Shuffleboard abstraction #31

ky28059 · 2022-03-29T07:26:37Z

This PR introduces convenience API wrappers around shuffleboard for ease of use. To add a simple entry,

ShuffleboardTab tab = Shuffleboard.getTab(tab);
GRTNetworkTableEntry entry = new GRTNetworkTableEntry(shuffleboardTab.add(name, value).getEntry());

becomes

GRTShuffleboardTab tab = new GRTShuffleboardTab(tab);
GRTNetworkTableEntry entry = shuffleboardTab.addEntry(name, value);

To add a listener,

ShuffleboardTab tab = Shuffleboard.getTab(tab);
shuffleboardTab.add(name, value).getEntry().addListener(this::method, EntryListenerFlags.kUpdate);

becomes

GRTShuffleboardTab tab = new GRTShuffleboardTab(tab);
shuffleboardTab.addListener(name, value, this::method);

or

shuffleboardTab.addListener(name, value, this::method, EntryListenerFlags.kNew | EntryListenerFlags.kUpdate | EntryListenerFlags.kImmediate);

if custom flags need to be set. .addListener() can also be chained to add multiple listener entries at once. There is also .addToggle(), which functions the same as .addListener() except exists specifically for boolean listeners to wrap the displaying of a fancy toggle switch:

shuffleboardTab.add(name, value).withWidget(BuiltInWidgets.kToggleSwitch).getEntry().addListener(this::method, EntryListenerFlags.kUpdate);
// becomes
shuffleboardTab.addToggle(name, value, this::method);

To set the position of entries from code,

ShuffleboardTab tab = Shuffleboard.getTab(tab);
GRTNetworkTableEntry entry = new GRTNetworkTableEntry(shuffleboardTab.add(name, value).withPosition(col, row).getEntry());

becomes

GRTShuffleboardTab tab = new GRTShuffleboardTab(tab);
GRTNetworkTableEntry entry = shuffleboardTab.addEntry(name, value, col, row);
shuffleboardTab.addListener(name, value, this::method, col, row);

While addEntry(), addListener, and addToggle() support the extra position arguments, it is not required and omitting position arguments will simply cause the entry to placed in the first available location when shuffleboard is populated.

ky28059 · 2022-03-29T07:29:31Z

Todos:

Make .list() support position and size setting
Consider whether it's worth it for entries to support size setting (probably not, since we don't really need to resize anything past 1x1 and it makes the method signatures even more convoluted)
Add .grid()?
Make .list() take in an optional parameter to control the LABEL_POSITION layout property?

ky28059 · 2022-03-30T07:31:37Z

Positioning GRTNetworkTableEntrys has changed a bit: it is now (similar to lists)

shuffleboardTab.addEntry(name, val).at(col, row);

for better composability and less constructor clutter (especially paired with .withSize()).

// Without these methods, GRTNetworkTableEntry would have to have a 2, 4, and 6 value constructor, as well
// as a static method most likely to handle the case where size is provided but position is not.
shuffleboardTab.addEntry(name, val).at(col, row).withSize(2, 2);

Speaking of lists, I should document those at some point. Also note that listeners, due to their structure, have not changed in how positions must be supplied in the method signature; only .addEntry().

e3l · 2022-03-31T04:14:57Z

src/main/java/frc/robot/commands/tank/PlebAutonSequence.java

+                }
+                forceTimer.reset();
+                internalSubsystem.requestShot();
+                shotsCompleted++;


how about 'shotsRequested'?

e3l · 2022-03-31T04:15:43Z

src/main/java/frc/robot/commands/tank/PlebAutonSequence.java

-                finalTimer.start();
-                complete = true;
+            // Force a shot if we haven't shot in 4 seconds
+            if (forceTimer.hasElapsed(4)) {


8 seconds driving + 8 seconds of forcing for two balls is just too much time in a worst case scenario

Perhaps the 8 seconds of driving time is the issue? I kinda don't want to decrease the force timer out of fear that the first shot may be forced out too early.

I think shooting two balls at once fixes this

e3l · 2022-03-31T04:16:59Z

src/main/java/frc/robot/jetson/JetsonConnection.java

-         * @return The distance from the camera to the hub.
-         */
-        public double getHubDistance() { 
+        public double getHubDistance() {
            return hubDistance; 
        }


it'd probably be clean to return a pair of (distance, theta) and automatically include setting consumed to true as part of that method

…equence changes

…GRTCommandBased into shuffleboard-abstraction

ky28059 · 2022-04-01T22:52:14Z

src/main/java/frc/robot/subsystems/internals/InternalSubsystem.java

@@ -204,9 +204,21 @@ public void periodic() {
            // Spin the top motor on a timer
            exitTimer.start();
            motorTop.set(0.5);
-            stagingExitBall = true;


So the way the two ball shot delay was fixed was to remove the final ball state entirely? Perhaps just commenting out the !stagingExitBall check on line 166 would be a better solution which still represents the ball count semi-accurately?

ky28059 · 2022-04-01T23:05:02Z

src/main/java/frc/robot/subsystems/internals/InternalSubsystem.java

@@ -215,18 +227,18 @@ public void periodic() {

            // Reset states
            // If the only ball in the system is the one we just shot, mark the shot as completed
-            if (ballCount <= 1) shotRequested = false;
+            if (ballCount < 1) shotRequested = false;
            rejectingChecked = false;


This is a little more complicated. If we skip the stagingExitBall timer and don't wait the 0.5 seconds, rejection logic won't check the color of the second ball. If the two balls are of different colors, it will shoot both either rejecting or shooting (which is bad in both cases).

Perhaps a solution is to only skip the stagingExitBall check if the balls are of the same color? Might be a bit messy to implement

e3l · 2022-04-05T20:57:05Z

src/main/java/frc/robot/RobotContainer.java

+        // If setting initial position manually, start it at a position assuming we are facing the hub
+        // at a distance `hubDist` inches and 0 on the y axis.
+        // TODO: is this worth having be a flag at all?
+        if (MANUAL_START_POS) {


i feel like we need a starting position anyways, we might as well make it something like 70, 0 instead of 0, 0 and keep this flag on. Cuz auto will overwrite it anyways

e3l · 2022-04-05T21:04:04Z

src/main/java/frc/robot/subsystems/ClimbSubsystem.java

        }
    }

+    @Override
+    public void periodic() {
+        sixPosEntry.setValue(sixEncoder.getPosition());


i feel like setting motor powers should go in periodic. this means we have a state with the desired power, and in periodic we decide what to do with it

e3l · 2022-04-05T21:05:26Z

src/main/java/frc/robot/subsystems/IntakeSubsystem.java

-    private static final double DELAY_LIMIT_RESET = 0.2;
-    private Double switchPressed = 0.0;
+    private double driverPower = 0;
+    private boolean driverOverride = false;


what if we assumed that if driverpower isn't 0 then we want to override

i really don't like having all these powers and overrides

That works, but it means that we can't override with 0 power (something that RunIntakeCommand does for 2 seconds after driver input is 0). I guess we can remove driverOverride entirely since we're not using the intake camera to detect balls at all.

e3l · 2022-04-05T21:09:22Z

src/main/java/frc/robot/subsystems/TurretSubsystem.java


-            // System.out.println("JETSON r: " + r + " theta: " + theta);
+            // Reset offsets when we refresh rtheta from vision.
+            resetOffsets();


idt we should reset offsets, we should try not to and see how driver practice goes. Because ideally offset accounts for bad interpolation tuning, which will persist even if new vision data comes in

Hmm, I thought offsets were for rtheta drifting which would be reset when vision sends data?

e3l · 2022-04-05T21:11:00Z

src/main/java/frc/robot/subsystems/TurretSubsystem.java

-        shuffleboardTurntablePosEntry.setValue(Math.toDegrees(turntableEncoder.getPosition()));
-        shuffleboardFlywheelVeloEntry.setValue(flywheelEncoder.getVelocity());
-        shuffleboardHoodPosEntry.setValue(Math.toDegrees(hood.getSelectedSensorPosition() / HOOD_RADIANS_TO_TICKS));
+        turntablePosEntry.setValue(Math.toDegrees(turntableEncoder.getPosition()));


can we have all the setvalues together at the end

ky28059 · 2022-04-09T06:01:01Z

Merged into #32 .

GRTShuffleboardLayout, .list(), positions in code

9d0ba93

ky28059 added 4 commits March 29, 2022 00:37

Position and size methods for shuffleboard layout

fa5b47e

Internals debugging and networktables

4bc05fd

Move .addEntry() to list-like .at() position syntax

ff0ca1a

Auton chooser, split up localize coordinate method

e329d83

ky28059 marked this pull request as ready for review March 30, 2022 07:28

ky28059 and others added 4 commits March 30, 2022 12:17

2 ball auto + more debug networktables

8794024

Run intake in DeployIntakeCommand, PlebAutonSequence restructuring

d9d2fba

Put hood and flywheel refs on shuffleboard, reset offsets on vision data

5a01184

internals and turret tweaks

e9aef3a

e3l reviewed Mar 31, 2022

View reviewed changes

ky28059 and others added 6 commits March 30, 2022 23:43

Separate driver and auto intake override, partially revert PlebAutonS…

6108af2

…equence changes

Merge branch 'shuffleboard-abstraction' of https://github.com/grt192/…

46639fc

…GRTCommandBased into shuffleboard-abstraction

Jetson data pairing

0a7c144

Shoot 2 balls at once, rtheta feedforward, turntable tolerance fix

1c7d9a7

tweaked and tested shooter interpolation table

4f383e3

Two ball fixes

0bba935

ky28059 commented Apr 1, 2022

View reviewed changes

ky28059 and others added 6 commits April 1, 2022 16:15

Only shoot two balls without delay when they are the same color

58db64d

.widget(), better internals shuffleboard updating

d94c188

Intake shuffleboard organization

e3023e4

Climb shuffleboard organization

207fa74

Intake top limit switch + reset abstraction

827e399

Preliminary intake deploy tuning

2c64a8c

e3l requested changes Apr 5, 2022

View reviewed changes

ky28059 added 2 commits April 5, 2022 21:30

Intake tuning, shuffleboard manual zero

e53ea54

Turret setValue() organization

76f28f3

Don't force if the turret is retracted

39fa9b6

ky28059 closed this Apr 9, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Shuffleboard abstraction #31

Shuffleboard abstraction #31

ky28059 commented Mar 29, 2022

ky28059 commented Mar 29, 2022 •

edited

Loading

ky28059 commented Mar 30, 2022 •

edited

Loading

e3l Mar 31, 2022

e3l Mar 31, 2022

ky28059 Mar 31, 2022

ky28059 Apr 1, 2022

e3l Mar 31, 2022

ky28059 Apr 1, 2022

ky28059 Apr 1, 2022

ky28059 Apr 1, 2022

e3l Apr 5, 2022

e3l Apr 5, 2022

e3l Apr 5, 2022

e3l Apr 5, 2022

ky28059 Apr 5, 2022

e3l Apr 5, 2022

ky28059 Apr 5, 2022

e3l Apr 5, 2022

ky28059 commented Apr 9, 2022

Shuffleboard abstraction #31

Shuffleboard abstraction #31

Conversation

ky28059 commented Mar 29, 2022

ky28059 commented Mar 29, 2022 • edited Loading

Todos:

ky28059 commented Mar 30, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ky28059 commented Apr 9, 2022

ky28059 commented Mar 29, 2022 •

edited

Loading

ky28059 commented Mar 30, 2022 •

edited

Loading