SQL实现分组排序和组内排序

在对表做排序时，经常会遇到需要先按某一个列排序，再按这个列分组的统计值来对子组或者说其它列排序的需求，下面是一个典型案例实现的技术路径：

利用Mysql find_in_set group_concat

select * from user order by find_in_set(name,
 (select group_concat(name order by time desc)  from 
 (select name,max(time) as time from user group by name) as tt),time desc;

利用相关子查询

SELECT * ,(SELECT MAX(time)  FROM user a WHERE a.name=t.NAME)  AS mm
FROM user t
ORDER BY mm desc,t.time DESC