One of the reasons for choosing a numbered list is that it supports duplicate combinations of properties. Does anyone know how you can import into a numbered list from a CSV containing duplicate rows? It is not acceptable to have to add anything to the CSV prior to import. For example the CSV has 3 columns headed List1, List2, Num1, with two data rows List1_1, List2_2, 1 and List1_1, List2_2, 1. The numbered list has Properties List1, List2 and Num1. If I were to Insert 2 entries into this list and then manually set the properties for each entry I could achieve the desired outcome, so why can't I do the same with an Import?
LIst imports are designed to support both creating and updating data - very often it's a mixture of both, sometimes known as 'upserting'. Some rows might fail, but if you correct the source data and re-import then only the previously-failed rows will be added to the list. Yes, you can denormalise the data manually (and use different composite keys for different imports), but most administrators spend a lot of effort trying to keep their master data manageable - and that usually means normal in the relational sense. If two items in a list are the same, there's rarely much value in them being distinct items, and quite a few disadvantages - you can't individually identify them for update without extracting the internal #numbers and pinning them back to the source somehow, for example.