Misleading N:N Interpretation in "Employees" Example Due to Overlapping Relationship Lines

Description:

While exploring the "Employees schema" example in ChartDB: https://app.chartdb.io/examples

I noticed a potential issue with how relationships are visually represented, particularly between the following tables:

• employees
• salaries
• titles

In this schema, both salaries and titles are independently related to employees (1:N relationships). This is consistent with a typical HR database design, where each employee can have multiple salary records and multiple titles over time. ([chartdb.io]1)

However, in the current visualization, the relationship lines between these tables overlap or intersect in a way that creates ambiguity.

Observed Issue:

From a visual (human) perspective, the overlapping lines can be misinterpreted as implying a direct relationship between salaries and titles.

This can lead to an incorrect assumption that:

• salaries ↔ titles form a many-to-many (N:N) relationship

In reality, no such relationship exists in the schema. Both tables are only connected through employees.

Why This Matters:

This issue is not about incorrect schema logic, but about visual misrepresentation.

Because ChartDB is designed to help users understand database structures quickly, this kind of ambiguity can:

• Mislead beginners or students learning database design
• Cause incorrect assumptions during documentation or system analysis
• Reduce clarity in presentations or team discussions

Expected Behavior:

The visualization should make it clear that:

• salaries and titles are not directly related
• All relationships are mediated through employees

There should be no visual cues that suggest a relationship where none exists.

Suggested Improvements:

• Improve auto-layout to reduce or avoid overlapping relationship lines
• Introduce clearer directional arrows or relationship labels
• Allow optional “clarity mode” that prioritizes readability over compact layout
• Provide visual separation or grouping for hub entities (e.g., employees as a central node)

Additional Context:

This appears to be a common visualization issue when representing ternary-like structures (three related tables sharing a central entity). The current layout unintentionally creates a “visual shortcut” that humans interpret as a direct relationship.

---

Thank you for building ChartDB — it’s a very powerful and useful tool. Improving this aspect of visualization would significantly enhance its usability for learning and communication.