-reader over the +0x10 base
Sylpheed RE agent
2026-08-26 07:55:12 +00:00
b9643c63f6
re: REGN is a tetrahedral navigation mesh — the cell→geometry link, decoded
Sylpheed RE agent
2026-08-26 07:53:13 +00:00
0d51cdb403
re: regn_decode.py — read a REGN object through its own POF0 fixup table
Sylpheed RE agent
2026-08-26 07:53:13 +00:00
67cce0e4c0
re: correct my last two savegame sections -- they analysed the COMPRESSED stream
Sylpheed RE agent
2026-08-26 07:50:29 +00:00
91074d1ba7
re: all savegame variation is in the header; slot region identical -- and my heuristic overstates
Sylpheed RE agent
2026-08-26 07:44:55 +00:00
2aa5894212
re: the save embeds live guest pointers, and a three-way diff of the captures is invalid
Sylpheed RE agent
2026-08-26 07:39:13 +00:00
7c1507a466
re: REGN cell payload indexes neither points nor planes -- static search exhausted
Sylpheed RE agent
2026-08-26 07:32:08 +00:00
1006523767
re: the REGN cell index does not reference section 1 -- two powered tests
Sylpheed RE agent
2026-08-26 07:26:45 +00:00
a20c5d56f9
re: refute my own BVH reading of REGN section 1
Sylpheed RE agent
2026-08-26 07:20:39 +00:00
488e8599b4
re: REGN section 1 slot census, and an index test with no power
Sylpheed RE agent
2026-08-26 07:14:39 +00:00
eb278c7941
re: REGN sections 0/1/2 have strides, and section 2 is a plane list
Sylpheed RE agent
2026-08-26 07:07:10 +00:00
fd71f37d8f
re: RATC +0x10 is a u16 pair too, and no bit is a clean predicate
Sylpheed RE agent
2026-08-26 06:59:08 +00:00
a191569167
re: the RATC +0x0c interval is authored, not a keyframe extent
Sylpheed RE agent
2026-08-26 06:52:44 +00:00
1fc68e391d
re: RATC +0x0c is two u16s forming an ordered interval, not one number
Sylpheed RE agent
2026-08-26 06:46:50 +00:00
0800e40c80
re: lead the .slb page with its settled state, not its history
Sylpheed RE agent
2026-08-26 06:39:40 +00:00
2634632787
re: refute the block-id chain -- 8 of 281 is coincidence
Sylpheed RE agent
2026-08-26 06:34:59 +00:00
f189f5fc1f
re: the scattered block bytes are an XACT-shaped cue record naming a real cue
Sylpheed RE agent
2026-08-26 06:29:06 +00:00
effda99390
re: w0 and w1 are the last two entries of the leading wave's seek table
Sylpheed RE agent
2026-08-26 06:22:56 +00:00
640d860b23
re: the seek table's sample total is not the wave length -- durations corrected
Sylpheed RE agent
2026-08-26 06:17:07 +00:00
00e9a932ea
re: name two fields inside the 12288-byte block's trailer
Sylpheed RE agent
2026-08-26 06:09:46 +00:00
85548bf661
re: the 12288 region is a padded metadata block, and the assignment rule is 100%
Sylpheed RE agent
2026-08-26 06:04:40 +00:00
4e180b3e1c
re: enumerate all 9661 waves, and measure a bank-to-wave assignment rule
Sylpheed RE agent
2026-08-26 05:58:29 +00:00
532e7dac74
re: refuse a seek-based recovery of headerless bank durations
Sylpheed RE agent
2026-08-26 05:52:02 +00:00
0aed00d4f6
re: stage dialogue with per-line audio durations, in both languages
Sylpheed RE agent
2026-08-26 05:46:03 +00:00
856fd64b3b
re: full bank manifest -- 4114 banks, 390.9 minutes, and the durations self-validate
Sylpheed RE agent
2026-08-26 05:39:16 +00:00
16a19219e8
slb: guard the wave-boundary identity with a test, and expose data_at
Sylpheed RE agent
2026-08-26 05:33:01 +00:00
e676210e9b
re: close both remaining .slb questions in the backlog -- both were my errors
Sylpheed RE agent
2026-08-26 05:30:57 +00:00
c3ff6aa0e8
slb: read Channels instead of assuming mono -- and retract the TCAF_608 conclusion
Sylpheed RE agent
2026-08-26 05:27:42 +00:00
b98910cab3
re: fix the slots-vs-lines wording at its actual location
Sylpheed RE agent
2026-08-26 05:23:35 +00:00
0be28c5fe2
re: sweep all six language packs -- and Japanese is not missing captions
Sylpheed RE agent
2026-08-26 05:23:06 +00:00
4b54ac2bf0
re: the seek chunk's layout is readable, its packet count is not
Sylpheed RE agent
2026-08-26 05:16:12 +00:00
931deb42d8
re: withdraw my own generalisation -- decoded length cannot locate a bank boundary
Sylpheed RE agent
2026-08-26 05:10:58 +00:00
e368c7064c
re: ran the decode test -- the bank data does not continue past the TOC window
Sylpheed RE agent
2026-08-26 05:04:50 +00:00
a3c6a6502c
re: the four .slb offsets are a segment-packing phase, not a header size
Sylpheed RE agent
2026-08-26 05:03:56 +00:00
2aa7f9a5a5
re: the .slb data offset is the .pNN segment's 2048 grid phase
Claude
2026-08-26 05:01:26 +00:00
746ca56a74
tools: slb_segment_phase.py — derive the .slb data offset exactly
Claude
2026-08-26 05:01:26 +00:00
c2e7f8f8e4
re: the flat entity count is explained -- arrivals are not spawns
Sylpheed RE agent
2026-08-26 04:59:32 +00:00
f32bfb1433
re: withdraw the DisableInterval / delayed-route association -- it is a clustering artefact
Sylpheed RE agent
2026-08-26 04:53:13 +00:00
ea155ea035
re: the wave timetable's t is SECONDS, measured against the live speed law
Sylpheed RE agent
2026-08-26 04:47:10 +00:00
71d294ece1
re: narrow what the .slb header is -- four things it is not
Sylpheed RE agent
2026-08-26 04:37:59 +00:00
220e250453
slb: the seek chunk gives the data offset structurally, and breaks the 28 ties
Sylpheed RE agent
2026-08-26 04:32:23 +00:00
f47d013c94
re: the unvoiced script lines are not the resupply case -- 0 of 14 have a cutscene twin
Sylpheed RE agent
2026-08-26 04:24:37 +00:00
ce726eaa9c
re: the cue-less cutscene pages ARE the resupply lines, and they cross-check the movie work
Sylpheed RE agent
2026-08-26 04:23:52 +00:00
e3c5eeae58
re: check whether the scan's 99.62% transfers to the banks it is actually used on
Sylpheed RE agent
2026-08-26 04:18:53 +00:00
b7b29dcb05
slb: the declared data size is an upper bound, not an exact one
Sylpheed RE agent
2026-08-26 04:12:45 +00:00
Sylpheed RE agent