0 Fork 0

Files

T

zwitschi 4d27109668 refactor(tools): group the sixty scripts by what they talk to

One directory with sixty scripts stopped being navigable, and the grouping is also the sandbox boundary:

  Tools/unreal/    commandlets and the two map builders, with ue_common.py and ue_placement.py as the
                   shared imports (they need the engine)
  Tools/blender/   builders, exporters, the FBX preset, the MPFB character pipeline, blender_screenshots
                   (they need the Blender Flatpak)
  Tools/review/    the visual loop: render_screenshots.sh, judge_screenshot.py, hero_compare.py,
                   build_comparison_md.py, e2e_asset_check.py, hero_iterate.sh, visual_loop.py,
                   visual_targets.json
  Tools/gen/       image and texture generation: the ComfyUI driver, deterministic texture writers,
                   reference fetching
  Tools/input/     OS-level input injection
  root             _common.py, check-lfs.sh, addon_compat_probe.py, sibling_asset_audit.py, tests/

Every moved script now resolves the repository root as ../.. and puts the tools root plus its siblings on
sys.path, so 'import ue_common' still works from unreal/. The test suite bootstrap does the same for the five
folders. render_screenshots.sh takes the folder as a third argument to run_cmdlet, and the e2e driver,
hero_iterate.sh and the comparison generator were repointed.

51 README rows now link to their new location, the README gains a Layout section naming each folder's
sandbox, and the architecture notes and Blender README follow the moves.

Verified: 108 tests pass, prettier is clean, judge_screenshot.py and hero_compare.py run from their new
homes, and the e2e run was aborted mid-way (it had been rebuilding an empty DemoMap for 15 minutes) and is
being rerun against the new layout.

2026-11-10 12:40:30 +01:00

3.8 KiB

Raw Permalink Blame History

Controls

Keyboard and mouse. The same list is shown in-game, bottom-right of the frame (AGigLineHUD), and the packaged Linux build ships with these bindings.

Action	Key	Notes
Move forward	`W`	4 m/s walk
Move back	`S`
Strafe right	`D`
Strafe left	`A`
Look	Mouse	Captured on launch; `Escape` releases it
Jump	`Space`
Reset to start	`R`	Returns to the grounded spawn point
Pause / release mouse	`Escape`	Second press resumes and recaptures the mouse
Swap first / third person	`V`	First person is the default; the body is hidden in that view
Show / hide debug overlay	`F3`	FPS, frame time, worst frame, view mode, speed and player position

Notes

Movement is relative to where you are looking (controller yaw), so the capsule stays upright on the sloped street.
The mouse is captured on launch, so looking works without clicking first. Escape toggles the capture as part of pause.
Bindings are Enhanced Input actions in /Game/Input built by Tools/unreal/ue_create_input_assets.py; the mapping context is IMC_Default.

Changes since `0.1.0`

V swaps first and third person (0.2.0, IA_ToggleCamera). The shipped default view is still first person, as the owner decided for 0.1.0; third person adds a collision-tested spring-arm camera and a visible, animated body. Escape still pauses and releases the mouse in either view.

Debug overlay

F3 toggles a top-left readout: smoothed FPS and frame time, the worst frame in a rolling two-second window, the active view (first or third person), player speed, a frame counter, and the player's position in metres from the project origin. The first two seconds of a session are a warm-up window and are excluded from the worst-frame figure, because shader and streaming hitches there made a healthy run read as a 400 ms stall.

It is off in the built demo — a player should not see an FPS counter over Hollywood Boulevard — and on automatically when the process is started with -perf or -qaselftest, so a performance capture never depends on someone pressing a key first. In a running build it can also be forced with the console variable gigline.Debug 1.

Cost: about 5 % of frame time at 1440p in the demo (2.64 → 2.77 ms average), measured in the same self-test route. The performance release check is recorded without the overlay for that reason; the overlay run is kept beside it as Docs/perf/0.2.0-walk-frametimes-overlay.csv.

Review and harness switches

These are command-line switches for the screenshot pipeline and QA runs, not player controls. They exist so a capture is reproducible rather than dependent on a human pressing keys at the right moment.

Switch	Effect
`-thirdperson`	Starts the pawn in the third-person rig, so a capture shows the body without anyone pressing `V`
`-walk`	Holds forward input from the first frame — what makes the stride shot possible
`-qaselftest`	Runs the scripted walk / reset / pause / transect checks and quits; also enables the debug overlay
`-perf`	Enables the debug overlay for a manual frame-time check

The walking shot is driven per shot rather than by the switch: the shot actor enables the walk only while a tracking (ShotFollow) camera is being captured, and returns the pawn to spawn afterwards, so the fixed-camera shots still find it in place.

Testing input without a physical keyboard

Tools/input/input_test.py injects OS-level keyboard/mouse events through a virtual /dev/uinput device (see Tools/input/input_emulation.py), e.g.:

python3 Tools/input/input_test.py --sequence move:2,look:3

It drives the focused game window exactly like real hardware, which is how the controls were regression-tested on the packaged build.

3.8 KiB Raw Permalink Blame History