ListenableGlesRenderer2

<SharedAssetsT : Renderer.SharedAssets> ListenableGlesRenderer2(
    surfaceHolder: SurfaceHolder,
    currentUserStyleRepository: CurrentUserStyleRepository,
    watchState: WatchState,
    interactiveDrawModeUpdateDelayMillis: @IntRange(from = 0, to = 60000) Long,
    eglConfigAttribList: IntArray = EGL_CONFIG_ATTRIB_LIST,
    eglSurfaceAttribList: IntArray = EGL_SURFACE_ATTRIB_LIST
)

@WorkerThread
abstract fun createSharedAssetsFuture(): ListenableFuture<SharedAssetsT>

When editing multiple WatchFaceService instances and hence Renderers can exist concurrently (e.g. a headless instance and an interactive instance) and using SharedAssets allows memory to be saved by sharing immutable data (e.g. Bitmaps, shaders, etc...) between them.

To take advantage of SharedAssets, override this method. The constructed SharedAssets are passed into the render as an argument (NB you'll have to cast this to your type). It is safe to make GLES calls within this method.

When all instances using SharedAssets have been closed, SharedAssets.onDestroy will be called.

Note that while SharedAssets are constructed on a background thread, they'll typically be used on the main thread and subsequently destroyed there. The watch face library constructs shared GLES contexts to allow resource sharing between threads.

final suspend fun onBackgroundThreadGlContextCreated(): Unit

Called once a background thread when a new GL context is created on the background thread, before any subsequent calls to render. Note this function is called inside a lambda passed to runBackgroundThreadGlCommands which has synchronized access to the GL context.

If you need to override this method in java, consider using androidx.wear.watchface.ListenableGlesRenderer instead.

final suspend fun onUiThreadGlSurfaceCreated(width: @Px Int, height: @Px Int): Unit

Called when a new GL surface is created on the UiThread, before any subsequent calls to render or in response to SurfaceHolder.Callback.surfaceChanged. Note this function is called inside a lambda passed to runUiThreadGlCommands which has synchronized access to the GL context.

If you need to override this method in java, consider using androidx.wear.watchface.ListenableGlesRenderer instead.

@WorkerThread
fun runBackgroundThreadGlCommands(runnable: Runnable): Unit

Inside of a Mutex this function sets the GL context associated with the WatchFaceService.getBackgroundThreadHandler's looper thread as the current one, executes runnable and finally unsets the GL context.

Access to the GL context this way is necessary because GL contexts are not shared between renderers and there can be multiple watch face instances existing concurrently (e.g. headless and interactive, potentially from different watch faces if an APK contains more than one WatchFaceService).

NB this function is called by the library before running runBackgroundThreadGlCommands so there's no need to use this directly in client code unless you need to make GL calls outside of those methods.

@UiThread
fun runUiThreadGlCommands(runnable: Runnable): Unit

Inside of a Mutex this function sets the UiThread GL context as the current one, executes runnable and finally unsets the GL context.

Access to the GL context this way is necessary because GL contexts are not shared between renderers and there can be multiple watch face instances existing concurrently (e.g. headless and interactive, potentially from different watch faces if an APK contains more than one WatchFaceService).

protected final suspend fun createSharedAssets(): SharedAssetsT

When editing multiple WatchFaceService instances and hence Renderers can exist concurrently (e.g. a headless instance and an interactive instance) and using SharedAssets allows memory to be saved by sharing immutable data (e.g. Bitmaps, shaders, etc...) between them.

To take advantage of SharedAssets, override this method. The constructed SharedAssets are passed into the render as an argument (NB you'll have to cast this to your type). It is safe to make GLES calls within this method.

When all instances using SharedAssets have been closed, SharedAssets.onDestroy will be called.

Note that while SharedAssets are constructed on a background thread, they'll typically be used on the main thread and subsequently destroyed there. The watch face library constructs shared GLES contexts to allow resource sharing between threads.

protected open fun onBackgroundThreadGlContextCreatedFuture(): ListenableFuture<Unit>

Called once a background thread when a new GL context is created on the background thread, before any subsequent calls to render. Note this function is called inside a lambda passed to runBackgroundThreadGlCommands which has synchronized access to the GL context. Note cancellation of the returned future is not supported.

@UiThread
protected open fun onUiThreadGlSurfaceCreatedFuture(width: @Px Int, height: @Px Int): ListenableFuture<Unit>

Called when a new GL surface is created on the UiThread, before any subsequent calls to render and in response to SurfaceHolder.Callback.surfaceChanged. Note this function is called inside a lambda passed to Renderer.GlesRenderer.runUiThreadGlCommands which has synchronized access to the GL context. Note cancellation of the returned future is not supported.