>SELECT Ksongs.iid, Ksongs.cartist, Ksongs.ctitle, Ksongs.dlastprinted, ; > TRIM(Ksongs.cartist)+TRIM(Ksongs.ctitle) AS ufilter ; > FROM ksb!ksongs ; > WHERE Ksongs.ccatalog = ?vp_cCatalog ; > AND EMPTY(Ksongs.dlastprinted) = .T. ; > OR (Ksongs.dlastprinted > ?vp_SelectDate) ; > GROUP BY 5 ; > ORDER BY Ksongs.cartist, Ksongs.ctitle >>This worked great in v7 but fails as invalid group by clause in v9. Purpose of grouping was to insure that multiple versions of same artist/song combination printed as 1 item. I can't make sense of the syntax explanation in the v9 help file. Can anyone suggest how I might accomplish my goal using the v9 engine bahavior? Thanks
SELECT min(Ksongs.iid) as iID, min(Ksongs.cartist), min(Ksongs.ctitle), min(Ksongs.dlastprinted) as dLastPrinted, ; TRIM(Ksongs.cartist)+TRIM(Ksongs.ctitle) AS ufilter ; FROM ksb!ksongs ; WHERE Ksongs.ccatalog = ?vp_cCatalog ; AND EMPTY(Ksongs.dlastprinted) = .T. ; OR (Ksongs.dlastprinted > ?vp_SelectDate) ; GROUP BY 5 ; ORDER BY Ksongs.cartist, Ksongs.ctitleBTW, in other languages you can not use newly created column in GROUP BY clause.