Re-ran lane J's image-wide tempering-immediate scan at real instruction
boundaries and audited it site by site. Sixteen functions carry the two masks
inside a decoded instruction, 67 occurrences, and a brute byte scan finds zero
orphans -- recall is complete. Of the sixteen:
* ONE IS A FALSE POSITIVE. 0x008cca30 has no temper chain at all; the four
bytes read as the second mask are the rel32 displacement of a call.
* FOUR ARE RNG ENTRY POINTS, not game code -- the tempering there is the
primitive's own. Two were known; 0x004f7670 and 0x008e6e30 were not.
So the figure is ELEVEN game functions with inlined draws over 28 sites, not
fourteen. Exactly ONE of the eleven is reachable from StrategyServer::ProcessTurn
(0x007aa240, depth 4) and one more from OnAllCombatDone_Tail (lane J's R2). The
other nine are map setup, the lobby, the network layer and two scripted
encounters, all reached through vtable slots with no direct caller -- so their
absence from the turn closure is proved for direct edges only, and that caveat
is stated as loudly as the result.
Both functions lane J handed over are read completely from the instruction
stream:
0x004f7670 (84 B) is Mars::RNG::NextUInt -- ONE WORD, UNCONDITIONAL, no loop,
no branch but the lazy twist. ECX is the generator OBJECT, where NextFloat
and NextInt take the object PLUS FOUR; ProbabilisticJump uses both conventions
0x6b bytes apart. Ghidra's size is right here, which is worth saying.
0x007aa240 (Ghidra says 944; the body is 953 and ends past the reported range)
is the encounter-detection roll. ONE INLINED NextFloat PER (contact, detector)
TRIAL, drawn BEFORE the accept test, so a detector holding neither of two
specific techs still burns a word and can never succeed. The accept test is
`thresh >= r` -- equality accepts -- derived from the fcompp/test ah,5/jp
encoding rather than the mnemonic. The outer repeat-until-no-progress loop
cannot redraw a pair: the "tried" bitset is filled above the back-edge target
and never cleared, so the whole call is bounded by |contacts| x |detectors|.
Three more draw entry points nobody had listed: a float range (one word, and it
NARROWS TWICE), a triangular integer range (at least two words), and a
truncated-normal range whose two draws are BOTH inlined and which costs two
words per attempt with an unbounded attempt count -- and which scales by 2^-32
where NextFloat scales by 1/(2^32-1). Two divisors, one image.
RECONCILIATION, stated honestly. The complete draw-site inventory of the
ProcessTurn closure is 22 sites: 21 entry-point calls plus the one inlined site.
This lane adds two previously-uncounted sources to that list and accounts for
NONE of lane Z's 18-20 words per turn with certainty, because both new sources
are gated and neither has been measured. Lane J's prediction that these two
functions would explain the gap is NOT confirmed. What is now provable is the
negative: there is no twenty-third mechanism, so the 18-20 words are distributed
among exactly these 22 sites. The search space closes; the count does not.
Also corrects, in place: my own Ghidra comment claiming the masks are a Mars
variant of MT19937. They are the textbook masks applied before the shift
instead of after -- (y & 0xff3a58ad) << 7 == (y << 7) & 0x9d2c5680, verified
over 200k words -- so mars::rng was never wrong, but a scan for the textbook
constants finds nothing in this image.
Fragment validated by generating to a scratch path (797 entries, no duplicate
name); the tracked header is untouched while lane Z is in flight. Eight
prototypes, nine plate comments and eight pre-comments written back to Ghidra.
The second save's numbers were not in the first draft's table. Correcting my
own correction: on a turn-19 Zuul game the counter moves 16, 21 and 44 times
between ProcessTurn entries, against 12-14 on an early two-colony game. So it
is not a driver-invocation counter and not a constant either.
Lane J landed rule 16 (inlined draws are invisible to call-graph sweeps) while
this run was in flight, and it bears directly on three claims here that rested
on direct-call sweeps. Those claims now rest on the behavioural measurement
instead, which is immune to it: ProcessNodeSpaceTravel moved the generator by
0 words on 16 observations and node-line decay on 8. The instrument does not
ask which function drew, only whether the generator moved.
Lane J's resolver map and this ledger pair up: the resolver has no
unconditional draw, so its prediction -- a plain fleet battle costs the same
18-22 words as a peaceful turn -- is testable with these hooks as soon as
someone builds a save where two hostile fleets meet.
We consume 18-22 generator words per turn and model none of them as a count.
All of it is inside StrategyServer::ProcessTurn; OnAllCombatDone_Tail costs 0
on every turn observed; the residual outside the two drivers is exactly 0. The
generator does not move between turns at all, so the interval a standalone has
to reproduce is closed at both ends.
The instrument reads generator STATE, not calls, and that choice paid: the
image has four draw entry points, not three (NextUInt 0x004f7670 is in no
lane's primitive set) plus inlined draws in twelve functions, two reachable
from the turn roots. A primitive-counting hook would have undercounted
silently.
Checked against the save files independently: the turn-6 autosave pair gives
18 words read from the two Sim.RNG blobs, and with twists == 0 that number
never passes through a twist implementation -- so the two instruments do not
share the hidden assumption they could have.
Corrections to combat-done-tail.md, in place:
* the node-line 0x20000-fleet check runs AFTER the Chance(0.5f) call and
cannot gate the draw; the expiry test is NodePath::RemainingLife 0x006e2130
and is now a formula rather than a description
* StrategyHost::Autosave is ret 8 and returns the std::string* in EAX
* SNMAllCombatDone IS delivered every End Turn (8 of 8) -- lane K's inference
was right; the stronger no-encounter reading is narrowed, not closed
* S+0x8 advances 12-14 times per turn, not twice
Node-line decay still has not fired. The hook reports the distance instead of
the absence: 51 of 53 lines are permanent, the mortal ones are dug ~1/turn by
the Zuul, each ~40 turns from expiry. It stays a labelled hypothesis.
Closes lane K's #1 ranked gap: FUN_007d5af0 under phase 6 of
OnAllCombatDone_Tail, the second RNG source in the strategic tail.
RNG inventory (the deliverable that matters). The resolver draws nothing
itself. Three sites in its subtree, each behind a function with exactly
one caller:
R1 RNG_NextInt @0x007bb69b (node cannon) 1 word w.p. 3/4, mean 4/3
R2 an INLINED RNG_NextFloat @0x007a84bd 1 word per back-eng candidate
R3 RNG_NextInt @0x00852ec7 (project pick) >=1 word per successful roll
R2 corrects combat-done-tail.md 3, which reported no NextFloat in the
subtree: the draw is inlined, so the only call-graph edge it leaves is
FUN_007a7f30 -> RNG_Twist, which reads as a bare Twist and is not one.
An image-wide instruction-boundary scan for the MT tempering immediates
finds 14 game functions with inlined draws; two of them, FUN_004f7670
(84 B) and FUN_007aa240, are reachable from StrategyServer::ProcessTurn
and are candidate mechanism for part of lane Z's unexplained 18-20
words per turn.
Also: the real body is 7641 B, not Ghidra's 7499 (which ends
mid-instruction); only 8 non-stack stores and 1 indirect call in the
whole function; 23 EVENT_* keys in the closure, five of them new to the
campaign's combat picture (plague is decided inside combat); the
resolver's subtree DOES write the SETurnResults accumulator at
S+0x2f4[PlyrIdx]+0x90 via FUN_007baef0 -> FUN_007b9df0; and
sizeof(Game::TacReport) = 0x94, enumerated twice.
Two structural errors made and corrected before publication are recorded
in 7.3 and 4 rather than quietly fixed.
Gates run separately: clean-room check OK; host ctest 36/36.
findings/objects/turncommands-block.md -- why lane W's 44-item recovery and the
save's 35/38/41/61/123 items disagreed, in mechanism terms, plus the per-list
element records read off all 22 distinct helper writers.
Three separate causes, only one of which was a branch:
* layouts.md is OFFSET-SORTED and Game::TurnCommands is one of the 89 classes
whose offset order is not write order, because each gate bool sits after the
payload it gates. Align saves against streams.json, never layouts.md.
* the writer's tail is 27 std::list members, each written by its own helper as
WriteInt(size) then size element records. The linear recovery keeps one item
per CALL SITE, guesses its kind from an element field and drops the count.
44 - 17 members = 27 = the list count.
* 8 prologue items + 27 zero counts = the 35-item empty block, which is why
every earlier save was bit-identical here.
Corrects lane O's section 11 in place: the fleet-move element is
{fleetId, nHops, nHops x systemId} -- a counted route vector, not a fixed
{fleetId, 1, destSystemId, 0}. Only the three-item reading makes the item totals
close (a four-item element needs 26 lists on human-turn2 and 28 on zuul-turn15;
the writer has 27). Also records that neither noderoute save carries a fleet
move, so the node-route UI does not queue through that list.
22 of the 27 lists are labelled hypotheses: read off the instruction stream, but
no save exercises them and no element value has ever been observed. List 14 is
observed but not understood -- issuing one fleet move also queues
{sameFleetId, 0, true} there.
save_reader.py deliberately left alone, with the reasoning written down: it reads
CD generically for the AIAgent bodies too, state_checksum's digest tree is built
from the generic tree so typing CD would add no named leaf to any diff, dispatch
by CDT ordinal would need new machinery in the oracle, and rule 8 is pointed at
exactly this move -- a layout mirrored into both readers from one reading is not
two checks. coverage: PROVED, 0 error 0 warn on all eleven saves, unchanged.
ghidra/addresses.d/lane-q.json: 6 entries (the writer, its Read pair, and the
four observed list helpers). Validated by generating to a scratch path; the
tracked header was not regenerated.
gen_addresses.py distinguishes "addr" (VA -> RVA) from "offset" (a member
offset). Three lane-K entries used "addr" for offsets and generated negative
RVAs, which broke the engine build. Two are re-expressed as offsets; the third,
StrategyServer_off_RNGPtr_S_frame, is withdrawn - addresses.json already has
StrategyServer_off_RNG = 0x16c for the S frame alongside off_RNGPtr = 0x168 for
S+4, which is precisely lane T's rule. The second call site found this lane
confirms it rather than correcting it.
The second turn driver, 0x007d92a0, read byte for byte. Corrects turn-driver.md
section 5: of the four subsystems it said live here, only bankruptcy does.
- 36-phase map with strides enumerated (EncounterResults 0x178, Encounter 0x74,
member 0x44) and the arity check that logs but does not return.
- Phase 7 is encounters.clear(), not a filter: the erase pair is the same
four-argument shape vector<Encounter>::operator= uses, and both arms converge
three instructions later.
- Bankruptcy: ProcessBankruptcy at phase 15, UpdateBankruptcyLimits at phase 31.
Three corrections to formula-gaps Q1 - the divisor is the double
-0.15000000596046448 not -0.15, the per-system income term is clamped at 0
before summing, and the 3.3 factor lives in .bss and is DB-loaded.
- Turn results are FILLED here (phases 6, 11, 18 write S+0x2f4[PlyrIdx]) but
rotated by ApplyEncounterResults and dispatched by SynchronizePlayer as event
0x25 afterwards. sizeof(SETurnResults) = 0x11c, enumerated five ways.
- BuildTurnEvents is misnamed: it is the setup/load/rejoin resync push, gated on
a pending descriptor, and references no EVENT_ string at all.
- TurnEvents_Write and TurnEvents_Read are swapped in Ghidra (layouts.json is
right). sizeof(TurnEvents) = 0x18, enumerated four ways.
- The autosave: StrategyHost::Autosave 0x00895210, its four localized paths, the
rotation that fires only on the post-turn call, the connection detach around
the write, and why the payload carries nothing time-, name- or machine-derived.
- Two RNG sources in the tail that nothing models: one NextFloat per expired node
line, plus draws inside the combat resolver. Both run before the autosave.
- S+0x8 advances twice per turn, not once.
Repo-wide correction: the research-event roll costs one or two RNG words, not
one. Fixed in unlock-cascade.md, addresses.json and lane-u.json; the captured
compare artefacts under verify/results are left alone as run records.
Produces the game states several board rows were blocked on, and reads each
one back out of the save before claiming it.
- CD/TurnCommands_v5 decoded: three saves with issued orders (research target,
research boost, build orders, system rates, colonize, fleet move). Layout in
SAVE_FORMAT.md section 11; ids cross-checked against the same file's Sys/Flt.
- Waypoint type 3 (NodeRoute) now has 16 instances in one save and the player's
own fleet is a mover in two. Type 2 is shown NOT to be produced by a Move
order for either node-drive race (Zuul and Human both emit type 3).
- research_roll_pending true at load in two saves; the ProcessTurn gate constant
at 0x00a2c788 is a float 0.5, so the flag survives into ProcessResearch only
while progress/cost <= 0.5 at the start of the turn.
- Budget tail: construction and a large maintenance are live; Nexp is empty in
all 11 saves and the 1.8 UI has no expense slider, so slot 12 looks dead.
VM140 restored to the original 8-file SavedGames set (autosaves byte-identical)
and left at the main menu; VM140 row set FREE.
Phase map for both turn drivers, byte-for-byte from objdump rather than the
decompiler. Corrections that matter:
- the end-of-turn tail is UNCONDITIONAL. turn-spine.md 2.4 step 11 said it is
deferred when encounters are pending; that came from reading an inlined
std::vector destructor as a branch (both arms converge at 0x007dcb38). There
is no branch on the encounter snapshot in the function, and 0x00794ad0 does
not build an encounter list -- it builds the 15 ship-action type ids.
- ServerPlayer::ProcessTurn takes one float argument and never reads it.
- the ResearchRollPending site runs AFTER ProcessResearch, not before, and
clears the flag only when the roll actually fires. Threshold is a strict
0.5f < progress/Cost. Both consumers located by a full displacement scan.
- EVENT_NO_RESEARCH's gate is 'completed nothing this turn', not 'nothing
available' -- 0x00584e50 collects RESEARCHED techs, not available ones.
- a fired research roll costs one or two RNG words, not one: the plague branch
draws a second NextInt and posts EVENT_PLAGUE_OUTBREAK.
- two StrategyServer bases four bytes apart, and two per-turn counters.
- new: the research refund at 0x008914aa, previously unmodelled.
45 addresses in ghidra/addresses.d/lane-t.json (669 -> 722 entries, no dupes).
Game::StrategyAIAgent::Streamable typed end to end. The writer is entirely
unconditional; the branch the decompiler shows is an inlined vector destructor
whose operator delete is marked noreturn. dsh is a std::map's _Mysize, so it is
the (pid, trns) count and not a scalar; lnat and lat are scalars, not elements.
AISystem::Write never reads its object, so every AISys body on disk is empty
regardless of game state. StreamableEnum<T> is a frame holding one int, which
makes SysMem/mts/nalat arrays of frames rather than of ints.
Named coverage 98.0% -> 99.9% on all four saves; state_checksum.py unchanged
and still PROVED. Section 7 lists the ten element layouts no save exercises and
says what workload would settle each.
29 addresses in the per-lane fragment; merges to 669 with no duplicate name.
Part 1 (notes side): findings/objects/svsctob-variants.md records the two maps
that are nowhere on the wire and were read out of the game -- EncID -> class
from a 23-entry dword jump table at 0x0052bf60 indexed by EncID-1 (0x0052bf00),
and xscn -> class from an exhaustive four-way _stricmp chain at 0x005a7050.
Twelve live EncIDs, four scenario names, and "indsys" =
Game::SVSOIndependentSystems whose Read and Write are both the shared `ret 4`
stub at 0x005f8ac0, so its empty frame is correct output rather than a
truncation. Also: SVSOSots::Read accepts NPCPlr and hastraps, which Write never
emits -- read-only backward compatibility, not a hole in the recovery. And a
correction to the recovery itself: SVSOCrowDefenders writes `dsys` INSIDE the
ndsys loop; layouts.json calls it a plain member and no save can settle it
because both counts are 0 everywhere. 13 addresses in ghidra/addresses.d/
lane-w.json; gen_addresses.py merges to 640 with no duplicate name.
Part 2: lane G found four defects in both readers and deliberately did not
patch the oracle mid-campaign. Fixed now, with tests, and byte-neutral.
1. Game::SystemParams field 1 is a string, not an int (empty string == four
zero bytes == int 0, so it round-tripped by luck).
2. ObservedTech/ObservedWeapon odet is a bool, not an int (byte-safe only
because a 4-char tag makes both items 12 bytes; 3 chars would not).
3. SpeciesRatios nv is a count, not a field.
4. ShipRecords srbd is a count, not a field -- and this one is behaviourally
confirmed, not inferred: srbd takes 0, 1, 3 and 4 across the players and
every non-zero count is followed by exactly srbd x 5 scalars.
Note that 3 and 4 were an ABSENCE in save_reader.py, not an error: ShipRecs and
civr were both A(..., "any"), so the fix had to add the shapes rather than
retype a field.
Byte-neutrality: every item's inflated offset is unchanged on all four saves
(38,933 / 39,843 / 40,300 / 35,771 offsets, sequences identical), so no item
boundary moved. state_checksum.py still reports coverage: PROVED on all four
with the same rebuilt byte counts. The /CreateParams and /Sim/players digests do
change, because they hash typed VALUES and two fixes change what a value is --
and the value-byte deltas balance exactly: odet items x 3 plus p1 items x 4.
--strict exit 0 on all four saves; tests 36 -> 48.
findings/objects/wire-schema-closeout.md carries the whole account, including
the proof that CD/TurnCommands_v5 cannot be typed without a save that has
issued orders.
Reads: TechTree::PrereqsMet 0x0057d8e0 (AND of ORs; zero groups TRUE, an empty
group FALSE), the prerequisite layout at TechDef+0x88/+0x98, the tail collector
0x00587cc3, and the head of OnTechResearched (RecordObservedTech unconditional;
the research-event roll gated on ResT == def && ResearchRollPending).
Live: 35 compared calls over three workloads, 0 divergences, tracecmp exit 0,
End-Turn oracle hashes unchanged. The EVENT_TECHS_UNLOCKED residual lane P
predicted and lane V measured is closed by running the cascade, not by posting
on completion.
12 new addresses in ghidra/addresses.d/lane-u.json (header 615 -> 627).
objects/layouts.json is a memory-layout projection: build() sorts fields by
off_abs (89 of 386 classes have offset order != write order) and merges
duplicate offsets into alt_tags, which is exactly the JewelsOfTheCrown
double-tag trap. Both losses are the substance of the on-disk format.
tools/streams.py is a second projection of the same recovery that keeps the
program order Lab.layout() already computes and the repeated tags, and drops
every memory fact — no off, size, sizeof, gaps or strides. The engine must read
and write the format, not inherit the original's ABI.
tools/gen_stream_schema.py emits sots-engine's include/generated/sots_stream_schema.h
under the same discipline as gen_addresses.py: generated, provenance header,
never hand-edited.
386 classes, 2042 wire items.
Lane P wrote the expected numbers into sots-engine docs/P-events-wiring.md §4 before
this run existed. Every one held.
First End Turn (ref-turn2 -> Launch -> End Turn, shim.cfg.recapb3 unchanged, build
eventlive-dd38117-20260908T0916Z, main dd38117 with no source change): 3 calls, 3
compared, 0 divergent, tracecmp exit 0 — where lane R's run exited 1 on
side.events.after.v.next_id orig=4 ours=3. Call 0 read turn=3,
events_turn_bucket_exists=true, events_next_id_in=3, events_in_turn_bucket=1,
events_dedup_risk=0, no events_scan_truncated, next_id 3->4 on both sides,
turns/turns_bytes 2/48 unchanged, node[144].progress 2879->5768, flag 1->2, rng
identical — the whole predicted list.
Five End Turns: exit 1 with 2 divergent calls instead of 3, each short by exactly 1
(next_id orig 7 ours 6, orig 12 ours 11) = the deliberately unmodelled
EVENT_TECHS_UNLOCKED. A 0-divergence result there would have been suspicious.
sizeof(Game::ObservedTech) = 44 measured live: observed_techs.bytes grew by exactly
44 on both completion calls (440->484, 484->528), confirming lane X's static pin
behaviourally. Non-researching players measured 880 = 20 x 44 and never moved.
One deviation from lane R, and it is a workload effect, not a defect: rng diverged on
call 9 because the completed tech had research_roll_pending set and the
OnTechResearched callback drew one word ours does not (left 374 vs 375). Lane R's
"RNG matched 15 of 15" was workload luck; the honest statement is that the rng region
matches on every call that does not complete a roll-triggering tech.
Zuul double roll CLOSED. No species-5 save existed, so this lane made one: a custom
game with only Zuul in the Available Species pool, 4 End Turns under the same compare
config. 8 calls, 8 compared, 0 divergent, exit 0; all four researching calls have
species=5 and advance the generator by two, not one (left 540->538, 522->520,
504->502, 485->483), and ours reproduced every post-state bit-for-bit. Save added as
verify/results/saves/zuul-turn5-species5.sav.
End-Turn oracle byte-identical to lane R's on the first turn, so none of this
perturbs the game. clean_room_check OK and host ctest 33/33, run as separate commands.
VM140 restored to the recap build at the main menu and released.
