timeout arg from an RPC call is used to setup client side RPC timeout
This mixing client setup and server side information used to execute the call.
client_car.clean_up(timeout=2)
Flint is using timeout
argument in server-side, usually with a very big timeout (cause it's about user interaction), so i don't think it's a very good idea to also setup the RPC timeout at the same time.
I have introduced another way to setup such a timeout with a context manager. Which I think is better.
with client_car.connection_timeout(2):
client_car.clean_up()
(as RPC calls are serialized (at least to the same object) using a context in such a way is fine, i think.)
But i would like to know is such use of timeout
(as argument) from calls is used at another place than Flint?
i.e. can we remove https://gitlab.esrf.fr/bliss/bliss/-/blob/master/bliss/comm/rpc.py#L602