Re: QUSLFLD

[Evan Harris <spanner@xxxxxxxxxx>]

I would add to this that I have seen a couple of occasions where the IBM
reference files were "out of touch with reality" requiring a RCLSTG DBXREF;
I presume this would also impact using this file as the basis of your
procedure whereas IMO the API would continue to work. Admittedly an unusual
circumstance, and maybe I have misunderstood what I actually rebuilt.

Also, in line with James' comments I would be nervous about using an
undocumented/informal method as opposed to an officially support interface.
It seems to me this has similarities with hard-coding off-set or other
values in API calls.....

Just another opinion

Regards
Evan HArris

<SNIP>
>My concern when using QADBIFLD is that IBM could change the format in
>a future release, causing the program to stop working..   (Or have they
>documented that this won't change?)
>
>You wouldn't have that problem with QUSLFLD...  it's designed so that
>they can make changes and still keep things backwards-compatible.
>
>On Fri, 4 Jan 2002 rob@dekko.com wrote:
> >
> > My stock and standard reply to this is that why use DSPFFD or QUSLFLD when
> > you can use traditional i/o against the file QSYS/QADBIFLD to get this
> > information?
<END SNIP>