[ros-users] driver implementation
Blaise Gassend
blaise at willowgarage.com
Wed Mar 10 17:51:21 UTC 2010
Hi Marti,
If you feel that your drivers would benefit from using the driver_base
package, feel free to use them. Should you chose to use driver_base, the
problems you are likely to encounter are:
- There is currently nearly no documentation.
- It is likely to change in the future, as it has not been reviewed, and
there are many things in the API that I do not currently like.
The advantages for you of using driver_base is that it makes it easier
to support dynamic configuration, self tests and diagnostics. When I get
a chance to add things like disabling the driver driver_base, you will
automatically benefit from the enhancements.
In general I try to make API changes as painless as possible, trying not
to break old code without a deprecation period. That makes it easier to
support my drivers that use driver_base. But I can't guarantee that this
will be the case. Also, if you are using boxturtle, you can be sure that
the API for driver_base will not change. It will only change when you
switch to the next distribution.
I hope this helps answer your question,
Blaise
On Wed, 2010-03-10 at 12:40 +0100, Marti Morta wrote:
> Hi,
> I am a new ROS user and in the group where I am involved, we are
> considering to use ROS as a communication middleware for a mobile
> platform. We have already read the documentation and done the main
> tutorials.
>
> We would like to develop the drivers for the platform and also for the
> different sensors on board. We have checked the hokuyo_node driver
> source code and we've seen that it depends on the classes driver and
> driver_node. We were thinking on developing our drivers following the
> same architecture, by inheriting from these two classes.
>
> However, we have seen in the driver_common wiki documentation that these
> classes are immature and for internal use. We were wondering which is
> the best way to implement our drivers. We would like to avoid data
> publishing without using the driver hierarchy as we have seen in other
> driver implementations.
>
> regards,
>
> Martí
>
>
> ____________________________________________________________
> Martí Morta Garriga mmorta at iri.upc.edu
> Institut de Robòtica i Informàtica Industrial CSIC-UPC
> Llorens i Artigas 4-6, 2a pl. 08028 Barcelona, Spain
>
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
More information about the ros-users
mailing list