C++ subscriber not working for first messages [closed]

asked 2019-10-01 00:35:28 -0500

teshansj gravatar image

updated 2019-10-01 01:49:45 -0500

Just out of curiosity, I modified the simple c++ publisher and simple python publisher. I got rid of the while loop and did the following.

c++

loop_rate.sleep();

msg.data = "hello world 1";
chatter_pub.publish(msg);
ros::spinOnce();
loop_rate.sleep();

msg.data = "hello world 2";
chatter_pub.publish(msg);
ros::spinOnce();
loop_rate.sleep();

msg.data = "hello world 3";
chatter_pub.publish(msg);
ros::spinOnce();
loop_rate.sleep();

python

rate.sleep()

hello_str = "1-hello world py"
pub.publish(hello_str)
rate.sleep()

hello_str = "2-hello world py"
pub.publish(hello_str)
rate.sleep()

hello_str = "3-hello world py"
pub.publish(hello_str)
rate.sleep()

Now when I run both publishers one after the other, rostopic echo gives me the following, which is expected

data: "1-hello world py"
---
data: "2-hello world py"
---
data: "3-hello world py"
---
data: "hello world 1"
---
data: "hello world 2"
---
data: "hello world 3"
---

The problem is, c++ subscriber callback is not triggered for the first messages. To be exact, it gets triggered after only the 3rd message for c++ publisher and after the 2nd message for python publisher. After that it works fine. This is the c++ subscriber output I get

[ INFO] [1569905819.772859593]: I heard: [2-hello world py]
[ INFO] [1569905819.874401168]: I heard: [3-hello world py]
[ INFO] [1569905825.212111970]: I heard: [hello world 3]

Python subscriber works fine

[INFO] [1569905819.678143]: /listener_17329_1569905813773I heard 1-hello world py
[INFO] [1569905819.772875]: /listener_17329_1569905813773I heard 2-hello world py
[INFO] [1569905819.875100]: /listener_17329_1569905813773I heard 3-hello world py
[INFO] [1569905825.012191]: /listener_17329_1569905813773I heard hello world 1
[INFO] [1569905825.112169]: /listener_17329_1569905813773I heard hello world 2
[INFO] [1569905825.212244]: /listener_17329_1569905813773I heard hello world 3

Can someone please explain what's happening here? I don't get it.


c++ publisher - talker.cpp

#include "ros/ros.h"
#include "std_msgs/String.h"
#include <sstream>

int main(int argc, char **argv)
{
  ros::init(argc, argv, "talker");
  ros::NodeHandle n;
  ros::Publisher chatter_pub = n.advertise<std_msgs::String>("chatter", 1000);

  ros::Rate loop_rate(10);
  std_msgs::String msg;
  std::stringstream ss;
  loop_rate.sleep();

  msg.data = "hello world 1";
  ROS_INFO("%s", msg.data.c_str());
  chatter_pub.publish(msg);
  ros::spinOnce();
  loop_rate.sleep();

  msg.data = "hello world 2";
  ROS_INFO("%s", msg.data.c_str());
  chatter_pub.publish(msg);
  ros::spinOnce();
  loop_rate.sleep();

  msg.data = "hello world 3";
  ROS_INFO("%s", msg.data.c_str());
  chatter_pub.publish(msg);
  ros::spinOnce();
  loop_rate.sleep();

  return 0;
}

python publisher - talker.py

#!/usr/bin/env python
import rospy
from std_msgs.msg import String

def talker():
    pub = rospy.Publisher('chatter', String, queue_size=10)
    rospy.init_node('talker', anonymous=True)
    rate = rospy.Rate(10) # 10hz
    rate.sleep()

    hello_str = "1-hello world py"
    rospy.loginfo(hello_str)
    pub.publish(hello_str)
    rate.sleep()

    hello_str = "2-hello world py"
    rospy.loginfo(hello_str)
    pub.publish(hello_str)
    rate.sleep()

    hello_str = "3-hello world py"
    rospy.loginfo(hello_str)
    pub.publish(hello_str)
    rate.sleep()

if __name__ == '__main__':
    try:
        talker()
    except rospy.ROSInterruptException:
        pass

Subscribers are identical to the ones provided in the example pages (c++, python)

edit retag flag offensive reopen merge delete

Closed for the following reason duplicate question by gvdhoorn
close date 2019-10-01 03:47:53.690422

Comments

2

The relevant part of the code is missing (the definition of the subscriber as well as what comes after it). But after you register a subscriber, it might take a little while until it is fully set-up and connected to all topics. So you could simply add a sleep for a second or so after the instantiation of the subscriber and check if this already solves you issue.

mgruhler gravatar image mgruhler  ( 2019-10-01 00:44:14 -0500 )edit

As I have mentioned at the beginning, I modified the code in the examples c++ and python. Didn't include it in the question since it would make the question too long to read. I'll append it at the bottom of the question

teshansj gravatar image teshansj  ( 2019-10-01 01:45:07 -0500 )edit

@mgruhler Yes adding a delay solves the problem. So it it happening because init functions exit before they actually finish initializing things?

teshansj gravatar image teshansj  ( 2019-10-01 01:55:15 -0500 )edit

Are you running all your nodes from a single launch file ? Try launching only the listeners nodes first and then after one or two secs the talkers nodes. As @mgruhler said, it's not related to init functions but to the time to connect to the topics.

Delb gravatar image Delb  ( 2019-10-01 02:06:02 -0500 )edit

I am launching them separately. And of course I am running the listeners first. To solve the problem, delay need to be added in the publisher after the init

teshansj gravatar image teshansj  ( 2019-10-01 02:46:13 -0500 )edit
2

Please see #q287548 and #q313491, of which your question appears to be a duplicate @teshansj.

You probably already did, but just to make sure: please try to search for previous Q&As before posting a new one. Try to use Google, as it's much better than the built-in search. Using something like first message lost site:answers.ros.org got me #q287548 and #q313491 as the two top results.

delay need to be added

don't add delays (time-based). A state-based approach will always be more robust.

Finally: if message reception (and acknowledgement) are really important, don't use publish-subscribe (in ROS 1). It's a fire-and-forget communication pattern, with literally no guarantees about anything whatsoever (see also #q203129). Use services or actions.

gvdhoorn gravatar image gvdhoorn  ( 2019-10-01 03:16:58 -0500 )edit
1

I've actually closed this as a duplicate.

If you don't agree @teshansj, please comment here, clarify and we could re-open if that would be justified.

gvdhoorn gravatar image gvdhoorn  ( 2019-10-01 03:48:33 -0500 )edit