>>It may also be used in dynamic cursor processing, e.g. you collect some entries to a cursor (e.g. within a loop) and then, when loop is over, you want to go through the cursor picking up only unique entries. The main problem of UNIQUE index is that it's not maintainable but for temp cursor it does not matter.
>
>
>Agreed. There is probably always an SQL equivalent but setting a UNIQUE index may be more straightforward.
>
>Peter
Yes, there are at least two more ways to do the same thing:
1.SELECT DISTINCT against the cursor
2.Regular Index on cursor and SEEK before each INSERT.
Probably, any way is good enough. You see, I take chances again :-)
Edward Pikman
Independent Consultant