User reported issue: radians incorrect at 9pi, -9pi and after

[oliver-phet]

There’s a bug with the angles on Trig Tour. With Special Angles marked and angle in radians, once you rotate clockwise past -8 pi – 5 pi/6, the angle resets to -8 pi instead of becoming -9 pi. The angles after that are incorrect by – pi.

@amanda-phet when I went to verify, I noticed 9pi has the same issue. But the sim self-corrects at 10pi. At 17pi and 18pi, the radians are off by 1pi, but the sim self-corrects at 19pi. Then at 21, 23, 27, 31, 34, 35, 36, 39. Same in the negative direction: -9, -17, -18, -21...

[amanda-phet]

Yes this seems to be a bug at 9π.

[jessegreenberg]

I was able to observer this, Ill take a look. The numerical value is correct, so it is just about the fraction readout.

[jessegreenberg]

We are going to change this representation in #80, so any changes here may overlap with that one.

[jessegreenberg]

The halfTurnCount in the model is one less than it should be. It is calculated here:

trig-tour/js/trig-tour/model/TrigTourModel.ts

Lines 204 to 209 in 151018b

	// set turn counts and angles
	this._fullTurnCount = Utils.roundSymmetric( ( targetAngle - remainderAngle ) / ( 2 * Math.PI ) );
	remainderAngle = targetAngle % ( Math.PI );
	this._halfTurnCount = Utils.roundSymmetric( ( targetAngle - remainderAngle ) / ( Math.PI ) );
	this.fullAngleProperty.value = this.constrainFullAngle( targetAngle ); // now can trigger angle update
	this.previousAngle = smallAngle;

When it works well, the remainderAngle is 0. But in all the cases where this is broken, the remainderAngle is exactly Math.PI.

[jessegreenberg]

OK, I believe this is fixed. @amanda-phet can you please verify? @oliver-phet thanks for the testing and list in #97 (comment) that helped identify the problem, and I made sure each value you found is working.

[jessegreenberg]

Well, I guess this behavior is going to change with #80 so I maybe this should be on hold.

[jessegreenberg]

#80 is ready for review, is this working as expected after the change in #80?

[amanda-phet]

This is working as expected. Thank you!

@oliver-phet do you want to reply to that user and let them know a fix will be published, and the representation is also changing?

[oliver-phet]

I think it's best to communicate to the user once the fix is published (unless that is a long ways out). Can this issue be kept open and assigned to me once it is published?