Have a data set with a 'parent' and 'child' column, storing paths through a tree (see http://fungus.teststation.com/~jon/treehandling/TreeHandling.htm for the modelling). At the moment, there's a stored proc that takes exactly one ID and builds a data set like the below. Trying to neatly re-write this as a view, so that I can simply add in the "…

gistfile1.txt

SELECT 
 * 
FROM data 
WHERE id IN(
  SELECT @id
    UNION
  SELECT child_id FROM paths WHERE parent_id = @id
)