|
I haven't got anything that would not allow it to be pushed. Please go ahead.
The behavior is still weird in some ways, e.g. it is strange when you can do insert (f) values (default), but select default(f) causes an error, as in
MariaDB [test]> create or replace table t (a int, s bigint unsigned as row start, e bigint unsigned as row end, period for system_time(s,e)) with system versioning;
|
Query OK, 0 rows affected (0.074 sec)
|
|
MariaDB [test]> insert into t values (1,default,default);
|
Query OK, 1 row affected (0.010 sec)
|
|
MariaDB [test]> select default(a), default(s), default(e) from t;
|
ERROR 1364 (HY000): Field 's' doesn't have a default value
|
But it's not weirder than it was before (and still is with "regular" virtual columns), when select default(f) returns something completely different from what insert (default) inserts, so it can't be an obstacle.
|