> From: "Ann W. Harrison"
> Certainly, constraints work as documented. Equally clearly, they
> don't follow the standard. The 92 standard recognizes two types of
> constraint, immediate and deferrable. ...

Yes, but original question of this thread was whether behaviour of check
constraints containing selects

> CREATE TABLE test_master (
> id INTEGER NOT NULL PRIMARY KEY,
> CHECK(COALESCE((SELECT max(id) FROM test_master), 0) < 10)

is possibly buggy because they do not behave as "expected".
So, even if Firebird supported sql92 standard, would that solve
the problem ?

Ivan