Thread: null = null
the following took me by suprise: egg=# select null=null;?column? ----------t if this were true, for example, then only one null key could be in a unique index, which is certainly not true. i thought the philosphical idea behind null in SQL was you didn't know, which can't have equivalence (as versus the pragmatic idea in C where null is a dead end pointer, which can have equivalence). or have I been too involved with my Oracle databases? __________________________________________________ Do You Yahoo!? Send FREE video emails in Yahoo! Mail! http://promo.yahoo.com/videomail/
chester c young <chestercyoung@yahoo.com> writes: > the following took me by suprise: > egg=# select null=null; > ?column? > ---------- > t This is a Microsoft-compatibility kluge that is going away in the next release, or at least not being enabled by default. regression=# select version(); version ---------------------------------------------------------------PostgreSQL 7.2b5 on hppa-hp-hpux10.20, compiled by GCC 2.95.3 (1 row) regression=# select null = null;?column? ---------- (1 row) regression=# select null IS null;?column? ----------t (1 row) regression=# set transform_null_equals to 1; SET VARIABLE regression=# select null = null;?column? ----------t (1 row) See the 7.2 docs at http://developer.postgresql.org/docs/postgres/functions-comparison.html regards, tom lane