On Wed, 21 Apr 2021 at 22:32, Peter Eisentraut <[email protected]> wrote: > Add DISTINCT to information schema usage views
What do you think of the idea of just getting rid of all these DISTINCTs and instead do a semi-join so that we don't get duplicate rows? For me, I don't really like the DISTINCTs as the behaviour is prone to change as the target list changes in the view. Maybe these views are standard enough that that's not going to happen very often, but it could. The 2nd reason I'm not a fan of using DISTINCT is that it's really chopping the problem off at the tail instead of the head. I've attached a patch that gets rid of all the DISTINCTs. David
use_exists_instead_of_distinct_in_information_schema_views.patch
Description: Binary data
