The timestamp calculation in documentation can lead to misinterpretations.
On https://harp-tech.org/protocol/BinaryProtocol-8bit.html#microseconds-2-bytes it is described that Timestamp(s) = [Seconds] + [Microseconds] * 32 * 10-6.

Having it written as Timestamp(s) = [Seconds] + [Microseconds] * 32E-6 would be more clear and help preventing mistakes as the original version can also be interpreted as [Microseconds] * 32 * 10E-6 which is wrong.