[ros-users] Can service answers be asynchronous?

Eric Berger berger at willowgarage.com
Tue Oct 5 16:24:45 UTC 2010


ROS communication protocols allow for it, so blocking is basically a design
choice in the client libraries.  You could either try modifying the client
library that you're using, or build something on top of it (at the cost of 1
thread per in-flight service request) that provides the behavior that you
want.
--Eric

On Tue, Oct 5, 2010 at 8:50 AM, Ken Conley <kwc at willowgarage.com> wrote:

> It's not on our roadmap for ROS 1.4:
>
> http://www.ros.org/reps/rep-0101.html
>
> For rospy, I was waiting until I redesigned the inner i/o engine. I
> won't speak for roscpp.
>
> best,
> Ken
>
> 2010/10/5 Stéphane Magnenat <stephane.magnenat at mavt.ethz.ch>:
> > Hello,
> >
> > I am about to start writing a bridge between ROS and ASEBA [1]. In this
> > context, ASEBA is a low-level (cross-microcontroller) middleware, whose
> > features I would like to expose to ROS. I already did this with D-Bus
> [2].
> >
> > As ASEBA talks to microcontrollers, there might be delays in answers, so
> > when an RPC-like query is performed on the Service, I would like the
> > function to return to the event loop and to be able to provide an answer
> > to the RPC later, when the data from the microcontroller will  have
> > arrived. As far as I understand, as ROS is message-based, this should be
> > possible. D-Bus provides an API for this [3]. I have not found any
> > information related to asynchronous answers in the context of ROS, and
> > if I have read the API doc correctly, ROS does not provide such a
> > feature in its service API. Is it correct? If so, do you plan to
> > implement this at some point?
> >
> > Thank you, kind regards,
> >
> > Stéphane
> >
> >
> > [1] http://mobots.epfl.ch/aseba.html
> > [2] http://svn.gna.org/viewcvs/aseba/trunk/medulla/
> > [3] for instance, with QtDBus, see:
> >     http://doc.trolltech.com/latest/qdbusmessage.html#setDelayedReply
> >
> > --
> > Dr Stéphane Magnenat
> > http://stephane.magnenat.net
> > _______________________________________________
> > ros-users mailing list
> > ros-users at code.ros.org
> > https://code.ros.org/mailman/listinfo/ros-users
> >
> _______________________________________________
> 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/20101005/f4566a5d/attachment-0003.html>


More information about the ros-users mailing list