RE: IBM not investing in i5/OS was: iSociety

From: Aaron Bartell

This is a pretty poor argument, Aaron. An app server is no more work

than Apache.

I am not talking about Apache, I am talking about the 5250 application
stack - I should have clarified. How does that change your comment back
to me?

Um... then I have no idea what you are talking about. Hoenstly, Aaron, I
have no idea why you hate application servers. What is your alternative?

That's part of the beauty of the Java stack; all of the enterprise level

functions are standardized, unlike just about any other stack.

There are very few, if any, Java stacks that I have seen that give you an
enterprise environment out of the gates.

I don't understand this, either. Are you saying you actually have to do
some configuration or tweaking to get the stuff to work? Then I guess
you're REALLY against SQL, since unlike J2EE, which is pretty much one-time
setup work, SQL is constant tweaking and configuration.

The nice thing about J2EE is that you can find support on the Internet for
just about any conceivable configuration, because somebody has already done
it.

Does Java=Bad? No. Java equals changing your IT infrastructure which
*can*
be bad. When I say "bad" I simply mean a whole lot of time was spent
changing infrasturcture without as much payback as what should have been
(talking in the long run).

I'm just wondering what all this has to do with the price of tea in China?
The stuff is there, ready to use, and it's obviously IBM's strategic
direction. You may not like it, but as I've said in the past, a lot of
people didn't like subfiles either.

It was one that had a place to enter a customer number, and then you could
click on an order, and then you could click on a line item - all on the
same
page. I just read it last week and am nearly 100% sure it was your
article
(I remember that grinning photo of you staring back at me ;-). Does that
ring any bells? Maybe it was for a different trade rag.

Uh, that article was on Ajax. It was actually a very cool setup that used
about 6 lines of JavaScript and 120 lines of EGL for a three-level
Ajax-enabled drill-down. That one actually didn't use Java *OR* RPG. But
there was no plumbing; I had to add six lines of JavaScript code to update a
hidden field depending on what I clicked on. If you consider that too much
plumbing, then we have some issues about what the term "programmer" means.

You're talking about objective??? When your primary argument has been

"it
has Java in it, so it must be bad!" Well, pleased to meet you Mr. Pot,
I'm
Mr. Kettle! <grin>

You are putting words in my mouth. My argument is this: IBM *should* be
providing us solutions with the same simplicity as the 5250 stack, but
they are instead listening to vendors clamoring for platform independence.

And my point is that they have. EGL is as simple as the 5250 stack to
implement. You can completely test the entire application on your desktop,
and then you simply export the application and deploy it to the app server.
I'd put that up against whatever it is you have to do to have a test version
of RPG-CGI or PHP and then deploy that to a production environment.

Here's the 64,000 dollar question: was I right about WDSC or not?

You won't find me fighting against this point. I love WDSC. I can easily
deal with it's short comings (i.e. crashes, bugs, etc) because it isn't
used
at run-time and I can always go back to SEU/PDM in a pinch. To sum it up,
development tools live in a separate space than a software stack - though
the lines are blurring more and more on that point.

Bugs? Which ones, Aaron? Crashes? When is the last time it crashed?
Seriously, I'm wondering.

But anyway, this horse is beaten. You won't look at the language on its
merits; because it uses an application server, you are dead set against it.
Me, I hate code generators, but if they allow me to build a pie graph based
on data from DB2 or from a called RPG program in minutes, then I have to
look at it seriously.

Joe