Ask Your Question

Trying to use Timesynchronizer...getting compile error

asked 2013-09-25 11:21:39 -0600

rnunziata gravatar image

updated 2013-09-26 15:14:14 -0600

Trying to use Timesynchronizer...getting compile error. Do not know what the error is telling me. Changed pgm to use ConstPtrs still same error:

It seem based on this reference Timesycronizer and other boost type templates are not officially supported by ROS within a class member function. Is my understand correct? If this is true what other interface to these functions is supported? Can the tutorial be updated to show how to use these in member functions or state that they are not available.

find_package(catkin REQUIRED COMPONENTS

  DEPENDS boost

Here is my code:

#include <string>
#include <ros/ros.h>
#include <tf2_ros/transform_broadcaster.h>
#include <tf2_ros/static_transform_broadcaster.h>
#include <std_msgs/String.h>
#include <nav_msgs/Odometry.h>
#include <geometry_msgs/TransformStamped.h>
#include <geometry_msgs/Quaternion.h>
#include <control_msgs/JointControllerState.h>
#include <geometry_msgs/Pose.h>
#include <geometry_msgs/PoseWithCovarianceStamped.h>
#include <sensor_msgs/JointState.h>
#include <message_filters/subscriber.h>
#include <message_filters/time_synchronizer.h>

using namespace control_msgs;
using namespace message_filters;

class PublishOdometry

     motor_position_joint1_new =0;
     motor_position_joint1_old =0;
     motor_position_joint2_new =0;
     motor_position_joint2_old =0;
     x = 0;               
     y = 0;
     theta = 0;

     pubOdometry = n.advertise<nav_msgs::Odometry>("odom", 10);

     // Listen and wait for initial pose from nav
     subInitialPose = n.subscribe("/initialpose", 1, 

void handelerOdometry(const JointControllerStateConstPtr& msg1, const JointControllerStateConstPtr& msg2)
     double wheel_track = 0.4;   //distance between two wheels

     motor_position_joint1_new = msg1->process_value;
     motor_position_joint2_new = msg2->process_value;

     ROS_INFO("J1PV=%f  J2PV=%f",motor_position_joint1_new, motor_position_joint2_new);

     if (timeold.toSec()==0)  // skip first set of joint cmd
       timeold = msg1->header.stamp;
       motor_position_joint1_old = motor_position_joint1_new;
       motor_position_joint2_old = motor_position_joint2_new;

     float j1 = (motor_position_joint1_new - motor_position_joint1_old); 
     float j2 = (motor_position_joint2_new - motor_position_joint2_old);

     motor_position_joint1_old = motor_position_joint1_new;
     motor_position_joint2_old = motor_position_joint2_new;

     double  dt = (msg1->header.stamp - timeold).toSec();  // elapsed time in seconds
     float   distance =    ( ((j1 + j2)*dt) / 2.0);             // distance in meters
     float   dtheta   =    ( ((j1 - j2)*dt) / wheel_track);     // rotation

     double dx = (distance*sin(dtheta));
     double dy = (distance*cos(dtheta)); 

     theta =  (theta+dtheta); //fmodf((theta+dtheta),(2*M_PI));                   
     x = x + dx;
     y = y + dy;

     ROS_INFO("x=%e  y=%e theta=%f ddtheta=%f ",x,y,theta, dtheta);

     timeold = msg1->header.stamp;    // new old time

     geometry_msgs::Quaternion  quaternion;
     quaternion.x = 0.0;
     quaternion.y = 0.0; 
     quaternion.z = sin(theta/2);          // set yaw
     quaternion.w = cos(theta/2);

     geometry_msgs::TransformStamped odom_trans;
     odom_trans.header.frame_id = "odom";
     odom_trans.child_frame_id = "base_link";
     odom_trans.header.stamp =  timeold;
     odom_trans.transform.translation.x = x;
     odom_trans.transform.translation.y = y;
     odom_trans.transform.translation.z = 0.0; 
     odom_trans.transform.rotation = quaternion;


     nav_msgs::Odometry odometry;
     odometry.header.frame_id = "odom";
     odometry.child_frame_id  = "base_link";
     odometry.header.stamp    = timeold;

     odometry.pose.pose.position.x  = distance;
     odometry.pose.pose.position.y  = y;
     odometry.pose.pose.position.z  = 0;
     odometry.pose.pose.orientation = quaternion;
     odometry.twist.twist.linear.x  = x;
     odometry.twist.twist.linear.y  = 0;
     odometry.twist.twist.angular.z = ((theta)*(M_PI/180));  // convert to radians



void calculateJointState(const float pv, const std::string name)
     sensor_msgs::JointState jointStateMsg;;
     jointStateMsg.velocity.resize(1 ...
edit retag flag offensive close merge delete


I think your problem might be related to how boost::bind works with member functions:

georgebrindeiro gravatar image georgebrindeiro  ( 2013-09-26 10:22:15 -0600 )edit

It seem based on this reference Timesycronizer and other boost type templates are not officially supported by ROS within a class member function. Is my understand correct? If this is true what other interface to these functions is supported.

rnunziata gravatar image rnunziata  ( 2013-09-26 12:37:21 -0600 )edit

Yes...I tried adding it to several places....a working example would be great

rnunziata gravatar image rnunziata  ( 2013-09-26 13:54:51 -0600 )edit

@georgebrindeiro is right. `this` is missing in the `bind()`. Every **method of a class** actually has extra argument which is the pointer to the instance of that class. When the method is called within another method of a class, `this` is passed implicitly. ROS require callback functions to have certain number of arguments, but it has no idea about the host class. So, when the callback is called the only parameters to be passed there are the messages and `this` is not provided. However your function can not work without `this`, because it will have no access to any of the class members. So `boost::bind` simply allows you to pass it as additional argument. Since you didn't do it the compiler complained.

Boris gravatar image Boris  ( 2013-09-26 17:13:38 -0600 )edit

Replacing the line for sync.registerCallback() with the sync.registerCallback(boost::bind(&PublishOdometry::handelerOdometry, this, _1, _2)); should be sufficient.

Boris gravatar image Boris  ( 2013-09-26 17:19:19 -0600 )edit

This solution worked and the code now compiles...thank you. How can this solution get into the tutorial so others like myself who are not steeped in c++ will not repeat this. Also can you make this an Answer.

rnunziata gravatar image rnunziata  ( 2013-09-27 02:31:06 -0600 )edit

2 Answers

Sort by ยป oldest newest most voted

answered 2013-09-26 12:43:39 -0600

updated 2013-09-27 06:28:23 -0600

Have you tried adding this to the boost::bind function call? That's what they usually require for binding member functions. That might be the only thing missing.

Citing Boris' comment, this is how you would do it:

sync.registerCallback(boost::bind(&PublishOdometry::handelerOdometry, this, _1, _2));
edit flag offensive delete link more


Wow thank you! Saved my day!

soroosh129 gravatar image soroosh129  ( 2019-07-05 12:19:03 -0600 )edit

answered 2013-09-25 15:21:28 -0600

updated 2017-10-03 20:25:11 -0600

130s gravatar image

From a quick glance you are specifying wrong argument types for calculateOdometry(). Instead of const JointControllerState, you should use const JointControllerStateConstPtr& (note ConstPtr& at the end). So, all the message callbacks in ROS accepting boost::shared_ptr as the arguments.

Although passing arguments by reference is not mandatory, it is recommended and considered as a good style.

Please also note that template arguments should not be pointers, but the types of the messages themselves. So simply fixing the callback function arguments will be enough to make it compile.

edit flag offensive delete link more


I have made these changes and I am still getting a compile error. Can this be some kind of compiler environment issue?

rnunziata gravatar image rnunziata  ( 2013-09-26 03:49:04 -0600 )edit

What I learn from my past experience (and I didn't manage to make it work), the timestamps of those to be synchronized have to match 'exactly'. Perhaps this is also the culprit in your case.

alfa_80 gravatar image alfa_80  ( 2013-09-26 19:53:30 -0600 )edit

@alfa_80, the problem that you are talking about is a runtime problem, but the question is about compilation. Anyway, you are right, the time stamps should be exactly same. To deal with misaligned time stamps there is ApproximateTime message filter:

Boris gravatar image Boris  ( 2013-09-26 20:43:13 -0600 )edit

@Boris: You're right.

alfa_80 gravatar image alfa_80  ( 2013-09-27 12:10:52 -0600 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools



Asked: 2013-09-25 11:21:39 -0600

Seen: 1,104 times

Last updated: Oct 03 '17