i think you hit the nail on the head. over a year and a half ago, this
company *bought* a package for shipping/receiving using hand-held scanner
(running Win CE) that of course sounded like the greatest thing since
sliced bread (IBM needs to hire their salesman!) after they purchase the
said software they realized that the product was a) not written yet
(vaporware) b) they wanted *me* to write the bulk of the data manipulation
on the "i" for the backend (so we were paying them to do it ourselves
basically. c) their crapware couldn't do even the simplest validation
(ensure that the ID number which is only 10 bytes long was scanned instead
of the VIN number which is 20 bytes long!) it was supposed to be in
production over a year ago and here we are just now running in parallel
(and still fixing their bugs!) needless to say the project manager(Six
Sigma guy) will now ensure that IT is involved in any future discussions,
etc. since my manager and I explained that by now *we* could have written
this thing using a web front-end that performed data validation at
ship/receive time, etc. and it would have been completed and in production
in 3 +- months (as simple as the project would be more like a month) but
i digress...well over $100,000 of *our* man hours (user testing, field
testing and that doesn't include all the coding time since i had to work
with their inadequate system, data transfer issues, etc) so it was a
*very* expensive lesson but i think they've learned that lesson and not
likely to forget it!)
this company has used IBM midrange since the beginning of the S/36 era and
love the platform and won't leave it...but management as it was replaced
over the years forgot how flexible, etc this thing is...and now they have
seen that i can use CGI-RPG and PHP for the "pretty oooh ahhh" screens and
do it all on the system no PC server required (or desired) it was simply
yet another case of (we didn't know you could do that on the as/400 (yes
they still call it that). but they haven't seen anything yet 8^)
Thanks,
Tommy Holden
From:
Joe Pluta <joepluta@xxxxxxxxxxxxxxxxx>
To:
Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Date:
11/23/2008 09:58 PM
Subject:
Re: Casino switches to AIX
Sent by:
midrange-l-bounces@xxxxxxxxxxxx
Adam West wrote:
Another post was about the Windows envirnonment not being good for a
business ERP. I would like to see more details on that. What we are being
told that everything is just peachy in windows. I would need evidence to
the contrary to help save the i installation.
I've been avoiding this discussion because it is very much an opinion
discussion, and in many ways I think it belongs in the midrange-nontech
list; we're talking about things like look and feel and application
features, neither of which really has anything to do with technology.
Your question itself underscores the point: you're told that everything
is "just peachy", which is a term rarely used to describe the technical
capabilities of a system.
That being said, I thought I'd weigh in here with my observation. The
truth is that, for a certain segment of the business community, a PC is
a better choice than an i. <GASP!> But if you think about it, that's
got to be the case - just as the converse is true: for a certain segment
the i is a better choice. The goal, then, is not to try to identify
using abstract measurements which system is "better" than the other, but
instead to find out which one is better for your particular business.
And to me, the biggest question is a fundamental one: do you need to
modify your software? This is a question that is almost completely
overlooked nowadays just as it was in the 80s and early 90s, although
for the exact opposite reason. These days, with the exception of open
source software, almost no packaged software comes with source code.
Either it has the features you want or you live with it; nobody would
even *think* of trying to modify Excel to act ... well, correctly
<grin>. On the other hand, when I was but a brash young "ute" (cf.
"youth", see "My Cousing Vinnie") in the industry, nobody would think of
buying ERP-level software that *didn't* come with source. That's
because almost immediately you were busy modifying things like invoices
and order acknowledgments, and changing how pricing worked. Ah, times
have changed.
Anyway, if you don't need to modify your software, it's almost a given
that you can find a package for Windows that will be "cheaper" than the
same package on the i (for the purposes of this particular post, I'll
forego the Total Cost of Ownership argument and just stipulate to the
initial cost benefit of Windows software). However, that means that you
have to run your business within the strictures of that software. It's
impossible to estimate how many companies have modified their processes
to live within the constraints of their software, but there are a lot of
them. And that's completely bass-ackwards to what I was brought up
with: in my day, software was not only a tool, it was a differentiator.
Having the best software meant you had better inventory control; or more
flexible pricing schemes or more accurate forecasting. You lose that
with packaged software: if you're running your business using the same
software as your competitor, then what exactly is the competitive
advantage of your software?
So there, then, is the fundamental issue. Do you build your business
around your software, or do you build your software to help your business?
If it's the former, buy Peachtree accounting or Microsoft's business
packages. But if you want software that gives you a unique competitive
advantage over your competition, then you need semi-custom software that
captures and builds upon your business rules. And for that, there's
simply no better language than RPG. (I've got a standing bet with
anyone in the entire computer programming industry: sit down with me and
a couple of real users and put together specifications for a custom MRP
generation system. We write the code in the language(s) of our choice.
I'll use RPG on the i, and I'll be done first. Nobody has ever taken me
up on it.)
We've lost sight of that. There are no semi-custom applications. Open
source may come close, I suppose, but the only attempts I've ever seen
at that are in Java, and despite the inevitable hate mail I get for
saying this, it's a horrible language for writing business logic. The
other way companies attempt it is by having lots and lots of switches, a
la the SAP model, but then you have to learn all their switches and
inevitably you run up against a situation where they don't have a switch
you need.
Anyway, this might all be moot. It's possible that businesses have
forgotten about the concept of using IT as a competitive advantage. The
closest they come now is having the flashiest web framework, but
honestly the core of most businesses is not their web site. As much as
I love EGL and the ease in which I can build great looking web stuff, I
still think that the core business logic is what differentiates a
company, and the only way to build that is with RPG.
So -- long, long response which doesn't really have an answer. But if
your company can be reminded that IT is not only a way to do things
faster, but also a way to do things that your competitors *can't*, then
I think that the i still has a chance.
Joe
midrange.com
