README doesn't say whether web Compose targets canvas or html #40
Description
When evaluating the overall Compose platform, it'd be useful to know whether the web target takes an approach like Flutter where things are mostly rendered with canvas and don't use native elements, or whether it'd be a little more like react-native-web targeting semantic, accessible html5 elements, or something in between.
Can your intended approach be documented somewhere?