Changing to neatatmo 2.3 binding, unclear how to migrate from 1.9

So I have been using the 1.9 binding for several years, however i finally started moving things over to 2.xx bindings when i upgraded OH2.0 to 2.3, i.e creating things files

So can I do auto discovery and then save it to files(is this possible?) or should i set up my things manually?
In the documentation it says:

Thing netatmo:NAMain:home:inside "Netatmo Inside" [ id="aa:aa:aa:aa:aa:aa" ]

and

Thing NAMain livingroom [ id="aa:aa:aa:aa:aa:aa" ]

Which is correct?

And does it need to be called NAMain and NAModule1, NAModule2, NAModule3… Or what is the correct naming configuration here, it is not stated in the documentation.

Finally it says :
For Welcome or Presence Camera, Netatmo servers can send push notifications to the Netatmo Binding by using a callback URL. The webhook URL is setup at bridge level using “Webhook Address” parameter. You’ll define here public way to access your OH2 server:

```
http(s)://xx.yy.zz.ww:8080
```

Your Netatmo App will be configured automatically by the bridge to the endpoint :

```
http(s)://xx.yy.zz.ww:8080/netatmo/%id%/camera
```

where %id% is the id of your camera thing.

Please be aware of Netatmo own limits regarding webhook usage that lead to a 24h ban-time when webhook does not answer 5 times.

What do I need to do to get camera working with an openhabian installation? And how quick is the motion detection on the camera, i.e how many ms from motion to a switch item in openhab has been triggered.

You could do automatic discovery and then translate the JSONDB parameters to a .things file, then delete the Things in PaperUI before saving the .things file (I don’t know what will happen if you save a .things file with Things with the same ID as an existing one).

Personally, I think it’s not worth the effort. The JSONDB is a text file itself so except for the fact that it is in a different folder and a different format it’s kind of all the same. Only the advantage is it gets automatically populated (autodiscovered) and it gets automatically backed up (see /var/lib/openhab2/jsondb/backups). So in my setup I use PaperUI for all automatically discover able Things (which is pretty much all of them anymore) and traditional text files for everything else.

Maybe both? I can’t say. I would just let autodiscovery handle the Things to avoid these sorts of issues. If you want to avoid PaperUI you can do autodiscover with Habmin, the REST API, or through the Karaf console. Then this sort of question becomes moot.

I think it can be arbitrary but again I’m not sure.

Hopefully a Netatmo user will chime in.

Hi,

I use manually things.

The NAMain is the big module (the master module).
The NAModule1 is the outdoor module.
The NAModule2, I thing is the rain module (I don’t have this module, I can’t tell you 100% sure).
The NAModule3, I thing is the wind module (I don’t have this module, I can’t tell you 100% sure).
The NAModule4 is the first indoor module.

Best Regards,
Fernando Gomes

Hi Kim,

According to the ESH docs the second one is the preferred way to use. So the syntax is Thing <thingTypeId> <thingId> [] . As a short cut it is possible to skip the Thing in the textual configuration.

Yes, absolutely. Those are the so called thingTypeIds which has to be defined by every binding. They tell the framework the purpose of each thing. Feel free to choose whichever name you want for your thingIds.

The cameras are not my speciality. I am afraid I cannot help you with that. Let’s ask @Lolodomo for an advice.

So is it ok, if I delete the first one in the documentation then, to avoid confusion?

Yes, that would be very much appreciated. But please correct them instead of deleting them.

Already filed a merge request… Now I just need 2 person to check it so it can be approved…