RE: Saving the System i: Fight Rather Than Switch -- MIDRANGE-L

From: Larry Bolhuis

Time to weigh in here. And just so you know I'm not picking on Joe, his
comment just happened to make my thoughts jell...:


Heehee... my thoughts are jello at this point, not jelled...  But I can
still respond!

Joe Pluta wrote:

So to expect the iSeries to somehow build simplicity AND power AND looks
into a single user interface architecture is asking a bit much.

Joe

    Joe, (first to pick on you :-)  I disagree wholeheartedly!  This is
one of the core tenets of the System i family from it's heritage way
back in S/3x land to the present day! The box IS Simplicity. It IS
Power. It IS Integrated. It can have looks if it's done right.  WE have
the best box on the planet to work with but guess what, so does
Rochester!  To say they can't do it is silly. They built the box, they
can extend it.


I think you miss my point.  Check my much more voluminous previous post: I'm
not saying that we can't have a standard UI, just that extending the 5250
architecture isn't the way to do it.  And even that's something of a
mis-statement; we extend the 5250 architecture every day.  But there's a LOT
more work required than meets the eye, and if it's not done properly, you'll
get junk.  Personally, I think the majority of the "advanced GUI" keywords
on the current DDS need to be axed; they're inflexible and not very useful
in anything but the simplest GUI.

At the other end of the spectrum, the pixel-painting approach of Windows or
X-Windows is WAY too calculation intensive for the iSeries.  It's a complete
waste of resources.

What we need is a very high-level GUI toolkit (something akin to HTML. Such
as XUL) that we adopt.  Even better, we need an iSeries-specific abstraction
layer which we can program our HLLs to and into which we can plug formatters
of any kind, be it HTML or some other interface.

    My company does mostly System i work. Most of our revenue is from
developers working on the platform. Most of that work and virtually 100%
of the new work is developing applications for the web. Applications
that run completely on the System i except for the presentation layer
which is in the browser.  Are our web tools awesome at this point? I
don't think so but they DO work! And since they are built on the best
business platform out there they work over and over and over.


I do the same thing, Larry.  That's what I do (although it's sometimes an
argument as to whether the web application server itself will run on the
iSeries (WebSphere) or on a sidecar (e.g., Tomcat on Windows or Linux)).  I
use a message based architecture where the only thing done in Java is
conversion between EBCDIC and the screen, and all the processing is done in
RPG.  It screams (show me another web-based application where you can hold
down the PageDown key and scroll at multiple pages per second!).

    So why is 'i' sliding backwards? A big part of that is US. Trevor
has posted at least twice in this thread :-) and made that point very
clearly. I completely agree but I think modified tactics needs to be
employed in our data centers.  When the CIO, IT Manager, Lead Developer,
or whoever says: "We need to swtich to Wintel" go on the Offense for OUR
platform. Make them probe that their solution is better than ours. Make
them show the true cost of ownership of their choice. When they say
things like "I didn't even know we had a System i (iSeries (AS/400
(S/38))) ask them what brand of air compressors are out in the shop. Bet
they dont' know unless the thing has recently blown up costing big
downtime. Ask them what brand of most things they have out there and
I'll bet the only ones they know are the brand new ones or mostly the
ones causing trouble. MAKE THAT POINT!


This is where we start to diverge.  I don't think Trevor's site really
espouses the same concepts you do below.  I think Angus is a viral
advertising site, attempting to win mindshare through gimmickry -- this is a
very common technique, I grant you, but I'm not sure it applies in the
server world.  I respect Trevor's efforts, but I'm not yet a convert to the
concept of viral advertising for the iSeries.  I think it's because I get
the idea that viral advertising is an implicit concession to the idea that
FACTS can't sell the box.

ASK QUESTIONS:

1) "So what anti-virus solution will you be employing and how much will
that cost in dollars and system performance? You know we don't require
any of that on System i. That's a cost you are adding to the system.

(snipping many other great questions)


These are excellent points, and while anybody with three functioning brain
cells can see that an iSeries far outperforms a Wintel server farm in all
but the most trivial front-loaded price comparisons, this still doesn't sell
the box.  Why?  Because the people making IT infrastructure decisions - be
it CFOs, CIOs, even DP managers -- are no longer technically competent to
make those decisions.

Back in the day at SSA, I remember that one of my primary jobs during my
stint in pre-sales (all programmers had to serve in pre-sales prior to
becoming managers, an outstanding concept) was to do line-by-line code
reviews with prospects.  Find me a CIO today who can do even a functional
technical analysis, much less a code review.  Hell, a code review of a lot
of the "production" code being written today would cause a good programmer
to run out the door screaming.  Thus the Betamax/VHS trap we find ourselves
in.

But maybe, just maybe, the concept is a little of both.  The best
combination of gimmickry and fact is the current Mac advertising.  Whenever
I see one of those ads, even the ones that miss, I can't help but say
"Brilliant!".  You might note that I'm using the catchphrase from another
commercial.  And here's the difference: while I can't remember which beer
uses that campaign, I am in the process of purchasing a Mac.  Or at least
considering one.  Okay, I'm not a beer drinker, but still, lots of viral
advertising only gets you to remember the phrase, not the product.  While a
good campaign based on both fact and fun might do better.

So maybe that's what we need.  A serious list of reasons to buy an iSeries,
and then a bunch of real viral advertising based on those premises.
Unfortunately, you also need dissemination, and that's going to be the
tougher bit.  And now I'm done for the day... maybe <grin>.

Joe