Astro Bot on AnyPS5: Compatibility & Test Results

No matching reviewed game. Unlisted does not mean incompatible. Browse all tests.
PPSA21564 · Windows · Full 3D title scene and gameplay entry
On Windows 11 with RTX 5080 and Astro Bot 01.007.000, the reporter restored missing /app0/args.txt and /app0/data/ paths: the full 3D title scene rendered and gameplay started without an AnyPS5 code change. The earlier black screen occurred with those assets absent; the separate Windows AMD driver timeout remains an independent result.
Passed the intro into gameplay at roughly 17–19 FPS in normal play; a heavy scene dropped to 7–8 FPS.
Build: astrobot b58df66c + PR #637; warm shader cacheTested: Oct 5, 2026Report: Oct 5, 2026
The branch author links a gameplay capture through the cinematic, Crash Site tutorial, world map and Sky Garden. This is a branch-specific milestone, not a stock-release result.
Build: oneandonlydean/AnyPS5 astrobot branch; capture predates a58477dbTested: Oct 1, 2026Report: Oct 2, 2026
On main 1be0a7ce the title compiles shaders and presents frames, then asserts defaultBusses.size() == 1 after 30–45 seconds. The reporter traced this to an unresolved %ASOBI_ROOT% path under /host that prevents sound/config.xml loading. Locally resolving the mount path passes audio initialization; this behavior has not been established on unmodified main.
Build: main 1be0a7ce (plus an unrelated local fchdir commit); Release; no Astro Bot integration branchReport: Oct 11, 2026
After the earlier sound-configuration stop is removed locally, the title reaches additional GPU/audio initialization and then asserts prop->m_pParent == this. Reporting the port's actual channel count and speaker angles in a separate local A/B test let textures allocate and draw calls continue for 140 seconds; no menu or gameplay was documented.
Build: main 1be0a7ce plus local /host mount resolution for #3480; experimental device-state values tested separatelyReport: Oct 12, 2026
Start-up logos render with sound, but the recorded GPU batch hangs when the first 3D scene begins. The display driver resets after about 3 min 38 s warm-cache or 6 min 40 s cold-cache. Logs end with vkQueueSubmit Vulkan result -4; repeated runs stop at the same scene.
Build: unmodified main d92a8dad3; repeated cold/warm cache runs. Separate diagnostics on main db81992c5 plus PR #4855Report: Oct 15, 2026
The title resolves imports, passes audio initialization, creates save data and connects the controller. The title-screen state then waits for input at a reported 26–31 FPS for minutes, but fullscreen layers remain black apart from a small UI region and thin bar; no visible title menu or gameplay was demonstrated.
Build: main f2d874c1fd; relinked with --windows; the reporter found the same symptom using v0.1.4 release libraries and relinkerTested: Nov 2, 2026Report: Nov 2, 2026
Restoring the missing /app0/args.txt launch settings and /app0/data/ asset path changed the previously black fullscreen image to a fully rendered 3D title scene (cosmos, ship and bots), and gameplay started. The reporter required no AnyPS5 code change; sustained gameplay performance was not measured.
Build: Windows main-based November 2 setup with corrected /app0 asset layout; the November 3 follow-up does not state a final commit SHAReport: Nov 3, 2026
PPSA21564 reports only. PPSA21567 regional tests in the same thread are not merged.
Early patch stack reached first draws but stopped on memory/audio problems. Later Windows gameplay supersedes this stage.
On the same RTX 5080/main build, a locally gated and reverted diagnostic kept pending GPU results; the 1920×1088 render target still contained only zeros. Driver rect-list execution tests passed, so the title's fullscreen draw inputs or state remain under investigation.
7 scoped test reportsLinux · Windows
Independent game compatibility records, source-linked test results, and practical setup help for the open-source AnyPS5 project.