Better qualify bad packets #8
Description
Currently we have the method is_bad_packet() which just evaluates if a packet is too short to contain the cmd/address byte. This might not even be valid, e.g. for status packets?
One case of bad packets I've seen is an 255 255 0 255 packet that showed up at the beginning of a log.
Other thoughts:
- Include check-sum checking in this method?
- Compare actual packet length (L+4) with the length byte (L)
- packet not long enough for at least one value byte = bad?
- does packet good/bad depend on the instruction byte?