Revision history [back]

You are just observing floating point precision errors. That is why we use 64 bit precision to store ROS time values.

I added this line to your program:

   ROS_INFO("delta between float and double %f %f float cast of time %f", f_grabtime, d_grabtime, (float) d_grabtime);

Where d_grabtime is a double instead of a float. And you get this output:

[ INFO] [1378710468.378363642]: delta between float and double 1378710528.000000 1378710468.378310 float cast of time 1378710528.000000