Yes, I think it's really for backward compatibility. I'm not sure if this is still true, but when unique indexes were first added to the product, their "uniqueness" was not even maintained. IOW, when you first create the index or REINDEX it contains only one entry for each distinct key value. However, as records are added or modified the uniqueness is not maintained. Pretty useless as far as I'm concerned. I'm sure some people have found some clever uses for them, though.

>Does anyone know of a valid reason for using the Unique index type? If there are duplicates, you should use Reqular and if they are truly key fields then Primary or Candidate should be used. (IMNSHO)
>
>Is it there for backward compatibility? There's nothing I've found to that effect.
>
>Thanks,
>Rock