Den 2012-10-23 10:32 skrev Frank såhär: > select R1.RDB$RELATION_NAME > from RDB$RELATION_FIELDS F1 > join RDB$RELATIONS R1 on F1.RDB$RELATION_NAME = R1.RDB$RELATION_NAME > where F1.RDB$FIELD_NAME ='EMP_NO' and > (not exists(select R2.RDB$RELATION_NAME > from RDB$RELATION_FIELDS F2 > join RDB$RELATIONS R2 on F2.RDB$RELATION_NAME=R1.RDB$RELATION_NAME > where F2.RDB$FIELD_NAME ='PHONE_EXT')); Although I can't see why it would give the indicated error, the query seems utterly pointless. R2 and F2 are not correlated and so the subquery will return all F2 records paired with all R2 records. I think it's a typo and that the subquery's join condition should refer to R2, not R1.
Kjell -- -------------------------------------- Kjell Rilbe DataDIA AB E-post: kj...@datadia.se Telefon: 08-761 06 55 Mobil: 0733-44 24 64 ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_sfd2d_oct Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
