Lots of good comments Lukas. This is the statement that caught my eye and
has been bugging me for some time:
"How many companies are switching TO the System i?"
I would be VERY surprised if a new customer would pick RPG on the System i
for new development, so I won't consider that an option. I do see value in
having all your processes/servers running on a single machine using multiple
LPARs. Would that be enough pull to get a shop who develops in PHP to buy a
System i? Think of all the knowledge learning they would have to do to
maintain it? It isn't huge, but it is a new knowledge set that will have to
live through the half-life of information and be kept up on as new things
come down the pipe.
I know a "Director of New System i Sales" position exists at IBM because I
met him in an executive briefing at COMMON. I regret I neglected to ask
where new sales are coming from (read, new customers). Anybody know how
many new customers the System i gets a month/quarter? I am not talking
about shops upgrading or previous customers that want to venture into new
business and choose System i.
I would think pre-packaged System i machines with things like
SugarCRM/Alfresco/Joomla/VOIP/etc would be attractive for a small business
owner if they could get it all for 15K. To have a one stop server that runs
90% of my business would be appealing.
Here's a site I stumbled across today that should be good for mining idea's
for software to run on the iSeries (i.e. PHP and JSF/JSP/Java):
http://www.osalt.com/
Aaron Bartell
http://mowyourlawn.com
-----Original Message-----
From: midrange-nontech-bounces@xxxxxxxxxxxx
[mailto:midrange-nontech-bounces@xxxxxxxxxxxx] On Behalf Of Lukas Beeler
Sent: Monday, January 14, 2008 1:39 PM
To: Non-Technical Discussion about the AS400 / iSeries
Subject: Re: i5 Youngsters
On 1/14/08, Vance Stanley <w_vance_stanley@xxxxxxxxx> wrote:
I have a manager/supervisor that is concerned about the "greying" of his
work force and the
ability to acquire younger programmers that are willing to work in the i5
environment. I
Well, i'm not a programmer, but i can assure you that there is very
few interest in the System i5 from young people (like me, 23). This is
mostly the fault of the System i itself, because the System didn't
really age that well.
argued that when I came into programming in 1991 I just wanted to work and
did not care what kind
of programming I was doing. I ended up doing COBOL programming in a
mainframe environment and at
That was almost 20 years ago. The IT industry has changed radically in
that time. IT has become standard infrastructure for almost everyone
in the first world.
the time I thought it was good experience. Anyway, my supervisor seems to
think the new crop of
programmers are not as flexible in that regard and only want to learn
whats hot. Any thoughts
Well, there certainly are people that only like to learn technology
that is hot and current.
But the problem lies elsewhere. There is no long term future for the
System i as it is right now. How many new LOB vendors do you see
releasing new software on the market, that is designed for the System
i?
How many companies are switching TO the System i? There are a lot of
companies staying on the System i, mostly because they have invested a
lot of development time into it in the past, but none of them are new
to the platform.
Few are going towards it, though. Not enough to be convincing.
I've been on Midrange-l for 2.5 years now. Most of the people i've
read here are well beyond their fourties. They only have to worry
about the next twenty years or so, in which many companies will still
be running their legacy applications on the System i.
For younger people, learning a platform that is already way past it's
prime time is wasted time. There is not much to be gained from a
platform that will probably be withdrawn from the market in the next
ten years.
Windows has an absolute dominance on the corporate desktop market, and
the server market is divided between Windows and the Unixoid operating
systems (Linux, Solaris, xBSD, and a lot of legacy OSs).
Linux/Solaris/*BSD can be made to interoperate rather well with a
Windows based infrastructure. The System i on the other hand, does
not. There are awkward limitations that have nothing lost in the 21th
century, like 10 char username limit, and the general limits on object
names etc.
Not the mention that the most used native programming language (RPG)
is based on punch cards, and looks like it escaped some bad 80ies
movie.
But in the end, these are reasons that appeal to the technical mind.
Businesses usually make decisions based around applications, business
needs, etc. Not technical quality.
And this is where the System i is really dying. There is no influx of
new applications. Sure, there are many companies that keep on
releasing new toolsets to work on the System i, but they're usually
for very basic functionality like CGI, generating XML, sending mails,
creating PDFs, etc. All this functionality is included with other
platforms today.
But these are not the applications a business needs. A business
usually needs a variety of LOB softwares. There are no new such
softwares being launched on the System i. On Windows on the other
hand, you'll have it hard to keep counting the new releases in a
single month.
The days of 5250 data entry are long gone. We're living in modern
times now. Most of this stuff is automated using EDI, XML,
WebServices, you name it. You'll still need to consolidate all this
information and present it to the manager, but nobody is sitting at
his desk day-in-day-out typing in invoices.
Of course, you can run modern software on the System i. But come on.
We've got a Java components in our software, but it runs faster on one
of our development machines (2.4 Ghz Xeon Quadcore) than on our 3800
CPW Model 515. But the latter machine cost four times the money.
There are other reasons why there are few new applications on the
System i, besides the dwindling user base, high price and the awful
performance: the cost of entry.
Suppose you're a little startup, maybe 5 people. You found a nice
niche market for an LOB software in your region. You want to start
developing:
Linux/Solaris/*BSD
You organize a few halfway decent machines for development. Say, 10k in
total.
Software is free, but maybe you need to add a big database. You can
start out with MySQL, Postgres, or Oracle Express. You download
Eclipse to begin your work developing whatever you want.
Total: 10k
Windows
You organize a few halfway decent machines for development. Say, 10k in
total.
You purchase an action pack subscribtion from Microsoft, costs 500$ a
year. You can start out with SQL Server 05 Developer edition. You get
5 free CALs for Visual Studio.
Total 10k5, plus 500 US$ recurring per year
System i
You organize a few halfway decent machines for working on. Say, 5k
total. You ask IBM if they give you a developer discount, they'll tell
you that you need a product in order to get one. You purchase a model
515 at retail price, say 15k. You only have a single development
environment, making testing a nightmare. You want WDSC, you'll pay
extra for all kinds of stuff. Oh, and you'll also need SWMA in order
to stay current.
Total: 20k, plus 1500 US$ recurring per year. Plus WDSC extras you might
need.
about how I can present a case to show that the system should be well
supplied with professionals
for at least the foreseeable future?
Well, i'll expect that the current generation of the old people
working on the platform will be enough till either the platform dies
off completely or IBM fixes it sufficiently to make it more
interesting for the current generation.
Sorry for the long rant. I really had to vent off some steam.
midrange.com
