Stuart,
you may think that your approach is something you came up with,
it isn't - what you do has been done since the database management
departmens existed and was one key part of an 370 mainframe
installation in the seventies.
They had a data dictionary, a central repository with metadata
and that is like having a "gold mine" because you can "pack and
repack" any information stored into that to any given UI.
In the seventies they had it on paper, some ugly thing we today calls
documentation, with the S/38 or was it when the AS/400 came it was
known as a Field Reference File, but it never did cover the UI (DDS
DSPF/PRTF) on the AS/400.
There is a historical reason for that, but this I will jump over.
In my product it is central and speeds up development 10X+ and
it can of couse be used to any UI as weel as DB definitions.
Does ZEND has it, no, does RAILS have it, no, does MS have it, no
and in general it goes out of "the box" and the discurse in modern
web development where people tends to think that pointing a file and
getting something on the screnn is equal to efficiency.
If PHP was such a miracle, why don't they just push a buttom in a
10 year old project and out of the push comes a top modern RIA based
application ?
The reason is that they have the same F**** problem that an iSeries
with a good old System/36 Environment application - is that they have
to rewrite not only the UI logic, but also their metadata statements
and their bussiness logic - they are as stocked as anybody else.
In fact this is simple and hasn't to become a rocket science, just
follow the roules of Keep it Simple, Stupid (K.I.S.S) and SKUNK WORKS
by "Kelly" Johnson at Lockheed Martin, and you do not need big companies
big analizys programs or directions to improve efficiensy på 10 times.
The only thing you need is a little leather noteblock where the in the
front is printet "THINK"
;-)
Fell free to visit my homepage and the demo and power of a central
repository at
http://powerext.com
/henrik
"Stuart Milligan" <stuartm@xxxxxxxxxxxxxxx>
Sent by: web400-bounces@xxxxxxxxxxxx
12-03-2010 20:02
Please respond to
Web Enabling the AS400 / iSeries <web400@xxxxxxxxxxxx>
To
<web400@xxxxxxxxxxxx>
cc
Subject
Re: [WEB400] Flex & RPG
Aaron,
" As you embrace Flex I would make sure to build for change in the UI
layer."
That's the whole purpose of what we do. We reengineer applications from
monolithic into MVC so the reengineered model part of the code can be used
with ANY UI. But out of the box people want a usable and maintainable UI
as
part of a rebuild process. We abandoned the Genuitec idea as they didn't
really understand enterprise style application dev requirements or weren't
interested in that market.
" so far HTML+Javascript+CSS has been the only safe technology"
We have been doing that for 10 years now but it's not as productive as
Flex
to develop for the end user. Strictly speaking we don't care what UI
technology our customers use, but we have to generate something from the
reengineering that will be mainstream. For most system I customers this
means providing them strategic advice and technology options that marry
new
world and legacy together. So we have become extreme experts in both
worlds
for enterprise apps. We have no responsibility for the ongoing development
of apps we reengineer, but we need to give people a running start. So that
naturally includes generating the right UI types and technologies.
"What's interesting to me is that Databorough has seen that those other UI
layers weren't a good fit for modernizing IBM i legacy code,"
No. They are very good. Those that use them love them. It's just that they
are proprietary to us, and as such they aren't mainstream. We are
reengineering experts not forward engineering. We don't have the
management
bandwidth to do them commercial justice. We aren't on the java bandwagon,
the rest of the world is Aaron. We are just getting people onto the wagon.
We also do C# and EGL so you could argue that we aren't on any wagon or
all!
Flex is just another option we are going to offer from our regeneration
options, and I was interested to hear of any IBM i specific experiences.
" I am still waiting for
Databorough, or other vendor, to come through with a solid way to talk to
modern interfaces using RPG+DB2+IBMi."
That can be done right now with any of the mainstream technologies. What
you
are hoping for is a UI technology that is driven by an RPG program on the
back end. Although we would love it to succeed, be we feel that it will
never be mainstream. Procedural programming is over. The world has moved
on
over the last 30 years. It's not to say that enthusiastic and competent
entrepreneurs/technologist like yourself and Nathan of course, won't
provide
some specialized solutions using this approach, and probably make some
money
doing so. As you know we have the resources and the know how to do exactly
what you are asking, and even the need for it to succeed. What you are
saying is the reason the rest of the world hasn't sat up and realized that
RPG centric development is the right way, is because there isn't a modern
UI
technology that delivers it. The only thing that is missing is this last
piece of the puzzle. Think about the players that have an interest in this
being true and yet haven't done anything: IBM, Microsoft, Oracle, Seagul,
Infor, Databorough, LANSA, Blue Phoenix. Surely we can't all be wrong?
Now don't think I am knocking RPG_Gui or any of your efforts. My intention
is to respond honestly to your questions. On top of that I am saying that
you can have your pick of ten years of development technology for free, as
I
know there is plenty of code in there that you can use in your RPG-Gui
project. Although people don't yet realize this, we have spent many
millions
of dollars and ten years in developing complete system reengineering
technologies for rebuilding enterprise apps into modern languages. You
don't
do that and not learn a thing or two about modern technology, technically,
commercially and practically. We also happen to have X-Analysis which is a
vehicle to get people into strategic modernisation while giving them value
up front, but most of our reinvestment over the last decade has gone into
our modernisation products we release this year. The response to these has
been a little overwhelming and we have had to bring the release dates
forward. Love to tell you more about those if you are interested some
time.
So if you want to open source our code go ahead. We have used it as R&D
proto-typing for a broader more optimum result, but there is some good
stuff
in it.
I certainly hadn't expected to get drawn into this level of debate when I
asked about flex!
Cheers
Stuart
midrange.com
