elliotwoods / dynamixelsdksharp Goto Github PK

So far this is mostly not important
But it could be useful for a number of requests in the future

• If a goal position send fails, then put it in a queue (with delay) for retry
• If a port fails to open, put it in a queue to retry
• Look for missing servos periodically

Quick boot feature

Send everything to its corrected zero position (and then once it arrives, turn off A1 torque)

Hardware errors often occur during group operations (e.g. refresh)
These need to be handled
Also we should sometimes try rebooting the servo

• CheckSafetyConstraints can temporarily shut down a servo (e.g. for 5 mins)
• CheckSafetyConstraints can put a servo into limited current mode

Refresh - the request was aborted

Wait until an exclusive lock is available before attempting to make a schedule request

E.g. shutdown the system at a certain time

should always run

My device (XM430-W350) has a secondary ID feature that enables multicasting of packets.
As it looks like the device doesn't reply to a broadcast ping with a secondary ID, they can't be detected with Port.Refresh().
Although it can be manually added to a Port instance with Port.AddServos(), as the model number can't be retrieved using pingGetModelNum() , the model number gets set to 0.
It would be great if this situation could be handled cleanly.

get this working correctly

Any errors specific to servos are logged with ServoID in record
Also level (trace, error, warning)

e.g.

• Test/ZeroAllServos
• Temperature warnings
• Failed to move

Can we just do background tries of polling servos?

Sometimes we have errors with things going missing. It would be good to keep track of what is available at any time.