Re: Displaying Journal Entries and File Format ID

[Vern Hamberg <vhamberg@xxxxxxxxxxxxxxxxxxxxxxxxx>]

SQL could provide a way to handle different formats, with its column 
renaming capability (field as newname) and its "case-end case" control 
structure. A UDF could be written to retrieve whatever controlling value 
you need, then use something like

case prodversion()
   when "Version1"
      then field1
   when "Version2"
      then field2
   else originalfield
end case
as originalfield

This way your original program uses the same field names. This can even 
handle data type changes with the various conversion (cast) functions.

HTH
Vern

At 08:21 AM 3/9/2004 -0800, you wrote:
> Interesting.  I'm as surprised as you are that there is no record format
> ID in the journal receiver.  Doesn't look like any easy way to
> automagically handle different formats.
>
> But then, your extraction app would have to be recompiled after every
> format change anyway, so I think you've got the right idea with using a
> cutoff timestamp.
>
> BTW, did you notice that one of the *TYPE5 fields is "Arm Number"?  It's a
> 5-byte zone-decimal field.  What in the world is this about?  InfoCenter
> describes this as "The number of the disk arm that contains the journal
> entry."  I am baffled as to the usefulness of this information.
>
> GA