Basic Idea:
TLDR: Columns which are both user-editable and calculated.
Something similar to Value for new rows which is dynamically applied to any blank cell in that column. (* with some caveats, see “Bonus desired features” below)
This would be a game changer for me, but curious what others think.
Example use cases:
- a generated Name column for a given table which includes (say) an emoji based on a “type” column, and some combination of other columns - but with the ability to override this with manual text
- a relationship column which includes a set of inferred values but which can be overridden manually
- a rich text field which is generated, but if edited becomes a concrete value (ie. even if the formula for defaults is changed, this field would be unchanged unless you cleared the cell’s content)
Some (perhaps less obvious) benefits:
- Currently if you have a calculated
Namecolumn for a table you lose the convenient ability to create new rows inline from a reference column (Works fine with non-calculatedNamecolumns). With this feature you’d get that same helpful experience even for tables with dynamic name columns. (Still not as nice as having alternate display types for rows, but that’d be a separate feature request)
Bonus desired features:
- For any column type which is a collection (eg. relationships or selects with “allow multiple” enabled) include a toggle which allows specifying that the calculated default values can’t be removed. So users can amend items but can’t remove any of the calculated ones.
- Allow Blank to be a valid override value, so if a given cell has the inferred content
[@A, @B, @C]I can hit delete on the cell to remove the default content. (This seems like it might be a more breaking change than the rest of this post though, and probably has some nefarious UX consequences which would need finesse)