Fixed handling of temperature in MultipleFT service

[triccyx]

Bug fix relating to temperature acquisition period for multiple FT sensor device

[pattacini]

Awaiting the CI and then merging.

[marcoaccame]

I have tested the changes on my HW setup and now the rate of TX of temperature is correct

[traversaro]

Just to understand, what happened before this PR? I guess that period was parsed as seconds instead of milliseconds, so the temperature in existing boards that used embObjMultipleFTsensors was streamed only every 1000/60 = 16 minutes? Or there was something else mitigating this?

fyi @HosameldinMohamed

[triccyx]

Just to understand, what happened before this PR? I guess that period was parsed as seconds instead of milliseconds, so the temperature in existing boards that used embObjMultipleFTsensors was streamed only every 1000/60 = 16 minutes? Or there was something else mitigating this?

fyi @HosameldinMohamed

@traversaro
Two things:

• Period read in msec but not correctly transformed in sec. For example, 10 msec in XML became 10 sec
• If you have used a period > 253 the period became always 253 (cast to uint8)

[traversaro]

Ack! So basically the temperature on the FT sensors of the foot of iCub3 (the only robot where embObjMultipleFTsensors is used at the moment) even if the .xml request 1000 (i.e. one seconds) was actually updated every 253 seconds, i.e. 4 minutes and 13 seconds.