[ros-users] ROS 2.0 Strategy review

[Brian Gerkey]

On Tue, Sep 29, 2015 at 10:52 AM, Michael Haberler <mail17 at mah.priv.at> wrote:
> *) it is my experience from the the precursor of the machinekit project, and I hear the same story from commercial users of realtime stacks: time and again way too much code is done in hard realtime environments - usually for a lack of understanding of the actual requirements, "playing it safe" and erring on the costly side, and sometimes it comes down to very human factors like job security.

I agree completely!  Far too often people think that they need
"real-time" when they don't (and sometimes don't even understand what
it means).

In any case, support for real-time-safe use of certain APIs is but one
design goal for ROS 2.  And of course it's up to developer of the
system to ensure that all the other requirements are met (real-time
OS, appropriate use of scheduling priority, sufficient CPU for the
workload, etc.).  We simply want to make sure that there's a way to
use our libraries to comfortably modularize your code without
sacrificing the *potential* to meet real-time requirements.  In other
words, if you know how to build a real-time system, we won't get in
your way.  That's a change from ROS (roscpp) today, which very much
does get in your way.