12-18-2016, 12:50 PM
I might do this at some point in time; however if I look at the required steps and what the final project could look like there are a couple of steps in between.
First step was to prove that it could be done talking to the nano over Wifi (check).
The next steps include (but are not limited to):
- figure out which protocol to use (probably UDP) - PoC connection issues with UDP
- run intensive tests on the existing platform (it would be nice if two or three more people could actually sign up to this and run tests)
- evaluate if we need another hardware module (e.g. netnano433.c) or if we just use the exisiting 433nano.c and extend it to net sockets
This last mentioned step hugely depends on whether we want to simplify module hardware to a maximum, i.e. provide a small matchbox-size device that would only have an ESP8266 plus sender and receiver or if we want to keep the Arduino in the design.
In order to take an educated choice on this issue I would really really love to have a couple of more opinions on this. Would people use it ? Is it worth porting the Arduino Code to ESP8266? The code is very timing-sensitive (surprise) and porting it will add new errors, new bugs etc.
If we find out that using UDP and socat works reliably and stable the whole remaining "workload" would be to modify the daemon startup (/etc/init.d/pilight) to launch socat depending on config.json - done!
If we would need a fault-tolerant network connect (i.e. have pilight try to reconnect to the wifi dongle if it had failed) then modifying the pilight code would be necessary (433nano.c or new module).
If people really want it and think it is cool to have a mini device of the size of a matchbox as a sensor then we might be looking into porting the pilight-usb code to esp8266, making it a new device.
If I look at my personal agenda which includes pilight as a fire and motion alarm detector that runs on my OpenWrt routers then I am fine with solutions 1 and 2. But I am ready to give something back to this project and think about solution 3 but we really really need a bit of feed-back from people using it.... plus a handful of folks being able to run tests and possibly re-flash an esp8266....
First step was to prove that it could be done talking to the nano over Wifi (check).
The next steps include (but are not limited to):
- figure out which protocol to use (probably UDP) - PoC connection issues with UDP
- run intensive tests on the existing platform (it would be nice if two or three more people could actually sign up to this and run tests)
- evaluate if we need another hardware module (e.g. netnano433.c) or if we just use the exisiting 433nano.c and extend it to net sockets
This last mentioned step hugely depends on whether we want to simplify module hardware to a maximum, i.e. provide a small matchbox-size device that would only have an ESP8266 plus sender and receiver or if we want to keep the Arduino in the design.
In order to take an educated choice on this issue I would really really love to have a couple of more opinions on this. Would people use it ? Is it worth porting the Arduino Code to ESP8266? The code is very timing-sensitive (surprise) and porting it will add new errors, new bugs etc.
If we find out that using UDP and socat works reliably and stable the whole remaining "workload" would be to modify the daemon startup (/etc/init.d/pilight) to launch socat depending on config.json - done!
If we would need a fault-tolerant network connect (i.e. have pilight try to reconnect to the wifi dongle if it had failed) then modifying the pilight code would be necessary (433nano.c or new module).
If people really want it and think it is cool to have a mini device of the size of a matchbox as a sensor then we might be looking into porting the pilight-usb code to esp8266, making it a new device.
If I look at my personal agenda which includes pilight as a fire and motion alarm detector that runs on my OpenWrt routers then I am fine with solutions 1 and 2. But I am ready to give something back to this project and think about solution 3 but we really really need a bit of feed-back from people using it.... plus a handful of folks being able to run tests and possibly re-flash an esp8266....