ilcato

You can put an empty string in the parameter or you can completely eliminate it.

Not sure

If you have some scenes triggered by global variables you can have such them triggered by switches in Homekit.

The purpose of the app is exactly that: the rooms (zones) and its associations with devices. Before importing Zones and Rooms you must first bind homebridge and have the devices visibile in homekit.

What's in your config.json?

Ok, they changed the type. I just published version 1.0.2 of the plugin that should solve your problem. Can you test it?

Ok. Probably Fibaro changed the signature of the device too. Can you send me the results of the following query on your browser: http://10.0.0.16:80/api/devices/150 where you need to replace the IP address with the one of your Home Center and the 150 number with the ID of your device. Thanks

It's not clear to me. Is anyone able to explain the problem in English?

upgrade node.js, homebridge and homebridge-fibaro-hc2 to the latest versions.

EADDRINUSE means that you are running homebridge two times; kill existing running instances of homebridge before launching another.

Some recent updates: Version 0.6.6 + Added support for Door Locks (tested on Danalock and Yale devices Version 0.6.5 + Fixed bug in contact sensors detection Version 0.6.4 + Added support for Humidity sensors Version 0.6.3 + Bug fixes in room grouping

The default value of the parameter if you don't specify it in the config.json file is 2 second, the same as the previous version of the plugin. @JossAlf, if you want to lower the probability of the failure you must set a refresh time higher than 2 seconds. Try 4.

Just published version 0.6.2 of the plugin. Added support for a new config parameter that allows to control the way in which the plugin checks the values of the characteristics that are updated outside the HomeKit world. Previously the plugin was polling the Fibaro Home Center every 2 seconds looking for updates. I think that this was the cause of the crashes of the Home Center that are described in the forum; probably because of a saturation of some internal logs files. I tested with some of you that were experiencing this problem a version of the plugin that completely disabled the polling and had the confirmation that the crashes were no more present. With this config.json parameter called "pollerperiod" you can: completely disable the polling by setting the parameter to 0 (the drawback of this is that if you have some triggers in HomeKit based on Home Center characteristic changes they will not fire and in order to view the real value of a characteristic in an HomeKit app you need to manually refresh the data. change the poller period from 1 second to 100 second. For those of you with the crash problem (I never experienced it) I suggest to change the value from the default of 2 second to an higher value in order to check if the problem disappear without loosing the capability of using HomeKit triggers or to have the changes happened outside of Homekit to be automatically reflected inside HomeKit This release corrects also a bug in the plugin that with iOs 9.3 make the blinds in some cases crashing.

I can send to you a modified version of the plugin to test the hypothesis that I describes before. Are you available?

It simply confirm that the HC2 is unreachable. On the topic of the crash of the HC2 I have a conjecture: the plugin polls with a specific Fibaro API the HC2 every 2 seconds (43200 times a day) in order to check if there are changes in the values of properties of the devices and present it to the user interface of the HomeKit apps it is possible that HC2 logs something in a file for every call and after some times it saturates the space available on the filesystem and crashes. Probably for configuration with specific devices or plugins. when HC2 reboot a cleanup restore the free space In order to validate this hypothesis I'm looking for someone of you that is exposed to the problem that is available to perform a test (unfortunately I'm not experiencing the problem). Thanks