i am kind of struggling with tables with a lot of data that need to be editable. The table holds financial data. The table has about 12 columns for which at least 10 need to be editable.
For the sake of ease the client prefers making them inline editable but there is a possibility more columns will be needed in the future and i'm afraid for the general scalability of this approach. Moreover, the application now is mainly for desktop but will in the future also be more and more used on a tablet (and i'm thinking portrait).
The other approach would be that the rows are not editable and that, for updating, you'd select a row and edit it in a separate inputfield region above. Same would be done for creating and duplicating.
Third option would be a mixture of the two: as a user you'd be able to edit some cells in the table but for creating a new or duplicating one you'd select the row and go into a separate inputfield region. Problem here is that the inputfield region above would always be visible and when you click on a row the data in that row gets displayed in the inputfield region but the editable cell (if so) would also show.
The thing is a lot of times the client wants to edit only one field and prefers the ease of inline editing.
I myself prefer the second approach but would like to hear some opinions from you guys. What would be your best practice?