3 ms·
If raw speed is not an issue you can always fall back on temp tables or a cursor for those cases where you might have extracted an initial dataset to then itera
by dwd 8y ago
If raw speed is not an issue you can always fall back on temp tables or a cursor for those cases where you might have extracted an initial dataset to then iterate over.
EDIT: A CTE (Common Table Expression) might be what you're thinking you need. You can do some fun recursive queries with them.