MQTT-SN SC Meeting

When:  Mar 27, 2025 from 16:00 to 17:00 (UTC)
Associated with  MQTT SN Subcommittee
  • Roll call
  • Approve agenda
  • Approve minutes from last meeting - https://groups.oasis-open.org/higherlogic/ws/public/document?document_id=72607&wg_id=3d1c9127-09c1-4b2e-8801-018dc7d3f558
  • Voting rights update 
  • Working draft progress:
    • Updated all the packet diagrams. The text is left justified, but MQTT diagrams are a mix of left justified and centred. Do people prefer left or centred?
    • Topic Type short name has been removed.
    • Topic Data/Data in PUBLISH packets changed to Topic Alias or Topic Name
    • Condensed SUBSCRIBE and UNSUBSCRIBE packet diagrams into one
    • Moved Packet Id in SUBACK to follow flogs
    • Changed Client Id to Session ID in PINGREQ. What do people think of that?
    • Short topic name in publish QoS -1
  • Working draft todo:
    • Update section 4.7.2 (Topic Aliases), once Simon has agreed to eliminate Short Topic Names.
    • Add Session ID to WAKEUP/SLEEPREQ
    • Section 3.9.3 - Can we provide an example on how to associate a PUBLISH received by a GW with a session handled by the same GW?
  • Working draft questions:
    • I suggest we move section 4.13 Example MQTT-SN Architectures, to the Implementation Guide. The whole chapter is Informative and should be labelled as such if left here. Also 4.14, Gateway Advertisement and Discovery? Also, I would suggest, PUBLISH QoS -1, section 3.9.2
    • Disconnect Retain Registrations should be on the SLEEPREQ packet instead, as DISCONNECT is no longer used to go to sleep. See issue 591 - should we just keep registrations?
    • In the Protection Encapsulation, do we need the Authentication Tag Length, as it is defined by the overall Packet Length?
    • Under what circumstances is PINGREQ sent by the server? (3.14 and 3.4.6)
    • Can we use a flag to indicate the presence of Topic Alias in SUBACK?
    • The transmission radius field in the forwarder packet is two bits. The broadcast radius in the SEARCHGW packet is 1 byte. I don’t have any real description for what this field does in the forwarder encapsulation. What are its possible values? (See 3.20.2.1)
    • Section 2.3 - would it be better for MQTT-SN to assign Reason Codes backwards from 255, in the opposite direction to MQTT? (check with Simon)
    • Issue 640 - size of Gateway Id (3.3.2) (check with Simon)
  • Working draft action items outstanding:
    1. update QoS2 descriptions and diagrams to match consensus
    2. review and update entire draft for use of Packet and Application, Will and  Retained Message terms
    3. Remove comparative references to MQTT
    4. be specific about DUP flag use
    5. re-review Topic Registration and Alias sections for consistency
    6. review and rewrite Operational Behavior sections inherited from MQTT-SN 1.2
    7. decide what contents we want for Appendix C
  • AOB
  • Confirm time and date of next meeting
  • Check for late arrivals

Location

Dial-in Instructions:
As per e-mail

Contact

Ian Craggs

icraggs@gmail.com

Related Content