Report rcl/rclc errors

[jimmy-mcelwain]

Fix for #128. RCL(C) API errors get reported to the user now. It just notifies the user that an RCL(C) API error occurred and gives the integer value of the rcl_ret_t value associated with that error that the user has to look up themselves.

motoRos_RCLAssertOK_withMsg() checks that the rcl_ret_t value that is passed in is OK, and if it is not, it not only sets an alarm indicating that, but it also sets an alarm indicating where it went wrong, like motoRos_Assert_withMsg() does. It's a bit ugly, something like what I described in the last paragraph of this comment may be prettier/do a better job of separating functionality/not duplicating code, but that would increase complexity a bit. It would require either a global "problems" struct, or several local "problems" structs in different files that get passed around that the hypothetical RaiseAlarms() function could reference.

This is a draft for now because I would like advice on what to do with motoRos_RCLAssertOK_withMsg(). I can squash the commits afterwards and rebase on main.

[gavanderhoorn]

It just notifies the user that an RCL(C) API error occurred and gives the integer value of the rcl_ret_t value associated with that error that the user has to look up themselves.

I would phrase it slightly more positive: users would lookup alarms raised by MotoROS2 in the troubleshooting documentation already, so this is just one more to look up.

It's just that in this case, it's as-if there's an additional sub code (a sub-sub code?), which happens to correspond to the rcl_ret_t value.

[jimmy-mcelwain]

Yeah that's a good point