Re: Utility

-----John Yeung <gallium.arsenide@xxxxxxxxx> wrote: -----

My point is that sjl seemed to be saying that CLPs from older
systems
can always be restored on newer systems and run perfectly fine. But
this isn't true.

Ok, almost always.

I had a secondary point which is that on systems other than IBM i,
the
issue of "restoring CLP objects" wouldn't even come up, because on
those other systems, people use shell scripts to do what most i
people
use CLPs for.

The things that would cause a CLP to run differently, for example exporting a spool file into a PF and parsing it, will also cause the script to fail. Why because the format of the spool file has changed, and it has changed for the script as well. If you are talking about the CLP calling a poorly written program, well, once again the script calling the same program will fail in the same way. There is nothing magic about scripting languages that makes them more fail-proof.

If everyone used Rexx scripts instead of compiled CLP objects, then
sure, we wouldn't even be having this discussion. (Though honestly,
for scripting, I think most people will find it easier to deal with
stream files than with members within source physical files.) If
everyone developed all their IBM i apps in IFS, with DB2 "merely"
serving as the database (which, granted, some people do, and it
seems
more and more are doing), then we also wouldn't be having the QSYS
vs.
hierarchical directory structure discussion either.

In my experience, this is pretty much totally false, and the hierarchical directory structure discussion is all white noise. People trying to pound a square peg into a round hole.