rsi_hw_iface: clarify how the RSI 'timeout' parameter works exactly

[gavanderhoorn]

From #54 (comment):

Is the timeout equal to the number of allowed missed packets? So at Timeout=100, 100*4ms have elapsed since the last answered RSI packet?

The text does not really explicitly state it like this, but assumes that the reader is able to understand it from the explanations further on in the same paragraph.

We might want to add a small example calculation that shows how timeout and actual time are related. Perhaps mention that timeout is measured in 'control periods'?

[BrettHemes]

I believe this was sufficiently addressed in #54

+Notice the RSIObject of type ETHERNET. Within this object is a parameter called Timeout. This parameter is set to 100 by default. The RSI interface operates at 250 Hz (4ms cycle). The kuka_rsi_hardware_interface does not have a period configured and is completely driven by the controller's output. Every controller RSI output has a IPOC timestamp which increments for every cycle. The kuka_rsi_hardware_interface will answer as quickly as possible. The answer includes the last IPOC received. If the connected PC with ROS did not manage to answer within the RSI cycle of 4ms, the IPOC timestamp of RSI has incremented. The IPOC included in the answer is not matched and the controller will increment a counter. When this counter hits the Timeout parameter (100), the RSI connection will shut down. If this parameter is lowered, the demand for real-time computation will increase.

[BrettHemes]

Documented on the troubleshooting guide via the wiki here

[gavanderhoorn]

👍 thanks again.