A simple improvement was to only perform the disable-before-next-set strategy if the caller's request mode is not equal to the current mode.
To keep things simple, SetControlMode was renamed to private method ApplyControlMode so we can still invoke it from c'tor.
Then, the new impl'n of SetControlMode() just calls ApplyControlMode() when caller's request mode is different. That takes care of direct-calls from team source, and indirect calls through enableControl().
Applied to both c++ and java.
Tested in java so far...
Change-Id: I934c06c5339d933918470659acd635e12eb4d113
Java...
added setStatusFrameRateMs() to modify the frame rate for status frames
added missing func that already exists in c++
isFwdLimitSwitchClosed()
isRevLimitSwitchClosed()
getNumberOfQuadIdxRises()
getPinStateQuadA()
getPinStateQuadB()
getPinStateQuadIdx()
added getAnalogInRaw() that doesn't count overflows (for potentiometers).
added setStatusFrameRateMs() to modify the frame rate for status frames
added getBrakeEnableDuringNeutral()
C++...
added GetAnalogInRaw() that doesn't count overflows (for potentiometers).
added SetStatusFrameRateMs() to modify the frame rate for status frames
added GetBrakeEnableDuringNeutral()
added kLimitMode_SrxDisableSwitchInputs to CANSpeedController::LimitMode
Patch set 2: Joe Ross, fixed two javadoc errors
Change-Id: I0bf871e138953de60eeacb547dc359f2125b1327
Increased wait delay to 4ms to cover worst case delays for solicted signal getters.
Specifically....
-Get firmware vers
-Get P,I,D,F gains
-Get IZone, Get CloseLoopRampRate
-Get IAccum (integral accumulator)
Change-Id: I313ea984832cce5182af8e5af5e775837fd54fdc
The other remaining closeloopramprate changes are also merged into this commit, so they may be redundant on gerrit.
Change-Id: Ic3108bb3669e487009b8f52412da3c2f44c42f6f
Change-Id: I4c41b560d879ffc6d8aa4681cad8f75297fe6cda
Fixed some issues with the camera server class
Change-Id: Ifda524b55f84053be004a404a2890905ded7b266
- IMAQdx typedef overrides were being ignored, resulting in incorrect types
(e.g. IMAQdxSession was a long instead of an int).
- Allocated byte buffers byte order was not being set.
- imaqDispose was incorrectly named.
Change-Id: I5d038d45e82755615f0a5bb928defb98f557f93e
The param is capped in the HAL C++ class to [1ms, 95ms] so that zero and negative periods are caped to 1ms, and so that caller can't pass an absurdly large value, which causes TALON is appear disabled.
Change-Id: I4207194be25a33bbd6ad281a75301ce6684659a5
Quite a few functions aren't wrapped, but the most critical ones for
vision should be.
This also fixes a couple of issues:
- nivision_arm.ini (and imaqdx_arm.ini) are now generated without need for
running the output on the RoboRIO.
- enum values are generated even if the value is not directly specified.
Only very basic testing performed to date.
The wrappers are still a bit incomplete (some structures and functions),
but are much more complete than the old ones.
Fixes artf3796.
Changes from initial changeset:
- Use // for comments.
- Add auto-generate notices to the beginning of each generated file.
- Include error number with error text in exception.
- Add free() function to structures.
- Fix out-of-order / non-array enums.
- Avoid duplicate calls to DisposedStruct.write() as .getAddress() does it.
- Refactor OpaqueStruct.
- Default to no null allowed except when overridden.
- Implement unowned return values.
- Add gen_struct_sizer script.
Change-Id: Ie0d102c45817ea8812d98fe4938d1a2255c61664
Adds isEnabled and getSetpoint functions to CANTalon classes.
Sets m_controlEnabled=false in Java if changeControlMode(Disabled) is
called.
Change-Id: I08fd0972df22ad83c5578dd43dd6b3536f3b365b
Added Get/clear routine for IntegralAccumulator
Added missing status check in GetFirmwareVersion(). I don't expect this to affect anything.
JAVA
Renamed getRampRate to getCloseLoopRampRate in java to match the set routines in java, and match all routines in cpp.
Added GetFirmwareVersion to java to match cpp.
Added Get/clear routine for IntegralAccumulator
Retested all three routines in java.
Change-Id: I4ce9d9c87a379b9d4a76aae226e2072876218688
Used to be that if you called Set less than ~20 ms after changing the
mode, potentially unwanted behavior could ensue.
Change-Id: I27cb3603286d8fddd894649787d88c0446b00615
