Is it possible to do a select distinct using the ordinal position of a column in a table rather than the column name?
What happened when you tried it?<BR><BR>-Doug
As distinct is applied to all columns in the select list, the question is moot.
What I need to do is get the distinct records from a comma delmited file which in some cases might not have the column names on the first line..
What happened when you tried it?
if you have an ordinal number for a column you can get the name by using the ADO method openschema or you can query the information_schema views. (The latter is not supported by all dbms. SQL server and Mimer does though, for instance.)<BR>
How can you try something when there is no way of knowing the syntax for it? Btw, you still have not answered me either.