Ron Mayer <rm_pg(at)cheapcomplexdevices(dot)com> writes:> Should the optimizer have found this optimization?
I can't get excited about it. Joining on a column that's mostly nullsdoesn't seem like a common thing to do.
regards, tom lane