Re: Using CHAIN instead of SETLL AND READE to begin a loop which reads a group

Frank,

I think that it depends on the number of duplicate keys.

I've only had situations where there are only two or three records with
duplicates and AFAIK it has always retrieved the first record.

Although I like to think of a CHAIN as a SETLL/READE combination, I'm
not sure that the internals actually work the same way in determining
which record to retrieve.

I know that I use a binary halving to "position to" in a subfile and if
I had sufficient number of duplicates I would not always retrieve the
first occurrence of the key field.

Frank.Kolmann@revlon.com wrote:
>
> Looks to me like the issue here is
> 'What does CHAIN retreive when used on a file with duplicate keys.'
> Long time ago I was told by IBM that the record retreived by the
> CHAIN is unpredictable.
> Experience tells that the retreival is FIFO.
> Has anyone noticed anything other than the above.
>