Re: Describing RPG IV

["Chris Pando" <chris.pando@xxxxxxxxx>]

On 2/12/07, James H H Lampert <jamesl@xxxxxxxxxxx> wrote:
> You seem to have pulled a "Dr. Watson" on this one, having
> missed nearly everything of significance that
> differentiates RPG from other languages.
>
> How about:

... snip ...

> the extreme difficulty in accessing files whose names and
> structures are not known at compile-time

I do this all the time. Before calling the program, I'll do a DSPFFD
(of the 'unknown file) to an outfile, and then open the file as
program-described, mapped to a buffer. I get the field names and
offsets from the outfile, and use the field names to drive
table-driven logic. I am currently resequencing employee number in
JDEdwards implementation (to accommodate an acquisition), and use only
one program to
update all of the files. It isn't difficult, much less extremely difficult.

I have a generic file comparison tool (matching record, it will print
out differences between two files with identical layouts) that does
not know the names/structures of files at runtime. I have found it
invaluable for regression testing.

www.brilligware.com/cp1030.html

Table-driven logic is a different way of thinking, but it really isn't any more
difficult.



As for a description of RPG/IV, I'd just call it C with native I/O.

Chris
--