The protocol layer gives occasional warnings about too big error in floating point number conversions about ping times. Likely explanation is that ping times are sometimes bigger (using higher exponent) than what we're targeted the network protocol for, so that too much of the precision is lost at the least significant digits. Straightforward move would thus be to move the "window" in which we expect the exponent to be -> to expect bigger numbers and less accuracy.
The protocol layer gives occasional warnings about too big error in floating point number conversions about ping times. Likely explanation is that ping times are sometimes bigger (using higher exponent) than what we're targeted the network protocol for, so that too much of the precision is lost at the least significant digits. Straightforward move would thus be to move the "window" in which we expect the exponent to be -> to expect bigger numbers and less accuracy.