Add QP_LIFESPAN to READER_QOS_MASK

[rotu]

Fix CI test
http://build.ros2.org/user/rotu/my-views/view/CycloneDDS/job/Fci__nightly-cyclonedds_ubuntu_focal_amd64/lastSuccessfulBuild/testReport/junit/rclcpp/TestNode/get_publishers_subscriptions_info_by_topic/

Signed-off-by: Dan Rose dan@digilabs.io

[rotu]

Build failures look spurious.

[eboasson]

Thanks @rotu , but there are some issues with this PR:

• The first is that this while adds a "lifespan" QoS setting on the reader, it doesn't actually implement any behaviour.
• The second is that the absence of "lifespan" on the reader is deliberate: in DDS, it is a QoS of the writer (see, e.g., page 2-110 of the DCPS 1.4 spec). There is certainly a reasonable argument in favour of adding a "reader lifespan", but there are also reasonable arguments why it doesn't make sense to do that.

In my view the real issue is that "get_publishers_subscriptions_info_by_topic" incorrectly tests whether one can set/get lifespan on a reader. In rcl that got fixed by ros2/rcl#571

After due deliberation it might turn out that adding a lifespan QoS on the reader is a worthwhile exercise, but that's a separate matter.

[rotu]

Thanks, @eboasson! Superceeded by ros2/rclcpp#1002