i think i was wrong about what was going wrong, i just finally figured out how to get meaningful debug info out of session manager and
env ANBOX_LOG_LEVEL=debug anbox session-manager
[ 2020-01-21 02:29:31] [Renderer.cpp:168@initialize] Using a surfaceless EGL context
[ 2020-01-21 02:29:32] [Renderer.cpp:251@initialize] Successfully initialized EGL
[ 2020-01-21 02:29:32] [service.cpp:41@Service] Successfully acquired DBus service name
I suspect single window mode stays black because the image itself isn't fully booting. In my experience with actual android devices, phones can be in a not fully booted but still adb accessible state. I will run adb logcat, i should have done that earlier:
01-21 02:35:18.776 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:19.777 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:20.778 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:21.778 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:22.779 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:23.780 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:24.280 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:25.281 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:26.282 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:27.283 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:28.283 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:29.284 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:29.784 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:30.785 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:31.786 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:32.787 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:33.788 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:34.788 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:35.288 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:36.290 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:37.291 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:38.292 378 543 I ServiceManager: Waiting for service media.player...
ad infinitum, producing approx 70 of the same log per minute
maybe i will build an image after all, something seems up here, and i don't actually think it's anbox anymore. i think anbox may be working perfectly fine (still testing on my originally successful build, and i still can't reproduce it i wonder if we will have the same logcat result? the suspense is killing me)
env ANBOX_LOG_LEVEL=debug anbox session-manager
[ 2020-01-21 02:29:31] [Renderer.cpp:168@initialize] Using a surfaceless EGL context
[ 2020-01-21 02:29:32] [Renderer.cpp:251@initialize] Successfully initialized EGL
[ 2020-01-21 02:29:32] [service.cpp:41@Service] Successfully acquired DBus service name
I suspect single window mode stays black because the image itself isn't fully booting. In my experience with actual android devices, phones can be in a not fully booted but still adb accessible state. I will run adb logcat, i should have done that earlier:
01-21 02:35:18.776 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:19.777 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:20.778 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:21.778 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:22.779 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:23.780 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:24.280 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:25.281 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:26.282 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:27.283 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:28.283 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:29.284 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:29.784 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:30.785 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:31.786 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:32.787 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:33.788 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:34.788 378 543 W MediaMetadataRetriever: MediaPlayerService not published, waiting...
01-21 02:35:35.288 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:36.290 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:37.291 378 543 I ServiceManager: Waiting for service media.player...
01-21 02:35:38.292 378 543 I ServiceManager: Waiting for service media.player...
ad infinitum, producing approx 70 of the same log per minute
maybe i will build an image after all, something seems up here, and i don't actually think it's anbox anymore. i think anbox may be working perfectly fine (still testing on my originally successful build, and i still can't reproduce it i wonder if we will have the same logcat result? the suspense is killing me)