Problem is simple - I need to get id (and other values by which I can not group) of rows with some maximum value.
Below two examples, which do the thing any suppositions to use a "better"/other way to achieve this? some kind of trick with analytic/aggregate function or anything else? thanks in advance Remigiusz --------------------------------------------------------- select r.id, r.wid, r.start_time, b.wid from ( select id, start_time, wid FROM wptv_listings l join ( select channel, max(start_time) stime from wptv_listingsWHERE visible='T' AND start_time BETWEEN <date1> AND <date2> group by channel
) m on l.CHANNEL=m.channel and l.START_TIME=m.STIME ) r JOIN wptv_blocks b on r.id=b.id select r.id, r.wid, r.start_time, b.wid from (select id, start_time, wid, rank() over (partition by channel order by start_time desc) rank1
FROM wptv_listings l WHERE visible='T' AND start_time BETWEEN <date1> AND <date2> ) r JOIN wptv_blocks b on r.id=b.id where r.rank1=1 -- --------------------------------------------------------------------- Remigiusz Sokolowski <rems@xxxxxxxx> WP/PTI/DIP/ZAB (+04858) 52 15 770 MySQL v04.x,05.x; Oracle v10.x Zastrzezenie:Niniejsza wiadomosc stanowi jedynie wyraz prywatnych pogladow autora i nie jest w zadnym wypadku zwiazana ze stanowiskiem przedsiebiorstwa Wirtualna Polska S.A.
---------------------------------------------------------------------WIRTUALNA POLSKA SA, ul. Traugutta 115c, 80-226 Gdansk; NIP: 957-07-51-216; Sad Rejonowy Gdansk-Polnoc KRS 0000068548, kapital zakladowy 62.880.024 zlotych (w calosci wplacony)
-- //www.freelists.org/webpage/oracle-l
