MechaCat02 147767c49d
Some checks failed
CI / Native — linux (pull_request) Failing after 1h21m4s
CI / WASM — Web (pull_request) Successful in 29m43s
CI / Formatting (pull_request) Successful in 24s
docs(port): the two steps a host needs that the container hides
Both were found by following RUNNING.md on this host from a clean checkout, and
both fail in a way that blames the wrong thing.

`godot --path port` without a prior `--import` produces THIRTY parse errors
naming half the project -- ScreenView, MenuAudio, Gamepad, MenuFlow, ExportTree
all "not declared", each cascading into "cannot infer the type of ..." -- and
then hangs. That reads like a port that does not compile. It is one gitignored
file: `port/.godot/global_script_class_cache.cfg`, which Godot writes at import
time and which `class_name` resolution needs. After `--headless --path port
--import`: zero errors, `main_menu` loads 16 elements and settles at t=56.

The exporter shells out to ffmpeg and ffprobe for the movies AND the audio, and
`sylph-ci:local` has neither -- so the documented `cargo run` fails minutes in,
after the screens are already written and it looks like it is working. The image
to use is `sylph-ci:ffmpeg`.

Neither bites in the agent container: its image carries an imported tree and its
own ffmpeg. That is exactly why neither was ever written down.

Also recorded: `--headless` CANNOT `--capture` (the dummy renderer never emits
`frame_post_draw`; the port detects this and says so), and `xvfb-run`, which its
message suggests, is not installed on every host -- it is not on this one.

Measured 2026-09-26 on fabi-Hyrican-PC: 31 s build, 4 min 11 s export, 149 MB,
`sylpheed-export check` validates all 16 screens against sylpheed.screen/3.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-26 17:56:52 +02:00
2026-09-13 19:31:49 +02:00

Sylpheed

A clean-room reverse engineering and port project for Project Sylpheed: Arc of Deception (Xbox 360, 2007).

Three things live here, in one repository so that a change spanning them lands as one commit:

The decoders crates/sylpheed-formats — the disc's formats, read and verified disc-wide
The port port/ — a Godot 4 project, plus crates/sylpheed-export which converts a disc into the open asset tree it reads
The corpus docs/re/ — what has been reverse engineered, with its evidence, its retractions and its dead ends

You need your own copy of the game. No game content is in this repository and none ever will be. The exporter reads the disc you supply.

The oracle is the real game

sylpheed-cli and the Explorer are tools for verifying our decoding. They are hypotheses under test and they have been wrong. When something must be checked against the truth, the truth is the game running in Xenia Canary, captured — not any renderer of ours.

This is stated first because getting it backwards is the most expensive mistake this project has made.

Layout

crates/
  sylpheed-formats/   the decoders. Disc-wide verified; the corpus is its spec
  sylpheed-cli/       headless tools -- render a screen, dump a table, probe audio
  sylpheed-viewer/    the Explorer: a human's window onto the disc. STATIC data only
  sylpheed-export/    disc -> the open, moddable asset tree
port/                 the Godot 4 project. Reads open formats ONLY
authored/             decisions that are NOT on the disc, each with its reason
data/
  base/               generated by the exporter. Gitignored, never hand-edited
  mods/               drop-in overrides. Yours
docs/
  re/                 the corpus: findings, refutations, method traps
  game/               how the game is navigated -- menus, modals, flight
  port/               the port's mission, its handoff contract, modding rules
                      -- and RUNNING.md, which is how you actually start it
  agents/             how the agent team works together
tools/                capture harnesses, probes, the share tool
exchange/             transient inter-agent files. NOT in git
docker/               the agent containers

Where to start

Xenia Canary is a separate repository: it is a fork tracking upstream, and it carries our instrumentation.

Conventions

Confidence is per claim, never per document: ✅ CONFIRMED · 🟡 PROBABLE · ❔ HYPOTHESIS · ❌ REFUTED. A withdrawn result is kept with its reasoning rather than deleted — that is why the numbers here can be trusted.

Description
No description provided
Readme MIT 820 MiB
Languages
Rust 57.6%
Python 27.7%
Shell 10.6%
GDScript 3.5%
Dockerfile 0.4%
Other 0.1%