Semantic model muddle

Struggling a bit with some facets of the semantic model in OH3. Trying a clean set up from scratch. Have read lots on forums and studied examples but still need some advice/opinions. Thanks in advance :slight_smile:
Example scenario: Fibaro Dimmer controlling lights in my Sunroom.
Sunroom - Group item, semantic class room. (makes sense to me ok).
Dimmer itself - Equipment item, label Sunroom dimmer, semantic class lightbulb (should this be slider class? or what if just add a point directly under the room and skip the equipment part altogether? what would I miss out on?)
Point - label Sunroom lights, class point. (what if label dimmer, semantic class control or switch??).
I’m getting muddled over choices for classes! How do these choices impact on those UI ‘cards’ or ‘tiles’. I want a consistent framework that makes sense but also that is best practice.
How do I just import my old (nice) sitemap from OH2 if I can’t get my head around this modelling? Thanks for your patience.

Waiting for some more experienced user respond to the post, let me simply point out that OH2 sitemaps are still supported by OH3. Semantic modeling allows you to build more complex visualizations but they do not replace the old sitemaps.

Hi Lionello, thanks for that clarification! I feared I had to ditch sitemaps (which took a lot of effort to design. It seems so far to me that these ‘cards’ or ‘tiles’ don’t provide the same UI experience but I’m still experimenting. They seem like those ‘dashboards’ that other systems offer but are more difficult to design with the mesmerising choices of classes.

OH3 is a significant step compared to OH2, it requires to user to understand new concepts. I keep my production in 2.5 and familiarize with OH3 with a smaller subset of my items.

My query has been answered/clarified in a ‘solved’ post Alessio Alexxio in late Dec 2020 titled “Missing Blinds Badge”. He had a similar concern about how/what semantic choices affect badges. It effectively answers my question and it should probably be included in the docs/tutorials somewhere as an excellent clarification with its screenshots!