===> even if you are programming to the OPM using MI,
there is
a translation of the result to NMI which is then given
to
CREATE MODULE for generation of the
RISC-instructions.
Leif,
I am alluding
to the 2 step process that a programmer has to execute to create an ile
program. 1st step is create the module, 2nd step is create the
program.
This lengthens the learning process of a
new programmer, increases the chances for bugs, ups the complexity all
around.
And for what reason? Increased execution speed? Performance can be
improved without increasing complexity by just shipping faster
hardware.
The much needed and relatively well done improvements in rpg that
came with ile did not require a new programming model or a link
edit.
===> fads come and go. Silver bullets shine and then
tarnish.
When the S/38 and the AS/400 came out, the OPM model
of
dynamic program calls (the CALLX instruction) was touted
as
a brilliant innovation, making link-edits obsolete,
allowing dynamic
replacement of modules (actually the ultimate in modular
programming). Then IBM discovered C, and the
BINDER,
and link-edited, bound programs were the holy grail.Said
to
faciliate 'modular programming' of all
things!
Then IBM discovered Java, and
dynamic binding was the rage and
a great innovation, etc, etc, etc. Then 'IBM is ready
for Linux" and
maybe we are back in C-land.
The pendulum swings back and forth,
and the clueless swing with it.
The ILE program model is incredibly complex. Some
mandated by
the segmented nature of the 'single-level store', but
most of it
unneeded, clearly bearing the marks of the 'second
system syndrome'
that Fred Brooks talked about.
Quoting dr. Frank again:
"Figure 9.5 shows how these components fit together to
create
an activation group. To summarize, we can say each
process
in the aS/400 has a PAWA. Within this PAWA id the PAGP
and
two or more ACTGRPs. Each ACTGRP has a PACB that
contains
multiple MBVs, an ActGrp directory, a PRT, a heap list,
one or more
heap spaces, auto storage segments, and static storage
segments.
Now, isn't that perfectly clear?"
Even Frank knows that something isn't quite right here,
hence the
almost apologetic quip about 'perfectly clear'.
Sigh.
-------
Hurry up with the nmi chapter of your
book! Write faster!!
===> working on
it...
midrange.com
