On Sun, Dec 13, 2015 at 10:23 PM, Bill Moran <wmo...@potentialtech.com> wrote:
> On Sun, 13 Dec 2015 22:15:31 -0500 > Corey Huinker <corey.huin...@gmail.com> wrote: > > > ALTER TABLE foo DISABLE [NONUNIQUE] INDEXES > > -- same, but joining to pg_class and possibly filtering on indisunique > > I would think that NONUNIQUE should be the default, and you should have > to specify something special to also disable unique indexes. Arguably, > unique indexes are actually an implementation detail of unique > constraints. Disabling a performance-based index doesn't cause data > corruption, whereas disabling an index created as part of unique > constraint can allow invalid data into the table. > > Just my $.02 ... > > -- > Bill Moran > I'd be fine swapping NONUNIQUE for ALL and defaulting to non-unique, or flatly enforcing a rule that it won't disable the index required by an enabled constraint.