Skip to content

FateWeaver

Maven Central Version

FateWeaver is a binary data logger for FTC robots. It writes one log file per OpMode, and AdvantageScope opens the file directly.

You log a value by name with Fate.log. There is nothing to create first, and every value is stamped with the time you logged it. A path keeps the type of its first value. If you later log a different type to it, FateWeaver warns once and skips that value instead of garbling the plot. Your own classes are logged by reflection, or with a schema you write when you want control over the bytes.

Quick start

import gay.zharel.fateweaver.telemetry.Fate
import gay.zharel.fateweaver.telemetry.log

class MyOpMode : OpMode() {
    private lateinit var drive: MecanumDrive

    override fun init() {
        drive = MecanumDrive(hardwareMap, Pose2d(0.0, 0.0, 0.0))
    }

    override fun loop() {
        drive.updatePoseEstimate()

        Fate.log("Robot/Pose", drive.pose)
    }
}

import gay.zharel.fateweaver.telemetry.Fate;

public class MyOpMode extends OpMode {
    private MecanumDrive drive;

    @Override
    public void init() {
        drive = new MecanumDrive(hardwareMap, new Pose2d(0.0, 0.0, 0.0));
    }

    @Override
    public void loop() {
        drive.updatePoseEstimate();

        Fate.log("Robot/Pose", drive.getPose());
    }
}

To group related values, ask for a table once and log into it. For example, val drive = Fate.table("Drive") followed by drive.log("leftVelocity", v) logs to Drive/leftVelocity. Slashes in a path become folders in AdvantageScope.

The schema system

Most of the time you never write a schema. FateWeaver infers one from the value you log:

  • Int, Long, Double, String, Boolean
  • Arrays, lists, sets and maps
  • Enums, encoded as a 4-byte ordinal
  • Your own classes, by reflection over their public properties
  • Classes with an AS_TYPE constant, which get that name recorded in the log

Writing a schema by hand

When you care about the exact bytes, write the schema and pass it as the third argument:

val optimizedSchema = CustomStructSchema<SensorReading>(
    type = "OptimizedSensor",
    componentNames = listOf("time", "value", "sensor"),
    componentSchemas = listOf(LongSchema, DoubleSchema, StringSchema),
    encoder = { reading ->
        listOf(
            reading.timestamp / 1000000,  // convert to milliseconds
            String.format("%.3f", reading.value).toDouble(),  // round precision
            reading.sensorId.take(8)  // truncate ID
        )
    }
)

Fate.log("Sensors/Optimized", reading, optimizedSchema)

The first log to a path fixes its schema. Later calls reuse it.

The schema registry

Register a schema once and every later path that logs that class picks it up:

// Register once during initialization
FateSchema.registerSchema<Pose2d>(customPoseSchema)

// Every Pose2d you log from now on uses it
Fate.log("Robot/Pose", drive.pose)

Where to next

  1. Installation adds FateWeaver to your FTC project.
  2. Getting started walks through your first logging OpMode.
  3. Schema system explains how types become bytes.
  4. Log management covers downloading and viewing your data.

FateWeaver is developed and maintained by the FTC community for the FTC community.