Things that could be done to DBMS -- 12 October 1998

 Query optimizer:

 	Optimize for clustering indexes (use S/T and G/H plans)

 		e.g. assume Auto-increment column => clustering

 		or measure the clustering in the stats

 	Optimize for multi-column lookup

 		e.g. For "a=v1 and b=v2" with index on (a,b)

 	Optimize for like-predicates

 		e.g. "x LIKE 'abc*'" => "x>='abc' AND x<'abd'" etc.

 	Query complexity-reduction by removing redundant predicates, e.g.

 		"x>4 AND x>3" => "x>4"

 		"x>4 OR x>3" => "x>3"

 		"x>-1" for unsigned x => trivially true

 	Better statistical data dn guesswork

 SQL/access plans:

 	Support "DISTINCT" in the projection stage via a Hash-system or sorting