Use case or problem
Imagine a long note. It has many headings and paragraphs (blocks). Each of those sections could be considered its own note. They’re not since Obsidian now allows linking to them. You can have the best of both words. Longer notes while still having the ability to reference parts of them. The problem is that the graph view doesn’t reflect that. The graph view shows these links as if they were just links between notes.
The line going between a note and a heading/block should be dotted, dashed, waved or colored. Also, while hovering over this connection, instead of the note name the heading/block name should be shown.
Current workaround (optional)
Related feature requests (optional)
Another possible solution:
- Draw note which contains referenced blocks as hollow circle.
- Draw referenced blocks belonging to that note as filled blocks inside the circle.
- This would be less cluttered by lines indicating membership to note and “block-nodes” would not be spread across whole graph but collected inside “note-nodes”.
Link-line from hollow circle to any internal or external circle would exist only when one block is referenced from another which is not referenced.
Thickness of the circumference would correlate to number of such links divided by diameter of hollow circle (I guess, we percieve the area).
force-based graph layout algorithm could be solved independently for each hollow circle containing filled circles for blocks and would also minimize diameter of hollow circle.
Then already existing algoritm would consider only whole notes (not blocks) and it would take into account their diameter to avoid overlap.
Then links would be laid over to interconnect block-subgraphs and global notes-graph.
# note [[K]]
# note [[L]]
# note [[M]]
# note [[N]]
- B ^B
- C: [[#^B]]
- D ^D
keywords: block, contain, container, disjunct, enclose, encapsulation, graph, nested structure, venn diagram, visualization
Related requests (similar visualization for different purposes):
This is awesome and much more elaborate than what I suggested. Hopefully developers implement some of this.
Great idea: Still trying to wrap my head around all of it.
Perhaps a setting can be used to toggle hollow circles on/off. Off: old graph behavior. On: New graph behavior (your proposal).
This way the user can choose when they need to see more complexity.
Optionally: maybe hollow circles can be turned on/off on a per “note-node” basis.