The gyro class no longer attempts to set a default deadband, but it still
has an optional SetDeadband() method.
The gyro integration tests were modified and still pass consistently.
Change-Id: I08a97b00b98b49b0a3c63306fcc809857523af2b
TestEnvironment::SetUp() will only initialize the first time it's called.
This allows the --gtest_repeat flag to be used to automatically repeat
the tests.
Change-Id: I20c857a37a88f48114d74ae68518d4a9d724d012
During calibration, the Gyro class sets the accumulator deadband to contain
whatever the farthest sample from the center was. The integration test
passes now.
A SetDeadBand method was added to the Gyro class for teams to set their own
deadbands.
Change-Id: Idbe4c279e2991b4daed4d4cf3bfaf605d4ee25c0
Timer::Get now compensates for the FPGA time rolling over after 71 minutes
UltraSonic::Ping doesn't bother disabling automatic mode, since it asserts
that it's not in automatic mode on the line before.
Change-Id: I6b0f45327c453abd8a846ec8da0f9676e210d909
SmartDashboard does usage reporting now (or will when it's implemented
in the HAL).
Global errors are raised in C++ when problems happen, since there is
no SmartDashboard instance. Previously, no error reporting was done
at all.
GetData was uncommented.
Change-Id: I3331eb9f09924d1d0028e3fa041f0cf68caa5cf5
Analog interrupts now work in C++.
The interrupts Resource was moved from a global in DigitalInput
to a static member of SensorBase.
An analog interrupt IT was added, and the digital interrupt one modified
to prevent a linker error.
Change-Id: I9a300daafed15e9666a4ccb405a509615e3dbb06
Modules aren't used anymore in wpilibc and wpilibj, so the hal functions
that references them and and JNI bindings for these functions have been
pulled out.
Both Counter classes were also modified because they still referenced
modules.
Change-Id: Ic01feb145a4ed5f08cd55f140867c721f5ee7b10
Analog triggers now work in Java.
Integration tests for analog triggers are included.
A message in the C++ analog trigger IT was fixed.
Change-Id: I50007c6901b8391d32c0e81becdbe18e48a7840f
Analog accumulators now wait for the amount of time a full sample
(including oversampling and averaging) lasts after
AnalogInput::ResetAccumulator() is called, so they don't return
old values after being reset.
This delay should be microseconds long and will only happen
when an accumulator is reset.
A new test is is the C++ TiltPanCameraTest that tests this behavior
with the Gyro class.
Change-Id: I1b3ffdeec187959f95c5e637a6d428c9a4bc2cf4
The LiveWindow singleton instance shouldn't be a global static
variable, since the order that global statics are constructed is
undefined, and it's required by other constructors.
Change-Id: I2edccc1f723f0ea41b1347379b3e3778a50afcdc
Some tests are more explicitly named now, and negative values are
tested with both voltage and current control.
Change-Id: I316ccfb7670a341e2f13f4ed3a514f283625409e
Previously, negative output voltages were unpacked incorrectly.
Now, they're requested as percentages, unpacked, and then scaled
in software based on the bus voltage. The output voltage and
bus voltage shouldn't ever be out of sync, since they're in
the same message.
Change-Id: I745fffd0faa6cbaab967240271e6cfa8653212e2
You can get the state of the USER button with GetUserButton() in
C++ or Utility.getUserButton() in java.
Change-Id: I923e62cab5e621ef43fed503acab5c0d751264fb
Calling CANJaguar::Disable() now makes NetComms stop periodic
setpoint messages, so the motor actually stops until EnableControl()
is called again.
Change-Id: Ib4eb4ad5f729be5c74e799f02ed000511de1e03d
When an error is reported or an assertion fails in C++, a line is now
printed with information about where the error occured, and a stack trace
is printed.
The stacktrace isn't implemented in the HAL because it's not
hardware-dependent, so StackTrace.hpp and StackTrace.cpp are gone.
The Eclipse project template is modified to include "-export-dynamic" in
the linker options, which is necessary for stack traces.
Change-Id: Ie86c14185b13ed603d0fe6467e87ba4f731b1913
The file name was never updated from the old path, a deadlock occured
sometimes. A "resource not found" message was also set when fopen()
returns an error, which should be an errno message because it's a C
library error.
Change-Id: Ic913a08f6f5d73219cb6625198f5a4519c039956
The CANJaguar class now stores a controlEnabled flag. When disabled, it
won't verify the control mode, which involves sending enable messages.
Change-Id: I8baa48eec31de6b4d3fee2b5a074320542a1bbef
A PDP channel number is correct now, the deploy script was changed to
kill Java programs before running and ignore useless messages,
the "Waiting for enable" message is only printed once, and the accelerometer
test is more robust.
Change-Id: I2226140d8c3e44c452e039c27f4f1cf11c952c42
This previous caused a bug where the Jaguar would sometimes
behave as if it were in the wrong mode, even when everything was
verfied.
Change-Id: Id5b96e0f7e64066eaaa0e5be4ba53fca76ba0703
