Thus it was written very recently in the epistle of Ted Ashton,
>
>   It might be more efficient to DBGET the master and then DBDELETE, expecting
> that most DBDELETEs will return 44.  Then again, it might not, depending on
> whether a DBFIND or a "failed" DBDELETE is more expensive.

From my testing here, it appears that they are about equally expensive, so I'm
likely to go with the DBGET/DBDELETE method, as it avoids some complexity and
a third call in case a delete is needed.  Advice still very welcome.

Thanks,
Ted
--
Ted Ashton ([log in to unmask]), Info Sys, Southern Adventist University
          ==========================================================
In an examination those who do not wish to know ask questions of those who
cannot tell.
                        -- Raleigh, [Sir} Walter Alexander (1861-1922)
          ==========================================================
         Deep thoughts to be found at http://www.southern.edu/~ashted