> Right you are. I guess I got the group by mentality because I don't
> really care what order they're in--just so long as records with the same
> values for certain columns are together.
>
> But this still leaves me a "column ambiguously defined" error when using
> addJoin():
>
> crit.addJoin(TableAPeer.ColumnX, TableBPeer.ColumnG);
> crit.addAscendingOrderByColumn(TableBPeer.ColumnQ);
> crit.addAscendingOrderByColumn(TableBPeer.ColumnR);
> List<TableB> tableBs = TableBPeer.doSelect(crit);
>
> generates
>
> SELECT <..TableB fields> FROM TABLE_A, TABLE_B, DBSCHEMA.TABLE_B WHERE
> TABLE_A.COLUMN_X=TABLE_B.COLUMN_G ORDER BY TABLE_B.COLUMN_Q ASC,
> TABLE_B.COLUMN_R ASC
>
> Where does the extra DBSCHEMA.TABLENAME come from? Did I miss that in your
> original response?
>
It appears this is glued on by the presence of
torque.dsfactory.programs.schema = DBSCHEMA
in the Torque.properties file. The docs say that this is only used
a) by Oracle
b) for the JDBC->XML and SQL Ant tasks
It seems (b) is false--it alters the way doSelect() works (by prepending
its value to the table name. Is this a bug?
Brendan
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
