Steam Input's per-game remapping doesn't reach a 2D app; device.steam_gamepad's options now change the buttons
inside the app: the launcher exports FRAMEPORT_GAMEPAD_LAYOUT/_KEYS, the agent's Podman wrapper mounts a per-game
key layout (agent v103). Diagnostics of such games carry gamepad.log and Steam's controller activation lines. The
game page shows a patch's on/off options under it while it's on.
CI: locales template refreshed; frame.main_process counts as a patch that works on the 32-bit bridge.
Halo CE VR's HaloActivity runs in :halo_game; Lepton starts it directly and watches only the main process, which
never exists, so it closed the game about a minute in ('Target app ... did not start'). Analysis field
separate_process (ANALYSIS_VERSION 23), default-on patch, triage app-separate-process, PLAYBOOK row.
SynthRiders is analysed as 'Vulkan (Unity boot.config)' but renders with GLES on Zink, so the dump was hidden in
Game settings and session triage's suggestion was dropped; the dump setting alone also never added the layer.
Games installed from an install link get a row on the game page (where it came from, the installed version) with Check for updates, and the same entry in the page's menu: it asks the link now instead of waiting for the 6-hourly check. Manifests that named no version (e.g. an app's first release) were never checked; the installed APK's own versionName (from the analysis) is now the version to compare with, in the automatic check, the button and frameport link-updates.
Updating a game from its install link downloads the new APK into its own folder; the scan then finds it as another version of a library game (#351 relinking), switches the entry to it and returns nothing, and add_from_link reported that as an unreadable APK, so the update stopped before the link version was recorded. add_from_link now takes the entry from the scan report; if the scan left the library's copy (still on disk), the link switches the game to the download like Update with a newer APK.
msm reported SteamVR's compositor engine time near 2^64, rising ~1000x faster than real time, so the summed GPU busy
was pinned at 100 % on an idle Frame. Each render fd is now tracked on its own; a delta longer than the interval
falls back to drm-cycles-gpu / GPU clock. Checked on the dev Frame: installed agent 100 %, this one 0.0 % at 231 MHz.
Flet 1.0 keeps toasts in the same stack as dialogs and pop_dialog() closes the topmost open entry. An install link's
job shows a "waiting for your answer" toast just as its question appears; when the toast came second, "Download
and install" closed the toast, the download ran, and the question stayed on screen. pop_dialog now skips toasts
(reproduced and checked with real clicks in the web build). Also a too-long test docstring line.
DETACHED_PROCESS left the restarted FramePort without a console, so every console program it started (ssh, git) opened
its own visible terminal, and the Frame connection's ssh kept one open. CREATE_NO_WINDOW gives it a hidden console,
as the dev shortcut does.
HardBullet's loading frames submit only a cube layer, which the Frame's runtime doesn't composite, so the loading
screen stayed black. Adapter setting cube_quads (default on) draws a cube-only frame's six faces as quad layers
around the viewer; frames with a projection layer are unchanged. Adapter rebuilt on current main.
Also: the foveation test expected a numeric FOVE_LEVEL to pass through (Valve's layer ignores numbers since #394),
a long description line and an import order in adapter_dialog.py failed lint.
The install page shows a manifest's name, description and icon only when the manifest's host allows CORS. GitHub file pages (github.com/<o>/<r>/blob|raw/<ref>/<path>) are now turned into their raw.githubusercontent.com address by the page (preview and the frameport:// link it hands over) and by deeplink.parse (a /blob/ page is HTML, so FramePort couldn't install from it). Release downloads can't be read by any web page: the page shows the release's name and the repo's description from GitHub's API instead of 'A game or app'. INSTALL_BUTTON.md says where to host a manifest.
Transformers Beyond Reality: Redux ended its first frames with Unreal's "Shader compilation failures are
Fatal." With Lepton's Fossilize layer loaded every graphics pipeline failed with
VK_ERROR_OUT_OF_HOST_MEMORY, even a constant-colour one; the pipeline cache, the app's CPU allocator and
Valve's FDM/RPO layers were ruled out on the device. Without Fossilize a few multiview vertex shaders
still fail in Turnip.
- adapter.vk_no_fossilize: the shim removes VK_LAYER_fossilize from GraphicsEnv's debug layer list when
the engine loads it (library constructor, same GraphicsEnv calls as zinkfix)
- adapter.vk_pipeline_fallback: a refused graphics pipeline is created again with a blank fragment
shader, then blank vertex + fragment shaders
- catalog entry (status unknown, headless only), triage vulkan-pipeline-refused, PLAYBOOK, DEVICE_NOTES
Headless on the dev Frame: RUNNING after 78 s at 72 fps (was: exit within 21 s). Headset check pending.
Settings -> Data picks the PC backup folder (e.g. a cloud-sync folder; old folders stay listed) and turns on
automatic snapshots (off by default): Quest launchers snapshot the saves when the game closes, under the launch lock,
keeping the newest three changed ones on the Frame, and FramePort imports them as automatic backups when it connects.
Agent v101. Checked on the Frame with BlazeRush (snapshot per close, one import per snapshot).
The cable wait ran ipconfig every 2 s without CREATE_NO_WINDOW, so the windowed Windows app flashed a console each
time, and it only tried the Frame once ipconfig listed this PC's cable address, which Windows can report differently.
Commands now run hidden and the wait connects as soon as the Frame's SSH answers over the cable.
The old radius setting exported FOVE_INNER/FOVE_OUTER, which Lepton doesn't pass into the container and Valve's
foveation layer doesn't read, so every size looked the same. Game settings now offers Default (large) / Large /
Medium / Small, exported as FOVE_LEVEL=lo/med/hi (the layer's own preset names); saved radius choices map to the
nearest preset. The agent's launcher update accepts FOVE_LEVEL.
The validator required at least one rewritten mmap call, so an Unreal build without the Binned3 reservation
pattern (Lies Beneath) failed to build with "0 rewritten, 0 left" once the patch was on.
On the Frame the atomic-branch detours never installed ("installation failed"): an mmap hint isn't honoured,
so the stub landed out of B range. Stubs now go into free gaps from /proc/self/maps within +-128 MiB of the
entry, mapped with MAP_FIXED_NOREPLACE (a relocated result is rejected); no gap = no hook. Every failure logs
the target and the failing step.
The detour helper from PR #34 (used by the metadata hook and PR #389's two geometry hooks) wrote a 16-byte
branch + literal in two steps while game threads may run the function (installation is retried from
xrPollEvent). Each hook now publishes a single aligned release-atomic AArch64 B to a stub within +-128 MiB;
the trampoline replays the one verified instruction and returns to entry+4. Every mprotect is checked;
failures before publication free the stub and trampoline and leave the function untouched; no stub in range =
no hook. Adapter artifacts rebuilt by us (the PR's binary isn't shipped).
Quest 3 is the default for every Quest game, but the dispatcher patch only knows OVRPort 3.4.3-aa54c3f: an
APK converted with another runtime (found with an already converted Batman copy) failed to build. It now
keeps OVRPort's Quest 2 answers and says so in the build checks.
cmd_shortcuts returns without "started" when it queues (later) or skips the change; cmd_uninstall indexed it
and failed after the shortcut removal was already queued, leaving the game installed. Found on the Frame.
Installing on the Frame restarted Steam to add the shortcuts, which also closed FramePort running there.
Settings -> Installing gets a Steam library mode: now (as before; restarts also wait while FramePort runs on
the Frame), later (no restart: Steam's next start re-applies and verifies the shortcuts through bounded
steam.service hooks) or off (no shortcuts; Play starts the game from FramePort). Restart Steam now bypasses
the holds and asks before ending a running game. Agent v99.
Based on the feature branch's two commits (a311498, cf05ce2), merged onto main.
- Files: line-style icons for the drive, PC VR and Linux sections (test_ui_consistency).
- test_pc_target: Quest games now also gain device.headset_model from its migration; PC VR games never do.
- test_surface_projection: frame.adapter is revision 2 since the dormant foveation diagnostic (#365).
Their entries open the install folder rather than Android storage, so they get their own "PC VR games" and
"Linux apps" sections next to Game storage. ui_smoke: linux-files step.
Foveation (Default / Eye-tracked / Fixed / Off, radius, test) moves from Customize to Game settings →
Picture. It's a user preference, not part of the recipe: existing recipes are migrated once and Share
working config leaves it out. Agent v98 set_settings rewrites only the launcher's foveation exports, so a
change applies at the next game start without a rebuild (checked on the Frame: Fixed adds the FDM exports,
Default removes them). The APK-side parts (Batman allocation fix, Climb 2 bridge) stay in the build.
The scale setting's label is now "Render resolution (supersampling)". ui_smoke: game-settings-dialog step.
Each drive the agent reports (SD card, USB, other Steam library drives) is a Files location with its free
space; read-only drives allow downloads only. Installed PC VR and Linux games get their install folder.
Agent v97: drives report read_only explicitly (checked on the Frame).
Games saw a Quest 2: OVRPort's dispatcher hardcodes the system name "Oculus Quest2" and a Quest 2
headset UUID (XR_META_headset_id), and OVRPlugin decides the headset type from the UUID first, then the
name. device.headset_model (Customize; Quest 3 default for every Quest game, also Quest 3S, Quest Pro,
Quest 2) rewrites both in the dispatcher after checking the original bytes; Quest 2 leaves the build
byte-identical. FrameBridge reports the same name to games that call the runtime directly. A library
migration adds the default to existing recipes (installed games show Update on Frame).
Lepton's ro.product.model ("Lepton") can't be overridden by boot properties, so Build.MODEL is unchanged.
Verified on the Frame: OVRPlugin logs "systemName Meta Quest 3" (was "Oculus Quest2").
OutRun Quest VR quits before its VR session with "Rec.709 color space unavailable": the Frame's runtime
has no XR_FB_color_space. Adapter setting fb_color_space (off by default; triage color-space-missing
suggests it) offers the extension when the runtime lacks it, reports Rec.709 and accepts any colour
space without converting. OVRPort's dispatcher can't forward the new functions, so it isn't on for every
game. Not yet tried on the device.
spoopyghosty0