一條簡單的sql

[zhuqiang9]

id time type

a 2006-11-14 13:55:17.320 net
b 2006-11-14 13:55:54.980 c++
a 2006-11-14 13:57:04.963 c
b 2006-11-14 14:00:29.843 java
c 2006-11-14 14:00:51.763 sql
我想select a,b,c這3种編號最新的一條數據

[冷箫轻笛]

select * from tablename t1
where not exists (select 1 from tablename where id = t1.id and time > t1.time)

[子陌红尘]

select t.* from 表 t where not exists(select 1 from 表 where id=t.id and time>t.time)

[fish_yht]

select top 1 * from 表
order by time desc

[crazyflower]

select id,max(time) from tb where id in ('a','b','c') group by id

[fish_yht]

理解错误
不好意思

[jamesming]

要是LZ的id中有E，F，G，**，二楼的方法也行吗？

[中国风]

棒场!
declare @ta table ([id] varchar(2), [time] datetime, type varchar(10))
insert @ta
select 'a', '2006-11-14 13:55:17.320' , 'net' union all
select 'b' , '2006-11-14 13:55:54.980' , 'c++' union all
select 'a' , '2006-11-14 13:57:04.963' , 'c' union all
select 'b' , '2006-11-14 14:00:29.843' , 'java' union all
select 'c' , '2006-11-14 14:00:51.763', 'sql'
select * from @ta a where time=(select max(time) from @ta where id=a.id) order by time asc

（所影响的行数为 5 行）

id time type
---- ------------------------------------------------------ ----------
a 2006-11-14 13:57:04.963 c
b 2006-11-14 14:00:29.843 java
c 2006-11-14 14:00:51.763 sql

（所影响的行数为 3 行）

[中国风]

select * from @ta a where time=any(select max(time) from @ta group by id)
id time type
---- ------------------------------------------------------ ----------
a 2006-11-14 13:57:04.963 c
b 2006-11-14 14:00:29.843 java
c 2006-11-14 14:00:51.763 sql

（所影响的行数为 3 行）

[allright_flash]

select id,max(time) from tablename group by id