|Re: [paho-dev] [iot-wg] Future Requirements and Releases of Paho|
Ian thanks for your comments. On 08/28/2014 02:34 PM, Ian Skerrett wrote:
Ian, Thank you for starting this discussion. I think the timing is good. Some suggestions/questions about future Paho requirements: - I am assuming the MQTT specification will evolve over the next couple of years. Would it make sense to start an incubator to try out some new ideas for the spec? Developers could contribute code that implements these ideas and this would help the MQTT TC in the evolution of the spec.
Interesting idea. By incubator, do you mean a separate project?
A low-level MQTT-SN C already exists (https://www.eclipse.org/paho/clients/c/embedded-sn/). Future developments include higher level C++ MQTT-SN client libraries, in the same style as the MQTT embedded C libraries. I am just about to contribute some Java MQTT-SN code that was written in IBM Zurich at the time MQTT-SN was developed. (MQTT-SN on the server side is implemented in RSMB which currently resides in the Mosquitto project).- Is MQTT-SN implemented in Paho? If not is this a potential future requirement.
Yes it could - I would like to see that. I think I have been wary of stepping into the remit of other IoT projects. We do have some simple MQTT exerciser apps. Then there is the interoperability test package which contains tests for clients and servers (to be improved) and a proxy to sit in between a client and server to display the packets being sent and received. Any ideas for helpful tools welcome - I have been using MQTT for so long now, it can be difficult to work out what would help newcomers.- I'd like to see a focus on creating tools to test/debug/deploy MQTT apps. Tools like MQTTLens I think will be critical to the adoption of MQTT. Could Paho be the home to these types of tools.
I was thinking if at some point there was no new release, perhaps not even a service release as it was so stable, would it go into a release train? Actually, we're not anywhere near that point, so I'm probably worrying unnecessarily.For the MARS release train, I think Paho would be a natural fit. Having stable, mature code that is updated on a regular basis is what our community will need for long term use. I also hope that we have other IoT project participate in Mars, so we can start thinking about an IoT package. This certainly should be a topic for future IoT WG meetings.
I hope this helps. Ian
-- Ian Craggs icraggs@xxxxxxxxxx IBM United Kingdom Committer on Paho, Mosquitto
Back to the top