Description:
There's a table called g：
CREATE TABLE `g` (
`id` int(11) DEFAULT NULL,
`a` int(11) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;
insert into g values(1,11);
insert into g values(1,22);
insert into g values(1,33);
insert into g values(2,11);
insert into g values(2,22);
insert into g values(2,33);
insert into g values(2,44);
insert into g values(3,11);
insert into g values(3,22);
insert into g values(3,33);
insert into g values(3,44);
insert into g values(4,11);
insert into g values(4,22);
insert into g values(4,33);
insert into g values(5,11);
insert into g values(5,22);
insert into g values(5,33);
insert into g values(6,11);
insert into g values(6,22);
insert into g values(6,33);
insert into g values(6,44);
insert into g values(7,11);
insert into g values(7,22);
insert into g values(7,33);
insert into g values(8,11);
insert into g values(8,22);
insert into g values(8,33);
insert into g values(8,44);
insert into g values(8,55);
insert into g values(9,11);
insert into g values(9,22);
insert into g values(9,33);
insert into g values(10,11);
insert into g values(10,22);
insert into g values(10,33);
insert into g values(10,44);
Execute the following statement：
mysql> select id,group_concat(a) from (select * from g where id=3 order by id,a ) t group by id;
+------+-----------------+
| id | group_concat(a) |
+------+-----------------+
| 3 | 11,22,33,44 |
+------+-----------------+
mysql> select id,group_concat(a) from (select * from g order by id,a ) t group by id;
+------+-----------------+
| id | group_concat(a) |
+------+-----------------+
| 1 | 11,22,33 |
| 2 | 11,22,33,44 |
| 3 | 44,33,22,11 |
| 4 | 11,22,33 |
| 5 | 11,22,33 |
| 6 | 44,33,22,11 |
| 7 | 11,22,33 |
| 8 | 55,44,33,22,11 |
| 9 | 11,22,33 |
| 10 | 11,22,33,44 |
+------+-----------------+
Why do two statements with id=3 return different results？
How to repeat:
Create the above test tables and data to replicate.
