12:30 < lulf> ppatiern: there is also a question of what we will do with mqtt as we have it today
12:31 < ppatiern> I don't like to have it as a different address space ... imho mqtt should be just a different supported protocol for accessing an address space I can already access using AMQP
12:33 < lulf> right
12:34 < lulf> to compare it with the kafka case... the equivalent would be to have a 'mosquitto' address space (or any mqtt broker really) and then have a mqtt-bridge for mapping the mqtt address into the
              'standard' address space
12:34 < lulf> but the mqtt-gateway is different in the sense that it is more like 'tunneling' rather than bridging?
12:43 < ppatiern> yes I see mqtt vs kafka differently ..
12:43 < ppatiern> mqtt just as a way to reach brokers/routers related address spaces with a protocol that is not amqp
12:43 < lulf> right
12:44 < ppatiern> kafka is for different "use cases" in the messaging business
12:44 < lulf> i'm wondering how to fit that into our definition of an address space
12:44 < lulf> :)
12:44 < lulf> perhaps the answer is mqtt support in the router :)
12:44 < ppatiern> yes :-)
12:44 < lulf> in one way, the mqtt-gateway is sort of a work-around for that
12:45 < lulf> although it has many other uses
12:45 < ppatiern> I don't see it as a work-around ... just as the MQTT support for the router but ... outside the router  ? :-)
12:46 < lulf> yes, i don't mean work-around in a negative sense... but that in an ideal world the router would support mqtt perhaps
12:46 < ppatiern> and maybe if you have a broker with AMQP only ... you could use the MQTT gateway for supporting MQTT
12:46 < lulf> yes
12:46 < ppatiern> just thinking ... not something that we are interested in :-)
12:46 < lulf> the multi-protocol support thing is something we need to define more clearly
12:47 < ppatiern> yes I think that this is the right word ... multi-protocol support that is different from defining an address space
12:47 < lulf> perhaps an address space can be 'accessed through a single connection per protocol'
12:48 < lulf> and the protocols supported are defined by the address space
12:49 < ppatiern> yes I like it