There are several things wrong here, which makes it a bit difficult to tell
what you want. Let's look at the subquery first.
SELECT idk FROM `mj_tmp`
GROUP BY Concat(ida,u)
HAVING Count(Concat(ida,u)) >1
First, you shouldn't use CONCAT() in your GROUP BY, as it breaks the
possibility of u
Thanks Jim,but that doesn't work for me because the order by
performs after the records have been selected.
At that time I already have the unwanted row.
I will try to explain a little bit further
SELECT *
FROM `mj_tmp`This select is suposed to return all rows
WHERE idk NOT IN but
> Hi , I'm trying to fetch the oldest row from a result
> set in wich the resulting rows have duplicates,
>
> I need only one row for each duplicate.
>
> The problem is that I need always the oldest row
> Like in the example
>
> SELECT *
> FROM `mj_tmp`
> WHERE idk NOT IN
> (SELECT idk FR
