When "Clone Cursor" is turned OFF, and a user goes to the "save as" screen and begins an export, we use the actual cursor which is tied to the data grid. The advantage to this is that the query does not need to be re-executed. The disadvantage to this is that the whole dataset must be held in your PC's RAM, because this is a scrollable dataset (it is this mechanism that allows scrolling in the grids). When "Clone Cursor" is turned ON, and a user goes to the "save as" screen and begins an export, we create a new, non-scrollable cursor. The advantage here is that as rows are read in and sent to the destination file, we don't have to hold them in memory any longer...and a minimal amout of RAM is used. The disadvantage is that for this to happen, the query must be re-executed. So...which setting should you use? If your query returns a LOT of rows (too many to hold in your PC's RAM), you should have "Clone Cursor" turned ON...even if your query takes a long time to execute. If your query returns a number of rows such that the entire dataset will easily fit into your PC's memory, then we should consider the execution time for the query. If execution time is slow, then leave "Clone Cursor" OFF. If execution time is fast, then it doesn't really matter.

This option has been renamed to "Display all exported results in grid" for version 9.1. It will be greyed out if the dataset being exported has already been completely fetched, e.g. OCI Array = 500 and only 125 records, or you have already scolled to the bottom of the result set. If either of these cases are not true, then if this option IS set, Toad will execute the statement again and fetch ALL of the rows for the export, and then also send the entire result set to the original data grid so you can see them there too after the export has finished.