Felix Kater <[EMAIL PROTECTED]> writes:
> On Tue, 8 May 2007 15:54:08 +0200
> Martijn van Oosterhout <[EMAIL PROTECTED]> wrote:
>> A unique index is not a "substitute" for a unique constraint, they're
>> exactly the same thing.
> Yes. For this reason I didn't have to implement *both* 'unique
> constraints' *and* 'unique indices' in my pg interface.
If you are trying to get away with a dumbed-down subset of SQL, be
prepared for people to refuse to use your tool ;-).
You have to support the unique-constraint syntax because the SQL spec
says so (and people are used to it), and you have to support the
create-index syntax because it gives access to functionality not
available through the constraint syntax. Unique indexes on expressions
for instance.
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
