Re: RPG IV -- MIDRANGE-L




-----Original Message-----
From:   Jon.Paris@halinfo.it [SMTP:Jon.Paris@halinfo.it]
Sent:   Tuesday, November 24, 1998 8:41 AM
To:     Don
Cc:     midrange-l@midrange.com
Subject:        Re: *** ADMIN: Ok, one more time, a POLL on 
www.midrange.com

>Reasons not using RPG IV include:

>No business reason to do the conversion.

I'm constantly hearing about the RPG programmer shortage.  I would have
thought that a language that has a minimal to non-existent learning curve
but brings significant productivity increases is reason enough.  I would
say it's hard to find a business reason _not_ to do the conversion.  Also
in this category, you're locking yourself out of almost everything new on
the system - IBM are not going to go on building CALL type APIs.

[Buck- Jon, I wish that a promise of future gains in productivity
[ would serve as a business case, but one can't put a dollar
[ amount on the benefit, while one can easily add up the cost
[ of doing the conversion.

 >Why the extra learning curve to do what I have now and what works now?

What learning curve ?  Absolute outside it is 4 to 5 days to use the
compiler in compatibility mode.  Note that I'm not advocating that everyone
switch to full ILE etc. - but every journey starts with a single step.

[Buck- I was able to CVTRPGSRC and add a new prototyped procedure in
[ one day.  Made it work 100% by noon the next day.  Lower case and
[ long field names take zero time to learn.

 >And, you'll see that alot of this also applies to your JAVA question...

Here I really have to strongly disagree with you.  We're talking about a
situation where there is no conversion available (and you probably wouldn't
want it anyway since we're talking about OO versus procedural) compared
with a simple command that does it for you.  We're talking about the
difference in learning curve of 5 days (I'm being generous) versus 3 to 6
months.  I fully understand that a switch to Java does indeed need a real
good reason. Also other RPG programmers will undoubtedly be able to read
your RPG IV even if they haven't been "converted".  They wouldn't be able
to read your Java.  Sorry but this is a "chalk and cheese" comparison.

[Buck- I will NOT work on another program that I did a CVTRPGPGM
[ on.  The juxtaposition between the old indicator-ridden ENDIF
[ ENDIF ENDIF ENDIF ENDDO ENDIF stuff and EVAL rc=myfunc(DataParm1:
[ DataParm2) is just too hideous.  There's WAY TOO MUCH RPG III
[ code that is just not going to benefit from CVTRPGPGM into
[ RPG IV. 
[To gain meaningful benefits from RPG IV (imvho) you must at
[ least use procedures and pointers; neither of which are
[ well-known to the RPG III community.  Neither of these are
[ going to give you a quantum productivity gain.  You need
[ ILE to do that. 
[When a shop goes ILE, there'd better be a new mindset that
[ goes ILE along with the code.  I humbly submit that the
[ thousands of RPG III programmers who moved up from the
[ S/36 are not going to deal well with that new mindset
[ in 5 days.

Buck Calabro
CommSoft, Albany, NY
mailto:mcalabro@commsoft.net



+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---