Here\'s a question I\'ve been racking my brain over. Let\'s say I have a table that has a series of timestamps and a part number as the primary key. The table stores incremental
ruakh is right. Alternative: Write a user-defined aggregate using SQL-CLR. This aggregate can run bottom-top-top over your rows and remember the first non-null value of each column.
You can also do this in the application. Make your program request rows in sorted order in batches of, say, 10 rows. Aggregate these batches like described above. If after the current batch is done a null-column is there get the next batch.