Two independent defects, both making the script report a correct checkout as
wrong. Neither is new; both were invisible because nobody ran it here.
1. `export/` is the exporter's OUTPUT and is gitignored (`/export*/`). A
checkout where nobody has run the exporter has no `export/` at all, so the
four `DECISIONS.md`/`BLOCKED.md` citations of `export/manifest.json` and
`export/screens/...` landed in "resolve NOWHERE" and the check exited 1 --
red for a state no edit can fix, which is the exact shape its own docstring
says it exists to avoid. `check-capture-citations` learned this for
`docs/re/captures/`; same rule now: absent BECAUSE UNBUILT is reported,
absent while the tree IS built still fails. Verified both ways -- `mkdir
export` and the same four go back to failing.
2. The selftest's peer-branch fixture cited
`docs/re/f5-a-press-snaps-the-plate.md`, which the consolidation made an
ordinary local file. The fixture stopped testing the scanner and started
reporting it broken; `PEER_REFS` also still named
`origin/auto/frame-blend-draw-path`, a branch that no longer exists. The
selftest now FINDS a peer-only path at runtime, and where none exists -- the
normal case on a clean checkout, measured: zero -- it says the class is empty
here rather than claiming a failure. The class itself stays: the next topic
branch that lands a finding recreates the condition exactly.
Measured: check 123 citations, 119 resolve, 4 unbuilt, 0 nowhere, exit 0 (was
exit 1). Selftest ok (was 🔴 BROKEN, on `main` too). The new generated-tree case
was confirmed to FAIL against the unfixed function first.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`committed()` built the directory set from `range(4, len(parts))`. Index 3 is
`docs/re/captures` itself, so every SUB-directory was registered and the root
never was. Eight pages cite the directory as a whole -- PROTOCOL's evidence
table, CONSOLIDATION, CONTAINERS-AND-AGENTS, decoder-loop, both HANDOFFs,
BLOCKED, verify-screen-blend-divergence -- and each of those looked like a
citation of a file that does not exist. `check-capture-citations` exited 1 on a
correct checkout, which is the failure mode its own comments warn about: "a
gate that is red on every clean checkout" teaches people to ignore it.
The selftest could not see this. It takes `next(iter(dirs))`, and an arbitrary
member of that set is always a sub-directory -- the one case that works. It now
asserts the root explicitly, with and without its trailing slash. Both new cases
were confirmed to FAIL against the unfixed function before the fix went in; a
selftest case that passes either way is decoration.
Measured: full check 203 citations, 212 captures, 0 dangling, 0 tracked assets,
exit 0. Before: exit 1.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The database mount below was one of three ways the decoder could not reach its
own oracle. The other two are here.
`run-canary` never looked in `Checked/`. It tried `Release/` then `Debug/`, and
both of those exist on this box -- an Aug 28 binary and a Jul 19 one. They boot
the game perfectly well and carry NO `audit_61` branch probe, so a probe run
against either returns zero hits that read as a finding about the game rather
than as a stale binary. Configuration is now the outer loop and location the
inner one, so a `Checked` build anywhere beats a `Release` build anywhere;
`$XENIA_BIN` still overrides everything. Measured here: `Checked` has
`audit_61_branch_probe_pcs`, `Release` and `Debug` do not.
The launcher also now says which instrumentation is missing BEFORE the run,
because the alternative is reading an empty log afterwards and guessing.
`build-canary` built `$PROJECT_DIR/xenia-canary`, which does not exist in this
container -- the source is bind-mounted at `/canary` and the launcher already
exports `XENIA_SRC=/canary`. CONTAINER-NOTES has carried that defect since
2026-08-29 with a symlink workaround and a warning to remember to delete the
symlink afterwards. It now reads `$XENIA_SRC` first, so there is nothing to
remember. Its default configuration moves Release -> Checked to match what
`run-canary` picks; the old default spent a full build on a binary nothing ran.
Two documented blockers are refuted rather than deleted, since the sequence of
wrong readings is what makes the right one checkable: the CONTAINER-NOTES
symlink dance (the warm build volume it was configured against is gone too,
removed in the 2026-09-18 cleanup, so the next build configures cleanly against
`/canary`), and `upstream-baseline.md`'s "`version.h` is never generated" --
`CMakeLists.txt` generates it at configure time now, with a stub fallback.
`decoder-loop.md` claimed the oracle was at `Linux/Release/` and that the probe
was on two side branches; both were true when written and neither is now.
Verified: five pick_bin cases against the extracted function body, and `strings`
on all three real binaries.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`/sylph-canary` was dead at both ends, measured rather than assumed:
* `run-canary-safe.sh` runs the Wine build and defaults to
`xenia_canary_i2d.exe`, which DOES NOT EXIST. The `xenia_canary.exe` that
does (2026-06-19) has ZERO strings for `audit_61_branch_probe_pcs` and zero
for `RE-INPUT`/`RE-DRAW` -- it predates all our instrumentation, and
refreshing it needs the clang-cl + xwin toolchain on `cross-build-wine`.
* both native launchers defaulted to
`xenia-canary-native/build/bin/Linux/Release/xenia_canary`. That directory
does not exist on this workspace -- the checkout is `xenia-canary/` -- and
Release is not even a target in that build tree (`ninja -n` lists 0 steps),
so its binary is frozen at 2026-08-28, before the probe.
Native launchers now default to
`xenia-canary/build/bin/Linux/Checked/xenia_canary`: the config this tree can
actually build, rebuilt today with the audit_61 probe (Canary PR #2), verified
by `strings` -- audit_61 8 hits, RE-INPUT/RE-DRAW 4. `Checked` is optimised WITH
assertions, which is what a behavioural reference should be: a bad state stops
loudly rather than being sampled.
The Wine launcher now ABORTS with the reason and points at the native one,
instead of failing somewhere downstream on a missing file.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`SYLPHEED_DB=/xenia-rs/sylpheed.db` named a path that no longer exists. The
xenia-rs repo was retired by the consolidation, its local clone was deleted, and
the `/xenia-rs` mount that served the database was removed in #61 because it
pointed at nothing. The env var stayed. So the decoder had NO database: `zq.py`
and `/sylph-dis` -- most of what a static-RE brief asks for -- could not run.
Mounted read-only at the container's repo root instead, which is where `zq.py`
looks when `$SYLPHEED_DB` is unset, so there is no variable left to drift out of
step with the mount. That drift is the whole bug: a path in an env var and a
path in a mount, maintained separately.
Read-only is deliberate. The host owns the file, DuckDB takes an exclusive lock
to write, and two agents plus the human sharing one database would corrupt it.
Regenerating means writing elsewhere and pointing $SYLPHEED_DB at it.
Missing-file cases now say so and print the command that builds one, rather than
starting an agent that discovers it mid-iteration.
Verified in the real agent image: DB mounted, no env var set,
`zq.py fn 0x824609C8` -> `Pak_FindEntryByName`, `zq.py classes` lists RTTI.
The brief's tooling table also claimed the old path, and said nothing about the
oracle binary; both corrected. It now records that the built Canary carries
RE-INPUT/RE-DRAW but NOT the audit_61 branch probe -- measured with `strings`
on both built binaries, zero hits; it is on two other branches (fork issue #1).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Removes the last reason this repository depends on its own history.
The pin was deliberate and carried its own exit condition, written into the
comment above it: "revert to the path dependency the day the tag is an ancestor
of `main`." Measured -- `formats-pin-2026-09-01` (e2630413) IS an ancestor of
`main` now, so the condition is met, and the cost the comment named goes away
with it: while the pin held, `sylpheed-cli` built from the workspace crate and
the exporter from the tag, so `tools/port/verify-screen` compared two eras
instead of detecting drift. They read one decoder again.
It also removes a failure mode nobody priced in. Depending on this repo by tag
is what made the #49 history rewrite break the build: every commit was replaced,
the locked rev vanished, and clean checkouts could not resolve it while the
rewriting machine kept working off its `~/.cargo/git` cache (PR #60). A path
dependency cannot fail that way. `Cargo.lock` now has ZERO references to
Sylpheed.git.
⚠️ This moves the exporter across a 243-file decoder change (8 467 insertions),
so it was gated on the full suite rather than a compile:
45 suites / 377 passed / 0 failed / 14 ignored, cargo exit 0
45 binaries launched, 45 reported (a SIGKILLed suite prints no result line
and would otherwise vanish from the tally)
corpus report: PRESENT for disc, res3d and iso
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Issue #49 removed game assets from git. It also, silently, removed the only
transport agents had for showing each other evidence: each agent works in its
OWN clone, so "commit the screenshot, the other one pulls it" was the mechanism,
and PROTOCOL.md's file table said so in as many words -- "evidence cited by a
finding -> git". That row sat directly above "🔴 Never commit game content",
which is how 76 MB accumulated: the two contradicted each other and the table
won, because it was the one that told you what to do.
WHAT REPLACES IT
* one host directory, `Sylpheed/docs/re/captures/`, bind-mounted read-write
into BOTH agents at /work/docs/re/captures. All three -- host, decoder,
port -- see the same files live, every citation resolves, and nothing can
reach git history. Read-write on purpose: showing each other a screenshot
is the point.
* PROTOCOL.md's table rewritten. Cited evidence -> present but never
committed; derived measurements (csv/tsv/txt/log/json) -> still git, they
are our numbers not game content; evidence that must cross MACHINES ->
attached to the issue or PR, because a bare clone has no captures.
Verified, not assumed: container A wrote a .png there, a SEPARATE container B
read it back, the host saw it, `git status` reported 0 changes, and
`git check-ignore` named the rule.
THE CHECKER WAS RED ON EVERY CLEAN CHECKOUT
A fresh clone/worktree/CI has no captures, so it called all 134 citations
dangling and exited 1. A gate that is red before anyone changes anything is one
people learn to ignore -- the exact failure this file already carries a comment
about. It now distinguishes "no captures here" (expected, explains itself,
exit 0) from "these are missing" (real, exit 1, unchanged when assets ARE
present). Both paths tested.
ALSO
* `sylph-decoder` no longer mounts `xenia-rs` -- retired repo, gone from disk,
the mount pointed at nothing.
* CONSOLIDATION.md closed: it still described captures as committed and the
history fork as undecided. Both are settled.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`crates/sylpheed-export` depends on THIS repository by tag:
sylpheed-formats = { git = "…/Sylpheed.git", tag = "formats-pin-2026-09-01" }
and `Cargo.lock` pinned it to commit `#1cd5b8b1`. The issue-#49 history rewrite
replaced every commit, so `1cd5b8b1` no longer exists on the remote and the tag
now dereferences to `e2630413`.
⚠️ The failure is invisible here: `~/.cargo/git` still has the old object, so
builds on the machine that did the rewrite keep working. A clean checkout --
CI, or the other desktop -- cannot resolve the locked rev at all. That is the
worst shape for a breakage, which is why this was measured from an empty
CARGO_HOME rather than trusted to a local build.
`CONSOLIDATION.md` called this dependency out as the hard blocker for the
rewrite. It was not re-checked before the push; this is the fix.
The dependency's SOURCE is unchanged -- `crates/sylpheed-formats` has tree
`55461e46` at both the old and the new commit, so the rewrite never touched it
and the build result is identical.
Verified:
* remote tag peels to e2630413, matching the new lock
* `cargo fetch --locked` from an EMPTY CARGO_HOME -> exit 0
* `cargo check -p sylpheed-export --locked` -> exit 0
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Issue #49: the repos carry code, tooling and docs only.
Untracks 143 screenshots, 3 savegame blobs and `tools/re-capture/ob_digits.png`
(a digit-template sheet cut from game frames) -- 146 files, 76.4 MB. They stay
in the working tree and are gitignored, so the pages' relative links still
resolve where the captures exist and nothing ships.
Derived measurements (csv/tsv/txt/log/json/jsonl/npy) are our own numbers, not
game content, and stay tracked -- they are what most claims rest on.
`check-capture-citations` had its contract inverted, and it is the half worth
reading:
* presence now comes from the WORKING TREE, not `git ls-files`. The assets are
deliberately untracked, so asking the index would report every screenshot as
missing and fail all 203 citations.
* a NEW failure: a game asset that IS tracked. A screenshot that sneaks back
in is invisible in review -- a binary shows as "Bin 0 -> 1234567 bytes" --
and is permanent once merged, since removing it later needs a history
rewrite. So that half has to be loud.
Verified:
* selftest 8/8, including the new rule
* scan: 212 present, 212 cited, 0 dangling, 0 tracked -> exit 0
* force-add one PNG -> "game assets TRACKED: 1", exit 1, selftest red
⚠️ This does NOT remove the blobs from history; a clone still fetches them.
That needs a filter-repo rewrite and a force-push, which is a separate,
human-run step.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`slb_leading_segment_disc` was SIGKILLed by the OOM killer in `docker/ci/run`
at its 7 GB cap, so the documented 45/377 baseline did not reproduce on a 15 GB
box.
`PakArchive` holds the whole concatenated payload in memory and `sound.pak` is
1.01 GB (sound.p00-.p04). `bank()` and `bank_named()` opened it on every call --
inside loops -- and five tests opened their own besides, ~26 opens in all. With
cargo's default thread count that is ~6.1 GB of archive in flight against a
7 GB cap with `--memory-swap` equal to `--memory`, so there is no swap to
absorb it.
The archive is immutable once open and every accessor takes `&self`, so one
`OnceLock` instance is equivalent to N private ones at 1/N the memory.
⚠️ The failure mode is worth knowing: a SIGKILLed suite prints no
`test result:` line at all, so it disappears from a scraped tally instead of
failing visibly. The run still reported "0 failed" -- true, and useless. Check
cargo's exit code (101), not the tally.
Measured in the capped container, 7 GB, default threads:
* before: SIGKILL (signal 9), 0 of 10 tests reported
* after : 10 passed in 3.17s
* (single-threaded before the fix: 10 passed in 22.64s -- the fix is also
~7x faster, because it no longer re-reads 1 GB from disc 26 times)
* cargo fmt --check clean; cargo clippy --tests -D warnings clean
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Docker on the dev boxes is rootful, so without `--user` every byte the build
writes into the bind-mounted repo is owned by root and the user needs `sudo` to
delete their own artifacts. This is not hypothetical: `export/` in a working
tree held 227 root-owned paths (149 MB) from earlier runs, and the `sylpheed.db`
regen in the workspace CLAUDE.md writes straight into /work, so it lands
root-owned every time.
The catch is that the daemon creates a named volume root-owned, so a `--user`
container cannot write /cargo or /target at all. So take ownership of both
volumes first -- once, and only when it is actually wrong, since a recursive
chown across a ~36 GB target volume is not something to repeat per invocation.
Both are sampled, not just one, because an older run can leave them drifted.
Volume names become overridable (SYLPH_CI_CARGO_VOL / SYLPH_CI_TARGET_VOL),
which is what let the chown path be tested without touching the real caches.
Placed above the corpus-mount block so it does not collide with #53.
Measured, not assumed:
* fresh root-owned volumes -> chowns once, then writes as uid 1000
* second run -> no chown, correctly cached
* `cargo check -p sylpheed-ppc` through the runner -> passes, exit 0
* a file touched in /work -> owned fabi:fabi, removable without sudo
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`GAME_CONTENTS.md` sat in the workspace root through the whole consolidation,
adopted by neither repo -- the corpus documents formats in depth but never said
what files the disc holds or where they sit.
Re-measured rather than transcribed, and that caught a real error: the source
placed `resource3d/`, `DefTables` and `MiscBin` under `dat/`. All three are in
`hidden/` -- which is why `SYLPHEED_RES3D` points at `hidden/resource3d`.
Counts, the language table and `media_id` (0x2D2E2EEB, from the XEX header)
re-verified against the retail extract.
Half the source was pre-RE speculation phrased as status -- `dat/*.pak` marked
"Unknown, magic bytes TBD" when the container is decoded disc-wide, plus a table
guessing each archive's contents from its name. Carrying that forward would put
claims into the corpus the corpus has already refuted, so it is dropped and the
page says what was dropped and why. INDEX.md stays the single authority on
format status.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Resuming happens on the other machine, so anything that lived only in this
workspace either comes into the repo or gets named as something to carry.
**The launchers come in.** `run-canary-safe.sh` (Wine) and
`run-canary-native-safe.sh` (native) are the ONLY sanctioned way to start Canary
from the editor — they force software Vulkan and refuse to launch while a
hardware Vulkan device is visible, because Canary on the AMD GPU takes VS Code
with it. They sat in the workspace root, outside git, hardcoding one machine's
absolute paths, which is exactly why they could not be checked in. They now
derive the workspace from their own location and honour `$SYLPHEED_ISO` and
`$CANARY_BIN`, so the layout is a default rather than a requirement. Same for
`run-canary-native.sh` (interactive, hardware Vulkan — not from the IDE),
`diagnose-freeze.sh`, `live-guest-state.sh`, `heaptrack-wrap.sh` and the
`asound-null.conf` the native launcher needs beside it.
Both safe launchers were run from `tools/` before this commit: the native one
reached content in 20 s (`VERDICT: HEALTHY`, title + 25,398 XMA lines, no
faults), the Wine one ran 25 s and logged 4 `ADV.wmv` hits.
**`docs/agents/HANDOFF-2026-09-18.md`** records what changed since 2026-09-16,
what is on the server, what cannot travel through git, and the traps this machine
paid for — chiefly that the guest-PC branch probe exists only on the fork's
snapshot branch, so a rebuild from `sylpheed-re` silently loses it.
**`HANDOFF-2026-09-06.md`** gets a correction note rather than an edit: a plain
clone works again now that the 545 MB branch is gone, and its baselines are two
baselines old.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
All three jobs used `dtolnay/rust-toolchain@stable`, so the gate resolved to
whatever stable was on the day it ran. A floating lint gate is not a gate: the
same tree goes green or red by date, and that already produced a disagreement
between two people reading the same commit — `collapsible_else_if` is `warn` on
1.92.0 and `allow`-by-default pedantic on 1.98.1, so both readings were correct.
1.98.1 is what run 206 resolved and what `docker/ci/Dockerfile` already pins, so
`docker/ci/run cargo clippy …` on a desktop becomes a true stand-in for this
workflow instead of an approximation. The header says how to bump: the three
refs here and the Dockerfile's `FROM` in one commit, in a PR of its own, where
the lints the new version turns on are the diff.
Also: the corpus-report step's comment quoted `207/0/14`, a tally two baselines
old. It now describes the shape of the problem without pinning a number that
decays.
Closes#15.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Finishes #16 in the three places its earlier remedies missed.
`tests/`: the last four local `disc_root()` copies now use `tests/common`, and
with them goes the one real hardcoded fallback — `ui_keyframe_record_disc.rs`
fell back to an absolute path on one machine, which made `unset SYLPHEED_DISC`
a no-op there. Control: with the corpus absent that suite now finishes in 0.00s
instead of 57.55s, so it skips rather than finding a disc of its own.
`examples/`: seventeen examples defaulted to `/disc`, the mount point inside the
CI container. Redundant there — `docker/ci/run` sets `SYLPHEED_DISC=/disc` — and
wrong everywhere else, where a missing corpus turned into a file-not-found
against a path that has never existed on the host. They now name the variable to
set, like the other hundred examples already did.
`docker/ci/run`: mount `$SYLPHEED_RES3D` and `$SYLPHEED_ISO` alongside the disc.
Only the disc was mounted, so an in-container run sat out the res3d and iso
suites while looking like a full one — the defect this issue is about, in the
runner itself.
Measured in the container on this desktop with all three corpora present:
45 suites / 377 passed / 0 failed / 14 ignored, and `sylpheed-corpus-report.txt`
now reports PRESENT for all three rather than for the disc alone.
Refs #16.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Project Sylpheed ships as a retail XEX2 (human decision, 2026-09-16), so
Phase 2's key harvest is closed rather than left looking unfinished.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
- Category pages linked each family as `<slug>.md`, relative to categories/,
where no family page lives. They now link `../<category>/<slug>.md`.
- Form pages linked a member into its *own* category directory, so every
VMX128 sibling (`vsldoi128`) pointed at vmx128/ although its family page is
under vmx/. They now link into the family's directory.
- Hand-written "Related" and sibling mentions linked other categories' pages
as if they were in the same directory. 109 are retargeted through the page
index; 29 that pointed a family page at itself (`vrefp128` on vrefp.md) and
6 naming instructions the manual has no page for are plain text now.
Regenerated at the existing Canary pin (f21ebd49e): upstream has moved on, and
re-pinning belongs in its own change. The generator reports 0 family pages
changed and is idempotent; the only dead links left are TEMPLATE.md's
placeholders.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Record the end state at the top of the page: four repositories archived, their
clones gone from this machine, the Phase 7 drops done or never present here, and
Canary's default branch moved off `xenia-rs`. The retention question and the
history rewrite stay open for the human.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Nothing here changes what a tool computes; it changes where tools look.
- tools/re-capture: 33 censuses globbed /work/sylph_extract, a path that has
existed nowhere since /work became a clone, so they matched nothing and
printed empty results. They now resolve the disc through a new disc.py
from $SYLPHEED_DISC and exit loudly without it (the #44 fix, generalised).
Nine scripts that imported siblings from the retired Reborn checkout or an
old session scratchpad now import from their own directory. unitgroup.py
only needs the variable when --pak is not given.
- sylpheed-xex: the loader only ever uses the XEX2 retail key. The dead
devkit key and a doc comment claiming a devkit fallback that does not
exist are gone; Project Sylpheed is a retail XEX2, so no XEX1 key either.
- sylpheed-viewer: real_font_rasterizes looked for /tmp/sylph_extract and so
always skipped. It reads $SYLPHEED_DISC now, and passes against the disc.
- Comments and docs that named xenia-rs, the Reborn repository or /work/*.pe
as places to look now name sylpheed.db, Canary's ppc_context.h and the
flat .pe; docs/re/README.md no longer says the native Canary build does not
run.
Historical records keep their original paths: findings that were measured
against /work/xenia-rs/sylpheed.db still say so.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Three untracked files in the project root had no home in either repository:
- the Xbox 360 technical reference report (a research compilation) becomes
docs/reference/xbox360-re-technical-reference.md, unchanged;
- XBOX360_ARCHITECTURE.md becomes docs/reference/xbox360-architecture.md,
trimmed to its platform facts. Its format "status" sections (PAK unknown,
mesh unknown, audio TODO) and the `just sniff` workflow predate every
decoder in this repo and were wrong;
- generate_export_docs.py becomes tools/generate_export_docs.py, the path the
committed xbox360-exports.* already name as their generator. It lived
inside a Canary checkout, so it now takes --canary and --out and fails
loudly on a tree that is not Canary.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The hand-written parts of the manual still described how the retired
xenia-rs interpreter behaved: its snapshots, Rust casts and helpers. Each of
those 490 statements is now either restated as what Canary's emitters and
x64 backend actually do (at the pinned canary_experimental commit), or
dropped where it only made sense for xenia-rs.
Checking them turned up claims that were wrong, not just outdated:
- VSCR[SAT] is never modelled in Canary (DID_SATURATE is a stub and mfvscr
cannot see it); the pages said saturating ops set it stickily.
- Canary does not implement lswi/lswx/stswi/stswx, dcbi, mtfsb0/mtfsb1,
vmsum*, vmhaddshs, vupkhpx/vupklpx, and most SPRs; pages described them
as working.
- Traps evaluate TO in Canary; stvebx/stvehx/stvewx store one element, not
16 bytes; mtmsrd writes only EE; fres/frsqrte/vrsqrtefp precision claims
and the stfs "rounds under RN / sets FPSCR" claim contradicted the spec.
- Reservations are a 64 KiB block bitmap plus a value compare, not
per-address tracking.
Claims that neither Canary's source nor a public spec settles are marked
unverified (NI at boot, vmaddcfp128 operand order, estimate bit-exactness).
Generated regions are untouched; re-running the generator changes nothing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Its data-table section globbed `/work/sylph_extract/**/*.pak`, a path that has
existed nowhere since `/work` became a clone. The glob matched nothing, so the
section reported "0 disc names" and an empty table with no error — while the
committed `docs/re/data/name-block-bases.txt`, generated back when the path
existed, records 13 450. The disc is required for that section, so a missing or
non-disc `$SYLPHEED_DISC` is now an error (exit 1), not an empty result.
Verified: without the disc, exit 1 with the message; with
`SYLPHEED_DISC=…/sylph_extract` (the original /work/sylph_extract), the section
reproduces the committed figures exactly — 13 450 distinct disc names,
50 / 261 data-table rows, `sub_82341A20 r30 0x82088F94 217 names 89% disc`.
⚠️ The artefact itself is deliberately NOT regenerated. Rebuilt from the current
database it changes findings — solved bases move, three 15/15 rows vanish, −280
lines — and the cause is the database, not this script: the string extractor's
Shift-JIS path requires kana, so kanji-only Japanese (e.g. the allocator's
`32B 未開放 %4d : 最大…使用量 %4d / %4d` debug lines) is not in `strings`, and
the bases those names pinned no longer resolve (sub_8285F2C8 r5 @ 0x820AE7AC:
15/15 before, 7 now). Regenerating would bake that regression into a cited
research artefact (`docs/re/structures/player-tuning-tables.md`).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Instructions and generated text that still sent readers to `xenia-rs`, which is
archived and deleted locally. Provenance and dated research records are left as
they are — "lifted from xenia-rs", or a finding stating what database it was
measured against, is still true, and rewriting it would falsify the record.
Changed, because each one tells someone what to do today or writes stale text:
docs/re/README.md the RE toolchain guide queried `xenia-rs/sylpheed.db`
with `xenia-rs/zq.py`, and its "Dynamic" bullet said
to prefer xenia-rs's probe suite OVER Canary. Now:
`sylpheed.db` at the repo root (how to build it and
re-stamp names), `tools/zq.py`, dynamic = Canary, and
`.rdata` reads from the `.pe` (offset = VA - 0x82000000)
in place of `--dump-addr`.
challenge-mission-gate.md, structures/achievements.md
reproduction commands `python3 xenia-rs/zq.py dis …`
crates/sylpheed-xexdb/SCHEMA.md
titled after the retired `xenia-analysis` crate,
citing `xenia-rs dis`, pointing at a `xenia-analysis`
source path
sylpheed-xexdb/src/formatter.rs
WROTE "generated by xenia-rs" into every disassembly
it produced (no test pins the banner)
sylph-xexdb.rs, db.rs "SQLite" / "DuckDB writer for xenia-rs"
sylpheed-formats/src/hash.rs
pointed at `xenia-rs/RE_SYMBOLS.md`, now at
`docs/re/RE_SYMBOLS.md`
Verified: every rewritten reproduction command was extracted from the doc and
run as written against the regenerated database — all 5 exit 0 and return the
disassembly they describe. The README's function count (25 676) is the database's
own. `cargo fmt --all -- --check` clean.
Not changed, deliberately: the README's Oracle bullet says Canary's native Linux
ELF "crashes / does not run". That is about Canary, not xenia-rs, and it conflicts
with a July note that the native build works — unverified either way here.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The generator had not been able to run correctly since the manual moved into
`tools/ppc-manual/`: it computed the repository root as `HERE.parent.parent`,
which now names `tools/`, so the XML, Canary's emitters and xenia-rs all stopped
resolving — silently, because both scrapers skipped what they could not find.
Every page's references had been pointing at paths that exist nowhere.
What each source contributed, measured on the 350 pages before this change:
Operation (pseudocode) 251 pages: fixed boilerplate "derives from the xenia-rs
interpreter"; 99 carry real hand-written seeds
C translation 337 pages: the same kind of boilerplate
xenia-rs snapshot 336 pages: the interpreter arm, pasted in — the only
per-instruction semantics on unseeded pages
links xenia-rs opcode/decoder/interpreter + Canary emitter
Now:
* semantics come from **Xenia Canary**, the reference emulator, read through
`git show` at a pinned upstream commit (`origin/canary_experimental`,
f21ebd49e9). Not our checkout: it carries instrumentation and lacked
upstream's `mcrf` fix, so it would have published probes and a wrong `mcrf`.
Each page embeds the emitter (`InstrEmit_<mnem>`), and for the 128 pure
one-line delegations also the helper that holds the semantics.
* decode references point at `crates/sylpheed-ppc` — the decoder that
produces `sylpheed.db` — as in-repo relative links.
* the boilerplate now says what is true, and the C translation guide maps
Canary's actual HIR calls, checked against `ppc_hir_builder.h` (including
that `UpdateCR(n, v)` truncates to 32 bits).
* `rust_scraper.py` -> `decoder_scraper.py` (interpreter half dropped);
missing sources are now errors, not empty results.
Verified:
consistency checks 455 XML entries, 350 families, 598 index keys
hand-written tails 386/386 byte-identical after regeneration
xenia-rs in generated 0
pages with a snapshot 349/350 (was 336) — `dcbi` has no Canary emitter at all
in-repo decoder links 910/910 resolve to a line holding the identifier
emitter boundaries brace counter == column-0 `}` rule on 521/521;
preprocessor model unit-tested (#if 0/#else/#elif)
idempotency re-run: 0 pages updated, 0 working-tree changes
Hand-written notes (outside the generated regions) are not rewritten here:
* 110 links into `../../xenia-rs/...` were dead; they now point at the file in
the archived repository (git.mc02.dev/fabi/xenia-rs @ 8401d4d). Line anchors
were dropped because the notes predate that commit — 0 of 441 old line
ranges match it — and a precise-looking wrong anchor is worse than none. The
link text, which carries the author's line numbers, is unchanged.
* 140 prose claims about xenia-rs's behaviour remain. 23 are verified to hold
for Canary too (the 32-bit CR0 truncation, OE left unimplemented); the other
114 need checking one by one, and some invert — e.g. `divdx` notes a correct
64-bit CR0 update in xenia-rs where Canary's `UpdateCR` truncates. Left for
a deliberate pass rather than a blind substitution.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`SYLPHEED_DB` is the documented contract for the static-analysis database —
`docker/decoder/sylph-decoder` sets it, and `docs/agents/CONTAINER-NOTES.md` and
`decoder-loop.md` list it — but no reader honoured it:
isl_cmdtab.py, name_block_bases.py hardcoded `/work/xenia-rs/sylpheed.db`, a
path that has not existed anywhere since
`/work` became a clone. Both failed on
every machine, before and after the
database moved.
zq.py used `$SYLPH_XEXDB`, a name invented in
#39 without grepping for the existing one.
All three now resolve `$SYLPHEED_DB`, then `$SYLPH_XEXDB` (kept as an alias so
nothing already written against it breaks), then `<repo root>/sylpheed.db`, and
refuse with exit 1 naming both variables and the build command otherwise.
`grab_tutorial.sh` hardcoded its helpers into the retired `sylpheed-reborn`
checkout. That copy's `skip_intro.sh` still calls the removed `vgamepad` and
exits 0 having pressed nothing — the failure this repository's own copy was
rewritten to make loud. So the script was already running a silently broken
helper; it now resolves its helpers from its own directory. Nothing exists only
in reborn's `tools/re-capture` (checked: 0 reborn-only files).
Verified against the same database, output compared byte for byte with the
ORIGINAL scripts (path-substituted copies, sibling imports resolvable):
resolution path isl_cmdtab name_block_bases
default (root) identical identical
$SYLPHEED_DB identical identical
$SYLPH_XEXDB identical identical
bad path exit 1 exit 1 (zq.py: exit 1 too)
and `isl_cmdtab`'s output is byte-identical to the body of the committed
`docs/re/data/isl-command-table.txt`.
⚠️ `name_block_bases`'s output does NOT reproduce the committed
`docs/re/data/name-block-bases.txt` (2,609 lines differ, `strings in the image:
7366` vs `7140`). That is the database, not this change: the artefact was
generated from the agent box's older 586 MB database, and this machine's is the
current generator's. Two databases are in circulation. Not regenerated here.
⚠️ Also not fixed, because it is a different bug: `name_block_bases.py` globs
`/work/sylph_extract/**/*.pak`, another path that exists nowhere now; that part
of its report is silently empty. It should read `$SYLPHEED_DISC`.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`the_derived_order_puts_primitives_last_and_that_wipes_screens` asserted that
switching primitives on still flattened at least one screen (`wiped_on > 0`),
and said to turn `include_primitives` on by default the day it stopped.
`cfcda55` made that day arrive — a keyless primitive that would hide the screen
is now forced to paint first — and the wire had been red since 2026-08-29,
invisible to CI because the runner has no corpus and the suite self-skips.
#38 chose option A: keep the default off, and pin the fix instead. The flatness
proxy stood in for "paint order is solved", and the fix's own record says that
is only partly true — "This does **not** make `include_primitives` safe by
default" (docs/re/structures/ui-prm-primitives.md). So:
* renamed to `no_build_is_wiped_with_primitives_on` — the old name asserted
the bug; it survives in the doc comment for anyone searching from #38;
* `wiped_on > 0` -> `assert_eq!(wiped_on, 0)`, with a message saying what a
failure now means (`forced_backdrop` regressed) and that it must not be
"fixed" by turning primitives off in the test;
* `include_primitives` stays `false`; no library code changes.
It is the only end-to-end check of the rule: the four tests in
`ui_forced_backdrop_disc.rs` pin its mechanics, and none composes with
primitives on.
Verified with the corpus present, including that it can fail:
fix 3 passed 125 builds draw a primitive, 0 wiped
forced_backdrop disabled FAILED 36 of 125 wiped — the new message fired
restored 3 passed 125 builds, 0 wiped
The control's 36 is exactly the "36 builds ... wiped by our own sort" that
`cfcda55` and `ui-forced-backdrop.md` report — the guard independently
reproduces the fix's own number, so it measures precisely what was repaired.
The control was a one-line `return false;` in a scratch copy; the working tree
was restored and `git diff` showed only this test file before committing.
Closes#38
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
