Replace `analogServer` with `multipleanalogsensorsserver` for publishing FT data – Stint 2 about robots-configuration

[Nicogene]

Today @traversaro and I discovered that after #482 all the robots that use embObjStrain have the publishing of the FT broken because embObjStrain has not been ported to IMultipleAnalogSensorsInterfaces (see https://github.com/robotology/icub-main/blob/001b897fb29aaa1cb58ff713361fc68b2f97e1ca/src/libraries/icubmod/embObjStrain/embObjStrain.h#L52-L55)

Same thing for the robots that uses canBusAnalogSensor (see https://github.com/robotology/icub-main/blob/001b897fb29aaa1cb58ff713361fc68b2f97e1ca/src/libraries/icubmod/canBusAnalogSensor/CanBusAnalogSensor.h#L42)

This is not caught by the CI because right now multipleanalogsensorsserver does not fail if the device attached does not implement any interface (see https://github.com/robotology/yarp/blob/872679fa605b95163da4a64e61c161ab9df3d485/src/devices/multipleanalogsensorsserver/MultipleAnalogSensorsServer.cpp#L403-L413)
In this sense, we can add that if all view fail we throw an error but this will not solve the problem will only make the ci fail for the robots we broke.

We spoke w/ @marcoaccame and we checked that right now both embObjFTSensor and embObjMultipleFTSensor (the ones compatible w/ MAS pipeline) does not support right now strain board.
We would need a change embobjFTSensor in icub-main for ignoring the temperature/imu sensor when using strain, on fw side we do not need anything. In this way, we could throw away embObjStrain since it is used only w/ analogServer.
Moreover, we need to migrate the xml files of strain to the "new" device.

The last resort is to roll back the robot that is now broken to analogServer but it would be a pity.

cc @pattacini @valegagge

[pattacini]

Thanks for the summary 👍🏻

We will be discussing this the next Tue at the board meeting.

[Nicogene]

The PR #505 has been merged, I will then proceed to apply those changes to all the robots.

cc @pattacini @traversaro

[Nicogene]

PR: #517

I will test it on iCubGenova11 tomorrow. I am not sure it will be cover all the cases, I hope the dry-run will help in this sense

[Nicogene]

Putting on hold since we need to takle this: robotology/icub-main#873

[pattacini]

Putting on hold since we need to takle this: robotology/icub-main#873

Done!

[Nicogene]

We can finally close this issue, we will track the work on the simulators on a different one.

[pattacini]

I am not sure it will be cover all the cases, I hope the dry-run will help in this sense

🤞🏻