The job assignments csv file ties the organization hierarchy, position hierarchy, and manager hierarchy into the user through each job assignment. It also defines when a manager would have access to that user, and their training records. That is it's function, based on my understanding with our implementation.

I suppose basically speaking it could only be used for the managerial role as the organization and positions information is not required.

In my experience, and someone please correct me if I'm wrong, the idnumber field has at least three purposes in the csv. They are: 

1) it serves as the id for that particular job assignment in your csv generation; 

2) it serves as the key for the manager assigned, use it as the managerjaid. As HR Import requires the managerjaid to exist in the file.

3) it also serves as the identifier in the mdl_job_assignment table that HR Import will use to update the record on each import

But that is as far as it goes. Since that id isn't the key in the actual table.

I'm not sure how it would impact the order, as I haven't seen it having an impact. The number just has to be unique for each record. In our csv the numbers are not in sequential order.

Hope my ramblings helped.

John