When I wrote the video script last week, I paid attention to how the introduction of features could be connected to a cohesive story. Based on that story, I built another prototype and sent out to experts to gather feedback.
Two things need to be fixed:
The “bookmark” space at the left could be used to show parent and children of the current card. (Dimension of a knowledge graph)
The webpage reference process need to be clarified. Because webpages organize content in intricate ways (e.g. wrapping multiple text blocks in layers of components), both object selection (native iOS support) and screenshot-like annotation need to be supported. Apps like LiquidText support both selection methods. But can they exist as more intuitive interactions?
The design will be implemented into a high-fidelity prototype, presented as animated video.