Major 215: remove "other" in "other Clients" - the same client may subscribe
global: when referencing a specific Control Packet e.g., CONNACK we should call this either CONNACK Packet or a CONNACK Control Packet. Currently the doc uses both styles which is inconsistent
466-474 PacketID in narrative and example disagree
557 change loose to lose
667 remove comma after full stop
765 - 767 narrative is ambiguous. I think this means that a Publish Packet received by subscriber which has used a wild card to subscribe will see the full topic name in the packet, but the server may have mediated it. The current wording is not completely clear.
857 should state that PUBREL is the third packet in the QoS 2 protocol flow. See 831 which states that PUBREC is the second.
883 should state that PUBCOMP is the forth and final packet in the QoS 2 protocol flow.
865 and 921 reword to Bits 2,1 and 0 .... set to 0,1 and 0 respectively.
919 and 1030 Section reference is wrong. DUP is section 2.2.2.1
Table (line 1202 to 1202) inconsistent use of PacketID and Packet Identifier (e.g., rows 4 and 8)
1255 and 1257 we introduce the term 'in flight window' which is not described elsewhere. The sentence could be rework and 'Where a single message is in flight and each delivery flow is completed ..." and "Where more than one message is in flight, message ordering..." thus removing the need for the term window.