Re: what was the single mgmt decision made in mid-1990s. period

Nathan,

Since I have a long standing relationship with the company, and they are comfortable that I can do the job very well, there is not much of a leap of faith required.

There are numerous areas in the Windows package that would require modification. One is even a major change to the key hierarchy.

This is a relatively small, but growing, company, with maybe 15-20 users. Adding a couple of servers would probably cost around $30K+ in hardware, installation, and basic software licences. So even though it's a commodity, it still costs money, which could have been applied to revamping their software. Note that they don't have any in-house techies, so their network support costs will probably go up too.

When you say "yank an entire database," their understanding when considering this route includes doing a data conversion of the most important elements of their business and forego the rest. My proposal is pretty much a rewrite of their existing system, since much of it dates back 15-20 years and is showing its age! So both systems would be "new." The existing database design is good to use as a starting point, but must be updated.

In your estimation, how much longer does it take to code an application doing it with a GUI vs. green screen? Or allowing both, depending on what type the user prefers?

After 10 years of RPG programmers continuously calling for a native GUI I'm sure they got the message, but they still seem to be backing J2EE tools and interfaces, and more recently - PHP, instead.

IBM may have *heard*, but they have not *gotten* the message!

IBM's services revenues (at least as far as application services are concerned) in the midrange arena have traditionally been relatively low. I think that it will still be a losing proposition for them compared to selling the hardware and generic pieces (OS, compilers, system utilities.) Their entire VAR / Business Partner model, which I recall going way back, is based on the realization that IBM is NOT good at the application stuff, so why not team up with the professionals that ARE good at it?

To summarize my original point: IBM is coming to the GUI party very late (or not at all!) and / or charging extra in areas where the feature should be included. To illustrate: Would your prospective clients today be willing to pay extra for TCP/IP support or would they laugh you out of the room?? In the case of TCP/IP support, at least IBM realized that in order to remain in the game they needed to include it as a no-charge feature and become fully integrated. They have not come to the same realization with regards to a native GUI, and that, IMHO, shows that they are out of touch with what their broad base of clients and prospects want and need to sell the i.

I see plenty of progress for the big iron boys (LPARS, Capacity on Demand, external disk arrays, blades, multiple CPUs, etc.), but a lot less attention to the basics. This is very frustrating for the little guys in the trenches.


-mark

At 2/17/09 09:39 AM, you wrote:

> From: M. Lazarus
> I am competing against a Windows based package.

Since the package already exists, you're comparing something concrete to something that must be developed, and that takes some faith. That must be a bit unsettling.

> A long term relationship with the company
> We know their business

So hopefully you know things that are missing from the package that they need, something that adds distinct value.

> They already own their iSeries hardware and software

Hardware has become somewhat a commodity. But what about their database? Does it serve them well? If so, that could work to your advantage. It can be pretty disruptive to yank an entire database and replace it with a package. Contrast that with the option of incremental application replacement, while retaining their database.

> They recognize the stability of the iSeries and i5/OS

That's good.

> developing a brand new system that looks "old" is a major
> strike coming right out of the gate.

Maybe I can help there. I'm good at cranking out new RPG / HTML / JavaScript based applications, if you'd be interested in collaborating with systems analysis, business rules, and whatever else they may need.

Otherwise you're probably aware of Webfacing, HATS, CGIDEV2, EGL, Websmart, Valence, Lansa, and other promising tools available.

> The decision makers at IBM just don't seem to get it.

After 10 years of RPG programmers continuously calling for a native GUI I'm sure they got the message, but they still seem to be backing J2EE tools and interfaces, and more recently - PHP, instead.

And IBM has transformed into an application development/services company themselves. I'm sure their priorities are well defined and their strategy is well thought out. They seem to be okay with declining hardware / OS revenue, as long as application & service revenue goes up.