[ros-users] Hybridizing ROS with Blocking TCP/IP Stack

William Woodall wjwwood at gmail.com
Sat Mar 5 07:17:48 UTC 2011


Is it possible for you to post an example (or the actual ROS node code) for
us to see how you are doing it, that might help us understand the issue,
also you might consider posting this on http://answers.ros.org/.

Thanks,

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
William Woodall
Graduate Software Engineering
Auburn University
w at auburn.edu
wjwwood at gmail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


On Fri, Mar 4, 2011 at 9:14 PM, Matt Bergsma <mjb4 at sfu.ca> wrote:

> Perhaps a bit of a weird question for you all. I have a pre-existing large
> windows code base that communicates via TCP/IP sockets. And looking to bring
> it into ros by writing a a ROS node that accepts the TCP/IP socket message
> and converts it to the appropriate ROS topic. The problem I’m currently
> running into is that ros::spinOnce doesn’t seem to like to behave and
> publish messages as requested. Currently We have a simple client TCP/IP that
> accepts the message, does some stuff to it (which works) than publishes to
> the move_base_simple/goal topic, well in theory. The publishing isn’t
> happening. Since we are using accept() to get a message from the socket
> (which is a a blocking call),  the translator node, announces to the ros
> network where its talking, spins once, than hits the blocking call to wait
> for a message, than handles it, and publishes to the appropriate topic.
>  After the publishing command, it runs ros::spinOnce() again, then returns
> to waiting for a socket message. Using printf messages I can see the
> appropriate messages showing its getting stuck at the blocking call as
> expected, but no messages are being sent on the ros topic. Any idea’s of a
> work around for this? Current thought is to spawn a thread for the TCP/IP
> stack, and then using some combination of flag, and semaphore to control the
> ROS publisher side of things. This code will need to be expanded to include
> a couple ROS subscribers in the future as well.
>
>
>
> Matt
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20110305/8635f61e/attachment-0003.html>


More information about the ros-users mailing list