RE: Free OS/400 -- MIDRANGE-L

I think that tools will evolve (like Dreamweaver, etc.) that will make
knowledge of the underlying languages nice to know, but the tool will do
the work much better, faster and accurately than we can do today with
current code generators.
-------------

I disagree with this.  In fact, it's so wrong it's scary.  There still isn't
a code generator today that can write a decent application; if there was,
the whole world would be using it.  And why is that?  Because code
generators don't understand the concept of an application.  Fred Kulack has
a great quote that he's been using for some time as his tagline:

"The stuff we call "software" is not like anything that human society
  is used to thinking about. Software is something like a machine, and
  something like mathematics, and something like language, and
  something like thought, and art, and information...
  but software is not in fact any of those other things."
Bruce Sterling - The Hacker Crackdown

Software is still art, folks.  I don't care what anyone says - when it comes
down to the making of a great piece of software, it still requires that
nutty, kooky kind of inspired madness that each of us has tasted at one
point or another in our careers.

Sure, I can probably design a DFU generator.  It might even have a few cool
rules.  And then the end user can put together tables to their hearts
contents and generate all kinds of pretty graphs.  But when it comes down to
designing a tiered pricing structure that requires rebates based on total
volume sales not counting discounted items over the previous three quarters
with a sliding scale based on the market penetration of the client, but with
exceptions for direct sales sites who participate in corporate funded
marketing efforts, well, I think I'm gonna have to call in a programmer.

The art of making a business run with a computer is not about getting the
end user to get the most out of a limited toolset, but instead it's about
extending the toolset to let the user be as productive as possible.  A
limited feature set is fine in a word processor, but not in an order entry
system.  The ability to tweak payroll bonuses based on production statistics
is crucial to companies trying to figure out how to properly apply
incentives to a demographically evolving workforce.  Work that crap into an
Excel spreadsheet sometime.

Everybody keeps comparing OS/400 and Windows, or even Unix.  This works fine
when you're talking about software as commodity, but the analogy falls apart
when you're talking about software as custom designed tool.  Where this
discussion misses the point is where it is so focused on the price that
we're forgetting that the customer will spend more for a top-of-the-line
hammer, as long as they perceive it to truly be the top of the line.

RPG and DB2/400 are still the top of the line for custom built business
operations.  If you think of each site as a lovingly restored and maintained
classic car, then the software is the big red rolling toolbox that is used
to keep that car maintained.  Now, if the customer base decides that they
need to go to fuel injection, then carburetor maintenance is going to be a
declining market.  But that doesn't mean that the entire marketplace is
going to go out and buy Hyundais or Kias.  UNLESS we refuse to learn how to
maintain an electronic ignition.

So what we need to do is figure out how these newfangled fuel injectors
work.  But we don't have to all become Hyundai salesmen.  Instead, we need
to learn how to make those fuel injectors really hum with that balanced and
blueprinted 455 Wildcat engine we have under the hood.

Enough of the metaphor.  The point is that there is still a place for custom
software, and we are the ones that keep that place safe.  If we all throw up
our hands and decide that some generated SQL code with a VB front end is
good enough, then we're going to see the end of custom software shops,
because frankly, folks, the new generation of Nintendo technicians can throw
those together quicker and better than we can, and in short order companies
like Oracle are going to have "application servers" that you can rent over
the web that will be cheaper than anything a custom shop can provide (Oracle
just announced this, BTW).  And if we drop ourselves to that lowest level of
standards, then the customer will by default go to the lowest bidder.  But
if we as a group design an architecture, or a set of architectures, that
combines the benefits of a GUI with the flexibility of an n-tier structure
and the power and maintainability of an RPG/DB2 back end, then we can still
provide a reasonable, cost-effective service to the community.

But sticking a screen scraper on an outdated order entry system is not the
solution, nor is rewriting your legacy applications into yet another
proprietary user interface.  What's required is a lot of lateral thinking, a
combination of a tactical quick fix and a strategic long term direction,
that provides both a growth path into the new technologies while still
taking as much advantage of the existing legacy programs and legacy
programmers as possible.

Or you can read up on Dreamweaver and polish your JavaScript skills, and ask
Larry Ellison for a job.

Joe

+---
| 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
+---