IoT SIG - 2024-04-08


  1. debacle

    I would like to collect some marketing bla, about the advantages of XMPP for IoT compared to other protcols, such as MQTT. So far, I have I have two: 1. Batteries included. Not (much) need for any plumbing work. Everything is there, e.g. SASL, roster, PubSub/PEP. 2. XMPP allows for seamless integration of {human,machine}-to-{human,machine} communication, while other protocols are either IoT or human chat. What else? flow, MattJ, any ideas?

  2. flow

    debacle, I always wanted to make a "the XML overhead is not a big as you may think" study, as I suspect that for a huge fraction of use-cases, using XML as data exchange mechanism is just fine and the perceived disadvantages of non-binary transports either do not have a significant impact and/or can be easily mitigated (e.g. by using compression)

  3. flow

    sadly, I never did such a myth-busting study, nor am I aware of one

  4. debacle

    flow We use stream compression (XEP-0138) for our measurement data (I believe, that it's pretty safe for our use case) and we can live with that. We have the payload data in JSON containers (XEP-0355), which nowadays is easier to work with, assuming JSON, Python and similar data sinks. I thought about alternatives, such as XML or ProtoBuf or ASN.1 for payload data, but pragmatism won.