Re: Logical file order of records

FWiW there is no legitimate "logical file order of records" [the subject] in SQL, so it seemed intuitive to me from the subject that the issue was non-SQL. SQL does not officially allow specifying an LF [SQE], nor does a query honor LF key specifications [CQE; excepting perhaps an OPNQRYF ODP with KEYFLD(*FILE)].

I am not sure the relevance of what the original specifications were or were not for the file, given the text I quoted. I was merely clarifying that the "the key data will be organized as requested" may include LIFO, FIFO, & FCFO, that these do not apply to SQL, and that a non-query\SQL interface may be able to get a consistent logical order; something which the OP was not effecting, presumably for lack of any duplicate-key-ordering specification, but which could be effected given the nebulousness of /copy/ and /exact/ were defined and corrected... but that was in a separate post.

Regards, Chuck

Dennis Lovelady wrote:

The database does include in that /complexity/, the ability to specify the ordering of duplicate key values; e.g. LIFO, FIFO,
FCFO, in DDS. Not all accesses to database data are via query,
so realize that non-SQL & non-query interfaces used to
read-by-key can actually effect a read-order to reflect
whatever was specified as the [build]duplicate-key-order
rule(s).

I didn't notice that the originator specified that DDS was
involved, and responded with SQL on my mind. Besides, if the
files had had that specified, I don't think the question would
have been asked. Right?