When communicating with instruments, there are three types of timeout to consider. One is handled by the communications library, one must be configured, and one must be handled by the library consumer.
As described in the command reference, RBR instruments will sleep after 10 seconds without input. The library tracks when a command was last sent to the instrument and wakes the instrument as necessary when sending subsequent commands.
This process should be transparent
to users of the library.
For implementation details,
see RBRInstrument_wake()
in RBRInstrumentInternal.c.
The second type of timeout, configured by the user via the \a commandTimeout value passed to RBRInstrument_open() or RBRInstrument_setCommandTimeout(), defines the minimum amount of time the library will wait for a complete response from the instrument.
Because the user-provided callback to read instrument data is defined to block, this timeout can only be checked between read operations.
Because character reads are implemented by the user
via the RBRInstrumentReadCallback() callback function,
any character timeout must also be implemented by the user.
On POSIX systems, this can be done by using select(3)
to determine whether a file descriptor is ready for reading
before a subsequent read(3) call;
or for termios serial port I/O,
via appropriate configuration VMIN/VTIME.