Files

T

CydandClaude Fable 5 63312e07f9 source410: literal 4.10 source reconstruction + BC++ 4.52 fleet toolchain archived

- BORLAND/: Borland C++ 4.52 (chosen over 4.5 by byte-match: CODE/RP/CW32.LIB
  is identical to 4.52's install lib). BCC32/TLINK32/TLIB/MAKE run natively on
  Win11; CODE/BT/OPT.MAK is the shipped BTL4OPT.EXE's exact flag recipe
  (extender = Borland PowerPack DPMI32, not Phar Lap TNT).
- restoration/source410/: the literal 1995-form reconstruction of the missing
  BT game source (never mixed into CODE/). Round 1-3 state:
  * 6 of 10 surviving original TUs COMPILE CLEAN under the period toolchain
    (BTMSSN, BTCNSL, BTSCNRL, BTTEAM, BTL4MODE, BTL4ARND) - first builds
    since 1996.
  * BT_L4/BTL4APP.CPP pilot reconstruction: 12/12 functions, Fail() lands on
    its binary-recorded line 400 exactly.
  * BT/BTCNSL.HPP: console wire IDs recovered from the binary's ctors
    (Killed=9, Damaged=10, ScoreUpdate=13, DeathWithoutHonor=15 [T1];
    TeamScore=12 flagged [T4]).
  * MUNGA/: 8 engine-header backfills back-dated from the BT412 WinTesla tree
    (VDATA numbering decomp-verified; AUDREND's OpenAL-era virtual removed -
    the period compiler is the drift detector).
  * Tooling: backdate.py (WinTesla->1995 header transform), compile410.sh
    (per-TU verification sweep under authentic OPT.MAK flags).
  * README: corrected roadmap - MECH.HPP is the capstone grown with the mech
    TU reconstructions; BTREG.CPP green = the header-family milestone.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

2026-07-19 07:33:26 -05:00

4.3 KiB

Raw Blame History

BTL4APP.CPP — reconstruction notes (2026-07-18)

Status: 12/12 functions reconstructed; the binary's Fail() line constraint (BTL4APP.CPP:400) is met exactly. Source basis: decomp BT411/reference/decomp/all/part_014.c:8052-8425, port semantic map BT411/game/reconstructed/btl4app.cpp, surviving headers BTL4APP.HPP / L4APP.HPP / BTL4MODE.HPP / BTL4ARND.HPP / BTREG.HPP / L4CTRL.HPP / CONTROLS.HPP / MODE.HPP / APPMSG.HPP / RPL4VID.HPP / RENDERER.HPP / LATTICE.HPP, idiom exemplars AUDSRC.CPP / BTL4ARND.CPP / L4MPPR.CPP / BTMSSN.CPP.

Address map (source order = binary emission order) [T1]

line	function	@addr	size
100	ctor	4d34c4	72
116	dtor	4d350c	44
125	MakeRegistry	4d3538	44
138	MakeVideoRenderer	4d3564	76
165	MakeAudioRenderer	4d35b0	100
189	MakeModeManager	4d3614	48
202	MakeControlsManager	4d3644	72
228	MakeGaugeRenderer	4d368c	56
249	MakeMission	4d36c4	48
264	MakeViewpointEntity	4d36f4	824 (anchor; Fail @ line 400 ✅)
461	RunMissionMessageHandler	4d3a2c	104
476	StopMissionMessageHandler	4d3a94	224

Note the CPP emission order differs from BTL4APP.HPP's declaration order (ModeManager before ControlsManager before GaugeRenderer in the binary) — the binary order wins. [T1]

Decoded constants — all resolve to named 1995 expressions [T1]

ctor base arg 1 = BTL4 (enum ApplicationID {RPL4, BTL4, BTW4}, APPMSG.HPP:47 — note the unreleased BTW4 slot).
0x50421 = BTL4ModeManager::ModeInitial; 0x400000 = ModePlasmaDisplay; the +8/+4 writes = ModeManager::AddModeMask/RemoveModeMask inlines (MODE.HPP:56-70, byte-exact).
2/4 = LBE4ControlsManager::{PrimaryThrustMaster, PrimaryRIO} (L4CTRL.HPP IOType).
-1 mode mask = ModeManager::ModeAlwaysActive (MODE.HPP); keyboard registration = ControlsUpdateManager<ControlsKey>::Add(ModeMask, Receiver*, MessageID, Plug*) (CONTROLS.HPP:319; the vtbl+0x14 call on keyboardGroup[KeyboardPC] @manager+0x1a0).
Video ctor 5-tuple = (DefaultRendererRate, MaxRendererComplexity, DefaultRendererPriority, VisualInterestType, 1) — signature from RPL4VID.HPP; the two leading globals are DAT_004e8238/DAT_004e823c (extern consts, RENDERER.HPP:38/41). ⚠ the PORT's btl4app.cpp maps these to the WRONG positions (globals as args 4-5) — flagged for back-porting.
Audio ctor = (GetFrameRate(), GetMissionReviewMode()) — BTL4ARND.HPP:95 proves arg2 is Logical mission_review_mode; DAT_004fd550 = L4Application::missionReviewMode static. GetFrameRate naming for FUN_0044e19c is [T3] (port identification, not header-proven).
All string literals byte-verified via section_dump (51ebd6…51ed40), including "*****Mech has no controls mapping!*****" and the camera/mech "Thrustmaster"/"L4" mapping names.

Judgment calls (flagged, not proven)

Banner copies the stale rp4lbe4.cc RP-clone banner because the surviving BTL4APP.HPP carries the matching stale rp4lbe4.hh banner — the pair was cloned from RP together. [T4]
Include list & order mirror the port's reconstruction, trimmed to 1995 headers. [T4]
switch vs if/else for the two class cases compiles identically at -O; switch chosen to match VWE style elsewhere. [T4]
Camera-case mapper resource = stack SubsystemResource (decomp-proven: 32-byte name copy, classID 0xf, modelSize 0x30); mech-case mappers take a CString name (decomp-proven: CString ctor + refcount at the call) with ClassID/SharedData DEFAULTED (0x7dc, &DefaultData passed by the compiler). [T1 shape / T3 default-arg reading]
mech_mapper = NULL; after Fail() reflects the decomp's puVar3 = 0 (Fail returns in OPT?); kept. [T1]
description->resourceAddress / ControlMappingsListResourceType (== 6) member/constant names follow the port's ResourceDescription reading; verify against the 1995 RESITEM/RES headers when that TU is done. [T3]
Two Check* lines were omitted in MakeViewpointEntity to meet the line-400 constraint; Check macros are invisible in OPT builds so the choice is evidence-neutral, but WHICH lines the original spent is unknown — revisit if a Borland build ever line-checks. [T4]

Line-constraint ledger

BTL4APP.CPP:400 = the only Fail in the TU ✅ met exactly.

4.3 KiB Raw Blame History