Re: Structured Programming Software Packages IMHO

<DISCLAIMER>
I am a former SSA employee that, among other things, taught AS/SET to
customers on-site and in a classroom setting.
</DISCLAIMER>

Yes, you are at a disadvantage if you do not use AS/SET when looking at BPCS
code.

Practically any higher level programming tool does things incomprehensible
to folks that program in the "lower" level language.  The same is the case
with AS/SET.  Especially if the AS/SET coding is poor.

If someone looked at the MI generated by the RPG compiler, and compared it
to the clarity that can be achieved when coding directly to MI, the same
things you mentioned would occur.  Or if I don't speak Spanish, but I have a
translator do the work for me, the result will not be as elegant as if I
spoke spanish directly to my target audience.   It's just the nature of the
beast.

All AS/SET does is translates simple, high-level constructs into RPG "code
snippets" for compilation.  (I know this is a gross simplification...)

Now, I've seen some pretty bad AS/SET coding in BPCS.... It gets much worse
if the AS/SET coder has no concept of structured programming technique.  But
if the AS/SET coder uses standard, structured constructs, I would bet that
the resulting RPG code is as good or better than 75 to 80 percent of what
the RPG coders can produce.  (There are good and not-so-good coders in every
langauge.)

The advantages of AS/SET are the speed of development and structure.  As far
as I'm concerned, those are the only advantages.  But it is enough when you
consider the size of the total ERP package.  Not getting the development
tool used for an ERP package in "penny wise and pound foolish."

Just my $0.02.

William


----- Original Message -----
From: <MacWheel99@aol.com>
To: "BPCS Users Discussion Group" <BPCS-L@midrange.com>
Sent: Saturday, June 03, 2000 3:52 AM
Subject: Structured Programming Software Packages IMHO


> Is it BPCS, or me not having access to AS/Set, or still programming from a
> green screen work station, or my middle age?  (I started programming in
the
> early 1960's & am now in my mid 50's.)
>
> BPCS 405 CD = RPG/400 on OS4 V4R3
>
> It seems to me that BPCS/400 implementation of object oriented programming
> has taken structured programming backwards at least 20 years.  In the
early
> 1980's I was programming MAPICS modifications which went to points of
> absurdity with embedded soft code & I thought was more obtuse than the
> symbolic machine language I was writing in the late 1960's.
>
> But I had no idea to what extremes of soft code abuse were possible until
I
> found BPCS using message members for every other line of some RPG programs
&
> a higher proportion than that in some DDS with no annotation of what their
> function is.  Yes I know how to look up what each one is, but when there
are
> scores on each page of a program with hundreds of pages, I need a tool
that
> will automatically illuminate each of those soft coded whatsits as I
scroll
> thru PDM & throw off to the side what it is, like when we move our cursor
> over some M$ windows or browser icon we get in tiny print what the
function
> is.
>
> I have generally had the expectation that as platforms advance, the tools
of
> our trade will be come more programmer-friendly & the ability to get great
> power out of our languages without any obtuse coding, that becomes more &
> more straighforward for a maintenance programmer to be able to comprehend
> what is going on.
>
> Some BPCS programs are well organized, structured intelligently, but  H U
G E
>   mind boggling, while others seem to be an RPG emulation of machine
language.
>
> Programs call programs seemingly to infinitum abandon ... it is like a new
> dimension of spagetti GOTO from early days of Basic before purists said no
> more GOTO because so many people were abusing it.  When I look at IBM
compile
> x-index charts there is no section or coding specifically listing all the
> stuff called by this program, like it does for subroutines & exception
output.
>
> I would like to have a hierarchical chart of who calls whom that is
something
> like a multi-threaded internet discussion forum.  The cross-reference
stuff
> from plain vanilla IBM is not really very intelligible.
>
> It used to be only the contents of a work field could be variable but with
> *LIKE field size has to be looked up in the compile x-index because it is
> like a field like another field like many links to original story & I am
> crying out enough ... I want this compile to tell me way over on the right
> hand side what the dimensions are of the fields being defined or used on
this
> line, so I can see if they are appropriate to the task at hand.
>
> As bad as this pretence at a high level language is, it is not as bad as
M$
> Win which I consider to be a GUI version of symbolic machine language by
> folks who missed the computer science class on what is a high level
language.
>
> Am I out of step with reality by expecting that software be developed so
that
> maintenance programmers have a prayer of deciphering the code in some
> reasonable time period?
>
> The trigger for this diatribe is I just spent a second nite of 10 hours
> digging into the JIT & SFC 600 610 620 series of programs trying to figure
> out what is causing them to intermittently invent lot control (which we do
> not use) batches within batches of transactions in which the original
correct
> transactions are misplaced & replaced with 2-3 bogus transactions in
> non-existant locations in a different facility than where those items are
> supposed to be.
>
> We have back traced this scenario to 20 other cases that date back to ....
> around the time in late 1999 I was merging 4,000 BMRs in 3 months with our
> modifications & congratulating myself on how well it was going ... we had
> delayed REL-2 until we learned here that it had Y2K fixes.
>
> Al Macintyre  ©¿©
> http://www.cen-elec.com MIS Manager Programmer & Computer Janitor
> +---
> | This is the BPCS Users Mailing List!
> | To submit a new message, send your mail to BPCS-L@midrange.com.
> | To subscribe to this list send email to BPCS-L-SUB@midrange.com.
> | To unsubscribe from this list send email to BPCS-L-UNSUB@midrange.com.
> | Questions should be directed to the list owner: dasmussen@aol.com
> +---
>



+---
| This is the BPCS Users Mailing List!
| To submit a new message, send your mail to BPCS-L@midrange.com.
| To subscribe to this list send email to BPCS-L-SUB@midrange.com.
| To unsubscribe from this list send email to BPCS-L-UNSUB@midrange.com.
| Questions should be directed to the list owner: dasmussen@aol.com
+---