While playing around with the semantic model i found some parts that i am not able to find into the model.
If we find answers to these question it may help other aswell to understand and to use the sematic model.
1.) I use the astro binding to get various time and information in my system. So Astro is for me a kind of information source. But how does ist fit into the semantic model. From the first impression i thing it in somehow in openHAB, but there is no location like that.
2.) I use our smartphone as one source for presence. The same question here how do they fit into the model. They are kind of movable an have no location. I could think about something like a presence zone or a mobile device location, but thats not in the model. This question fits too to the question how tho handle smart plugs that are changing position and purpose over time (Xmas Tree, fan, …)
3.) I use the Corona binding and the Tankerkönig binding to get information from the internet. How should i fit these infromation sources into the model?
Maybe i am wrong with my impression that my complete smarthome should fit into the model?
In general you don’t have to set a location for the equipment that doesn’t have an explicit or stable one. They will stay on the root level, but that’s fine. Also if you don’t find a class of equipment that matches, you can simply classify them as Equipment.
With some stretch of the imagination, smartphones can be considered NetworkAppliances.
For the Corona stats and other cloud services, you can make them WebServices.
@ysc Thanks for your reply. Just to refelct that i have understood it right.
1.) The sematic model ist not neccessarily a hierarchical struktur based on location that will held all the equipment at the end. Eqipment is able to “live” without a location. So is it a possible solution to collect all the locationless equipment in a location named “no Location”
2.) Classes “Location”, “Equipment” and “Point” are valid classes and not a kind of root that should not be uses. So my complete model could consist of “Location”, “Equipment” and “Point”
3.) Be creative in interpreting equipment Corona Stats -> Webservice; TV -> Screen; Smartphone -> NetworAppliance (very creative :-))
Unfortunately i was not able the create a setup that autogenerate the equipment/location pages.
Correct. Basically you can have from the root level:
Location [> Location]* > Equipment [> Equipment]* > Point
Location [> Location]* > Point
Equipment [> Equipment]* > Point
Point
(* denotes “repeated 0…n times”)
For example I have a Spotify equipment at the root level (because it’s a multiroom system) and a “General Scene” control point, which is not related to any equipment in particular.
The model treeview shows the locations first, then the equipment, then the points, then the non-semantic groups, then the non-semantic non-group items (the last two only when you check the “show non-semantic” box) in that order, at each level. I don’t see much value in having a virtual “no location” location.
Correct too. There may be additions to the ontology (new classes) after discussions on GitHub (openhab-core) then you’ll be able to specialize your equipment or location items.
@YSC I can confirm that in build #1999 the root tags are working.
As i said in other threads is see shortcomming in the ontology. From my point of view it should be extended to be more userfriendly. I there a discussion i could join? I did not find any.
