I couldn't understand the statement of the puzzle until I took a peek at the solution (v5) given by Leszek Gniadkowski. My solution is re-written to show my understanding of the puzzle since, thanks to Leszek Gniadkowski.

This is my third attempt to solve the puzzle. This solution is slightly simpler than the previous two.

This is my third attempt to solve the puzzle. This solution is slightly simpler than the previous two.

This is a second attempt to the solution that is slightly simpler than the previous one but with no change in the main logic.

My solution displayed: ProductID Start_Date End_Date TotalLines LinesOutOfService TotalCustomerCalls 522 2010/04/05 2010/04/06 345 5 180 522 2010/04/07 2010/04/07 120 4 50 522 2010/04/08 2010/04/10 345 5 170 5

Copy the given sample data into a cte. Process the start date and end date as directed in the puzzle. Sum up the CustomerCalls for each group.

The results: Date CustomerID Type Amount Balance 2011-01-01 CUST1001 INV 12000.00 12000.00 2011-01-02 CUST1001 PAY 3000.00 9000.00 2011-01-03 CUST1001 INV 8000.00 17000.00 2011-01-04 CUST1001 PAY 9000.00 8000.00 2011-01-04 C

This solution is contructed in sql 2012 evaluation edition making use of the OVER clause of the SUM function and also the CAST function for datetime to date format.

This solution is contructed in sql 2012 evaluation edition making use of the OVER clause of the SUM function and also the CAST function for datetime to date format.

Re-submitted without html tags/scripts The running total is obtained by using a simple recursive CTE