Re: No giving up on System i (was: I'm about to give up)

Pete,

I am not trying to absolve IBM. My issue is with those people who cry "the
AS/400 is dying because it has no native GUI". These people have not
experienced your GUI pain. I appreciate your experience, having seen a lot
of that first hand also. I don't have the smarts that you do to take System
i further to the web with Ruby on Rails, and I certainly don't have your
Java experience.

However, I am not convinced that anything IBM can do will solve the System i
GUI "problem", though. I still think it is in the hands of us - the
industry.

Look at HA, for example. IBM started with a product in that area, but
dropped it. Did the System i HA industry die? No - it flourished, producing
at least 4 major vendors and a slurry of others. Sure, things are now
different, but no one railed on IBM for their lack of HA tools.

How about change management? IBM had no tools there, and it produced at
least 3 major vendors, and a slurry of others. That industry is vibrant, and
the newest player is making huge waves!

Looking at GUI, there has been many players. Vendors come and vendors go.
All flavors of GUI have been thrown against the wall, and IBM played (late)
with webfacing. And, CGIDEV2 was a way for RPG programmers, with a knowledge
of HTML to get their apps into a browser, and therefore, the web. Sure, web
is not GUI, but people think it is, and that is the ultimate argument being
offered.

I talked recently with someone who was contemplating entering this market.
They asked why there was no obvious market in play. From my perspective, I
said there is no demand for GUI. IBM can also be blamed for that - they let
us run 20-year-old green screen code on today's i5/OS, and don't require us
to keep up with the newest in our own OS. Ultimately, though, it is the
System i IT shops who are not standing up and demanding GUI. They are being
pressured into GUI by users or business, and knowing nothing, spend time
wondering about what to do - but the market is not that visible,
unfortunately.

Having been through this experience for more than a decade, you are one of
the few who have asked for, or even researched, a GUI. Everyone else has
been sitting in the backroom, coding RPG, and not bothering with GUI. All of
a sudden - to them, not to you - GUI hits, and RPG - their ONLY language -
does not have GUI. Finally, the user wants a GUI, and the developers don't
know what to do. From that springs the "AS/400 has no native GUI, so it is
dead" cry.

I still find resistance to change when I work with new customers on GUI
projects. I find RPG developers who have never seen a line of HTML code.
There is little understanding of forms, controls, properties, event-driven,
object-oriented, visual development. Sure, there are more people on this
list who have been exposed to those things, but I said PHP to a System i
customer the other day, and I had to stop and explain what it was. System i
shops are primarily settled into a routine that is hard to change. Blaming
IBM won't change it, either.

Our industry could affect IBM, and could effect change. But they cannot even
agree on the name of the platform. Certainly, IBM can do many things for us,
and something like browser-from-RPG would help a lot of shops where RPG is
their only/major skillset. But, hearing IBM blamed for no native GUI to
cover up for a lack of modernization in System i shops is frustrating. And
the noise we hear on forums like this is even more frustrating. Should I
really use a 4GL tool that generates COBOL to replace all my RPG, because it
works in a browser? While some in IBM may have said this, other IBMers think
that is as outrageous as it sounds.

IBM does a lot that we don't like. But, more people should be like you and
work to resolve the problem, rather than shout "IBM is to blame" for
inadequacies in their own IT department.

Trevor




On 10/7/07 7:15 PM, "Pete Helgren" <Pete@xxxxxxxxxx> wrote:

Trevor,

Normally, I wouldn't take issue with much of anything you would say
about System i GUI adoption, but the common thread that weaves it's way
through the technologies that you mention, except perhaps i Nav, is that
rather than produce a "native" GUI engine early on, something that is
RPG driven and rendered, IBM ignored the problem for quite a while in
the early 90's and then, when IBM started to lose out to other GUI
driven technologies (like Windows and Mac), *then* they got with the
program and started down the GUI path. This was usually accomplished by
"bolt on" technologies rather than native technologies. I DO realize
that there are some hardware and software realities that made a native
GUI rendering engine difficult to accomplish, but rather than address
that in the long run, IBM stuck with the "bolt on"s. Those are things
like fat clients (Visual RPG), and Java. And, what made the transition
worse was that these bolt on technologies were slow and incomplete at
first, so as in my case, I had some really bad initial experiences with
GUI interfaces to the System i. So bad, in fact, that I looked first to
Windows (C++ and VB) and then to the web (Java Servlets). I don't think
I was alone in that experience.

Power 5+ System i's and i5/OS V5R4M0 have remedied most of the initial
issues moot. The System i is a fine server platform, one I prefer over
any of the other server technologies I work with. However, I hold IBM
accountable for those initial missteps and they still seem not to quite
get it when it comes to GUI deployment strategies on the System i. It IS
IBM's fault that we arrived we are are, relative to System i native
GUI's, and it IS IBM fault that GUI development is lagging in the RPG
world that most of us live in. The easiest way to fix that is to
provide free tools, or complete development packages that are VERY
reasonably priced (< $1000) to make that GUI development easy and
economical for the ISV's and developers who love the System i. That
loyalty should have been rewarded. Instead, IBM seems to take the
stance that they will charge for tools until a pain threshold is
reached, then back off to a "tolerable" level.

I love the System i and I am very comfortable wearing my Java shoes in
the System i home in which I live. But, some folks love RPG as well as
the System i and would rather not have to change languages in order to
develop 21st applications (web applications). These are the folks that
IBM has left behind, to the loss of IBM. I am not saying there is a
complete dearth of tools for RPG but can you name one IBM developed and
marketed web development tool that uses only RPG? (needing HTML or
rendering HTML is a given) I can't think of one. And there's the rub.

Yeah, developers who fail to move forward are part of the problem, but
IBM can't be completely absolved in it's role here.

IMHO

Pete


Trevor Perry wrote:

Mark,

Blaming IBM continues to be the game-du-jour. This is not the approach
needed to remove the stigma from our greenish hue.

Regardless of anyone's preference, IBM has provided newer interfaces to the
System i.

1. iSeries Navigator - GUI tool to manage the server.
- people don't use it, because it is not the ~familiar~ green screen. Not
IBM's fault.

2. WDSc - GUI IDE based on (a modern tool) Eclipse
- people don't use it, because it is not the ~familiar~ green screen
PDM+SEU, etc. Not IBM's fault.

3. WAS - Java middleware to deploy web applications.
- people don't use it because it is new and they don't like change. Not
IBM's fault.
- people don't use it because it requires Java. Not IBM's fault - they
provided wizards for you.
- people don't use it because it sucks resources. IBM's fault.

4. HATS, Webfacing, WDHT
- people don't use it because it is new and they don't like change. Not
IBM's fault.
- people don't use it because it is hard to use and requires lots of
maintenance. IBM's fault.
- people create crappy user interfaces with the tools, because they don't
have a graphic designer "design" the interface, thinking they know better.
Not IBM's fault.

5. Third party interfaces/refacing/reengineering.
- vendors don't use them because they won't invest in modernizing their
product, catering for their current "AS/400" clients. Not IBM's fault.
- vendors don't have the resources to invest in modernizing their interface.
Not IBM's fault.
- vendors don't invest the time in researching their choices. Not IBM's
fault.
- customers are fed a lot of crap from vendors about how wonderful their
"modernization tool" is, and remain hesitant about GUIizing. Not IBM's
fault.

.... and the list, and the blame, goes on...

The bottom line is that the ability to create a better human interface is
available today, and has been available for quite some time. If there is no
GUI interface for an application, that is NOT IBM's fault. And the industry
needs to stop whining about it. It is the INDUSTRY's fault that there are
few GUI interfaces to the System i and i5/OS applications.

Trevor