Sometimes a simple form is the wrong UI. If the user needs to compare many values and make careful edits, a table can be kinder than a long page of inputs.
Dense UI has a bad reputation when it is used to hide messy thinking. But some work is naturally dense. A user may need to review many rows, compare columns, edit a few values, and then save everything as one deliberate batch.
In that workflow, the table is not decoration. It is the workspace. If I reach for a table helper, I want something like TanStack Table to stay headless enough that the product still owns the editing model.
The useful details are small:
Rows changed: 4
Invalid rows: 1
Save button: disabled until errors are fixedChanged cells should be visible. Invalid cells should point to the problem. The user should know whether they are looking at saved data, unsaved edits, or a failed save.
I like batch save for this kind of workflow because it gives the user a review moment. Immediate save can be good for simple settings, but it can be risky when the user is editing several related values. A batch save lets the app validate the whole change before committing it.
The hard part is state. A table editor needs to know:
- the original value
- the current edited value
- whether the cell is dirty
- whether the row is valid
- whether the save is in progress
- what failed if the save did not work
If that state is vague, users lose trust. They start asking whether the app saved their work. That is a product failure, not just a UI bug.
Keyboard flow also matters. If the user is editing repeated values, mouse-only interaction becomes slow. Tab, Enter, Escape, copy, paste, and predictable focus can make the difference between a tool that feels usable and a tool that feels decorative.
For more advanced editors, the same principle applies. A rule editor or structured text editor, whether built on CodeMirror or another editor component, should not just accept text and hope for the best. It should help the user understand valid structure, show errors before execution, and make risky actions explicit.
The trade-off is complexity. A dense table can become a small application inside the application. It needs accessibility review, performance care, careful validation, and test coverage around dirty state. If the workflow only edits one or two fields, a normal form is probably better.
I would choose a dense table when:
- comparison across rows is part of the task
- users edit more than one item at a time
- validation depends on the whole batch
- keyboard speed matters
- mistakes are costly enough to justify a review step
I would avoid it when a simple form would be clearer. More cells do not automatically mean more power.
The goal is not to fit more data on screen. The goal is to reduce mistakes for users whose work already has many details.