Gariben's profile - activity

I'm getting a 'mismatched protocol' error when publishing a message (UInt8) from an Arduino to ROS:

Mismatched protocol version in packet: lost sync or rosserial_python is from different ros release than the rosserial client

It should be noted that the message still works, I'm just a bit concerned about the error message and would like to learn more about it.

There's a lot of threads around ROS answers stating that people are getting the error from rosserial because their message of choice was too large (message types such as IMU, etc.)

I took a look at the rosserial limitations and saw that for my particular board (ATMEGA328) that the buffer size is 280 bytes. Shouldn't the Uint8 type in theory only be 8 bytes, or is there something I'm missing?

Here are all of the ROS components from my program:

#include <ros.h>
#include <std_msgs/UInt8.h>


ros::NodeHandle nh;
std_msgs::UInt8 ROS_int;
ros::Publisher state("Device_state", &ROS_int);

void setup(){
  nh.initNode();
  nh.advertise(state);

}

void loop(){

 ROS_int.data = CurrentState;
 state.publish( &ROS_int );
 nh.spinOnce();

Is it perhaps that I'm using some outdated syntax on something?

Edit #1: @ahendrix I can see the messages on the ROS side, I haven't checked the frequency... but the error could be related to that. Here's an example screenshot from when I'm echoing the value vs. what rosserial is telling me. It does seem inconsistent. Does it have something to do with my setting perhaps? I may just try flashing it as Wolf suggests.

@Wolf I'm running Hydro on the laptop, and the ros_lib on the arduino is a few months old? perhaps. I remember in September or something I initially downloaded that, I'll give it a try and see if that fixes.

Hello, I was wondering if one can use ROS to modify wiimote inputs before they reach the Wii console.

For example, hooking the wiimote up to the computer, and modifying the state of the buttons via ros, thusly directly inputting to the console from the computer. The use of the wiimote is an important component though. I would eventually like to implement this while running Wii software. My worry is that the wiimote's signal can't be split between the computer and the console.

Hello, I have created a two-dimensional array, and much like MATLAB or Octave, I would like to turn it into an Image message. I was working to publish the array directory, and perhaps go from there, but that also seems annoying. If perhaps that's necessary, here's what my messages look like:

Image_Matrix.msg Header header uint32 width int32[] degrees

degrees.msg int32 height int32[] lasers

lasers.msg uint32 distance

Basically what I seek there is just an array, image_matrix, of width degrees and height lasers, containing the information distance. I ultimately want this to be a displayable image, but I also understand it might be better to simply advertise the matrix. I'm also aware in order to advertise it as an image, I'll need to scale the information distance from 0 to 256 in order to make it a grayscale image.

Thanks.

Do you know how I can access the data from the file rawdata.cc? I can't even find the file in my system. I want to try to extract the information I want from it. edit: Moved the conversation to github.

I may look to doing that one day, when I'm not so deeply engrossed in my current research. I'd love to contribute! Would you happen to know if the PCL HDLgrabber has this quality? I took a look at some of the classes and it seems that in the firing class it has the angle, and then in the return class it has the depth. I perfectly understand is this is outside of your well of knowledge though. I really appreciate your prompt help with my questions! Thanks again.

Hello, I am wishing to create a depth-image from my velodyne data. I would like to use a degree paramater (from 0 to 360), the height (y-coordinate) and the depth from the velodyne. I was wondering if you can return what 'degree' the point corresponds to. I've heard that the Velodyne stack is not yet able to return this kind of data.

Thanks!

Sorry, new to the forum. I did it.

That update did it. Thanks for your help, joq. Your stack is awesome and your super helpful. Just for reference I'll include pics for reference later. http://i.imgur.com/RCmRDsN.png?1 http://i.imgur.com/O0hljAx.png?1 Thanks again. I'll try to [solved] this guy.

Note that it also does not change as you make the min_range larger either. http://i.imgur.com/DBcWloU.png

I gave that a try, unfortunately no avail. http://i.imgur.com/xIOchHC.png I also changed the min_range at every instance I could find, but it still defaults to 2.0. This may have something to do with the problem. Perhaps it's being overidden to 2.0 somewhere.

I included screenshots in a reply above. I also reintialized RViz whenever I took them (though I don't think it matters). Any ideas why it would still behave the way it does? I'll give the launch dealie a try though.

I don't have the 'karma' to include links, *sigh*. I'll try to get a work around. normal: http://i.imgur.com/J6hgmZU.png altered: http://i.imgur.com/8NO7GFi.png

Hello, I am a user trying to use the velodyne stack with ROS. I'm studying human interaction, so I would like the minimum distance with the Velodyne HDL-32e to be as small as possible. I understand that it has limits (as stated in the manual), however, when I attempt to change the value of the min_range parameter, nothing happens. Rviz does not update, even when I reinitialize it or anything else. The value definitely changes (as can be seen in my screenshot), but the data still reflects the minimum range as being 2.0 metres. Can someone help me out with this? Is it a possible bug?

[See joq's answer for solution]