RE: Application design & architecture

Hi Walden,

This is why I hope to implement new standards NOW before we make an effort
to proceed with ILE.  The traditional resistance to this technology can be
softened if we implement standards now that enable SOA for the future.  In
other words, I think we have to lay the infrastructure under the guise of
"standards update", staff development and training, and modular design.  

I agree that the resistors will try to stretch this out for years.  Perhaps
its best to set the standards that point the way, and don't tell them where
the road leads.  Some day, they might just discover the truth for
themselves....  In the mean time, our apps will still benefit from the
modularization efforts.

I disagree that SOA cannot be demonstrated.  Arguably, SOA boils down to the
consumption of "services".  Postal address validation is a good candidate,
since everyone has a working understanding of address validation.  Write an
ILE RPG app to consume (via SOAP) an address validation service.  I think
Scott Klement offered a sample program to consume SOA components.

Eric DeLong
Sally Beauty Company
MIS-Project Manager (BSG)
940-297-2863 or ext. 1863



-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx
[mailto:midrange-l-bounces@xxxxxxxxxxxx]On Behalf Of Walden H. Leverich
Sent: Monday, May 08, 2006 1:06 PM
To: Midrange Systems Technical Discussion
Subject: RE: Application design & architecture


>Moving a wall and adding a toilet CAN be SOA.

Yes, it CAN be. Heck, I'll argue perhaps it SHOULD be. But for may
iSeries shops the reality is they're not allowed to use something simple
like a trigger, or SQL because too many people wouldn't understand it,
and it would take "too long." SOA isn't going to be on the table for
years.

SOA is going to have the same problem CASE tools and OO design had
before it. The payback -- and it's a big one -- doesn't come on the
first and second use, it comes on the 10th and 11th. Many shops are
pressed for enough time to "do it right" using the tools they have now,
to get enough time from management to do something that _will_ add to
the timeline for the first use (even if it saves _huge_ time and money
down the line) is a nearly impossible sell.

SOA also has an uphill battle because it's not visual. I don't mean
there aren't visual tools to help you, I mean SOA is an architecture,
not a presentation (as you well know). This makes it hard to sell to the
"higher-ups." I've seen numerous cases where web applications (and
client/server before it) were frowned upon until some programmer stayed
late and make a crappy, but cute, visual interface to sales inquiry. He
showed it so management and they literally saw the light. SOA has no
such sales mechanism.

-Walden