Problem solve Get help with specific problems with your technologies, process and projects.

The IBM iSeries Developer Road Map: Is it right for you?

The iSeries Developer Road Map isn't the first road map iSeries developers have had from IBM, but this one promises to make more sense to more people as they shift to using J2EE.

There's a new road map in town: the iSeries Developer Road Map.

It says, "Moving from traditional iSeries programming languages (RPG, COBOL, C) to J2EE is a long jump, but does not require a single leap. This chart describes a road map-discreet, achievable steps that move developers and applications on the path to an excellently implemented Web future."

We celebrate the fact that WebSphere WASN'T in the FIRST sentence. On the other hand, IBM created another version of the road map for COMMON titled "The iSeries Developer Road map at COMMON." This one starts by saying the following:

"WebSphere is IBM's key infrastructure solution which allows developers to build and deliver dynamic e-business solutions quickly. In order to make the transition from "green screen" RPG/COBOL applications to graphical Web-based applications, IBM is introducing the iSeries Developer Road map."

So maybe the new road map IS about selling us WebSphere -- again. Maybe IBM is worried we haven't bought it already.

Actually, most of the road map is built around the development tools: WebSphere Development Studio Client for iSeries (WDSC). Large companies (20 or more developers) may also look at IBM's Rational tool suite for more application design tools as well (not listed in the road map).

Also, this isn't the first road map iSeries developers have had from IBM. There WERE others:

  • ILE for the last 10 years
  • Also, somewhere in there was SAA.
  • There was RPG to Java.
  • And let's not forget the "original J2EE" map.

Flexiblity -- we're not all going the same way

What's better about this road map? IBM DOES nicely say that not all customers will get on AND off at the same points. In fact, we're not even all going to the same place. A road map with flexibility makes it more viable for a larger segment of customers. So will this one make more sense than the others? I'm betting it does.

Pulling out the road map

Here are the basic steps in the road map:

  • Green screen = where most iSeries developers are today
  • Better tools = WDSC host development
  • Better user interfaces = Web browser
  • Better architecture = MVC (Model-View-Controller)
  • Better portability = Java
  • Better scalability = J2EE

Green screen = most iSeries developers
This isn't anything big. This is what many of us are doing today -- using SEU.

Better tools = WDSC host development
Migration to WDSC host development tools: RSE (Remote Systems Explorer) and CODE (editor and designer).

IBM's value proposition here is greater productivity. But don't expect BIG gains here, and don't underestimate the learning curve.

Better user interfaces = Web browser
Convert existing 5250 interfaces to Web-based interfaces.

IBM's value proposition here is ease of use for users. Also, there's no interactive CPW cost IF you use WebFacing on a NEW 8xx server running Standard Edition software. However, there are limitations for an all-WebFacing solution that you need to be aware of.

Overall this isn't a bad decision, since Web interfaces are more strategic than older client/server interfaces for most uses.

WebFacing is nice because you probably already own the software: WDSC. It's the least expensive option for Web-enabling 5250 applications for users. It DOESN'T handle system screens, so you can use some workarounds to Web-enable CL commands or use HATS (IBM's recommendation).

HATS LE (Limited Edition) is bundled in iSeries Access. If you don't already own it, it is relatively expensive. In most cases, developers and administrators need system screens more than other users. Using secure Telnet over the Web might save real money, but it doesn't give you a Web "look and feel" for programmers. You need to decide if that's important to you.

The full version of HATS never became popular on the iSeries because of WebFacing. Will the LE version packaged with iSeries Access? It will definitely draw a portion of the iSeries base that wants programmer and administrator Web access.

Better architecture = MVC
The concept separates business logic from the user interface, providing a better architecture (MVC or Model-View-Controller architecture).

IBM's value proposition under this is that you have lower maintenance costs, as well as easy support for multiple clients.

IBM has been telling you for 10 years to use ILE to build service programs for your application logic and separate the business logic from the user interface as way to get away from green screen. Then comes WebFacing, which puts you on the Web WITHOUT MVC! Still, IBM is right about the value proposition, although it isn't one easily measured by short-term savings in most cases.

Making an architectural change is usually expensive, especially if done by hand in RPG. There are third-party tools that can help the migration and maybe save money here.

Another point: MVC is 15 years old at least. There are new, more advanced application architectures such as Generic Application Protocol Services Architecture (GAPS) that would provide higher savings on application development. Given the high cost for the change to MVC, it would be nice if the end-point had higher returns.

Better portability = Java
Move to Java for your business logic from RPG.

IBM's value proposition: platform portability for applications AND higher reuse through object-oriented development and better frameworks. The value proposition here is real. Given this step and the prior step are both relatively expensive, maybe doing these two steps in one step is smarter, faster and cheaper for you.

There are third-party tools that can lower this cost. And consultants who've done this before can save time and money on the engineering and development.

Better scalability = J2EE
Here, IBM recommends going to full J2EE: Enterprise Java Beans and Message Driven Beans. IBM's value proposition: high volume, e-business transactions.

OK. You may choose NOT to go here -- especially now. IBM hasn't made this a low-cost solution: You pay a lot more for the runtime and development tools. The real irony is that performance IS an issue, and yet we've been told this is for scalability.

It's true that you can create a scalable solution with full J2EE. It's ALSO relatively expensive. Customers that bought into IBM's story on this in the past paid the price with slow EJB containers, but WebSphere 5 improves performance over prior versions now.

Standard Java (J2SE) can be implemented with many open-source frameworks to get many of the J2EE benefits. The better parts of the architecture can be implemented in J2SE in many cases. Can Java and XML e-business applications scale WITHOUT full J2EE? Absolutely.

If you are a larger company with a good business case, then this might be a good choice. Over time, IBM will improve EJB container services and performance. Will they make if less expensive, too? Not likely.

Where we pull on AND off

This is how we expect to follow the road map in our company.

The phases we will use as is:

  • Better tools = WDSC host development
  • Better user interfaces = Web browser

The phases we will combine into one to save time and money:

  • Better architecture = MVC (Model-View-Controller)
  • Better portability = Java

The phase we MAY NOT select:

  • Better scalability = J2EE

We're able to scale our Web solutions adequately with other frameworks now. We get the same basic services: multi-user transaction integrity for e-business applications. If J2EE we're more affordable with better performance, we might consider a WebSphere implementation of it.

All strategies have exceptions in practice. Ours will have some, too. And we'll need help getting where we want to go from lots of sources: IBM, e-business experts, etc.

One final thought: Beware of IBM recommendations on other tools. iSeries customers have been told by IBM to use WebSphere Application Developer (or Site Developer) to make some of these changes. WRONG! IBM WDSC comes from Toronto. It has a MUCH BETTER set of tools to make these steps happen at a lower cost on the iSeries. Things that take real programming in Application Developer can be done with a simple wizard in five minutes in WDSC.

Where are you going on the road map?

Some customers were lost without one. Others already have the map they need. And some have already arrived where they need to be. Let me know your thoughts on how the road map will work for you. Will it help you or not? Write me at thevaluemgr@yahoo.com.

---------------------------
About the author: The Value Manager is an IBM iSeries IT manager trying to make the right decisions to deliver better value for his company. He welcomes your comments and feedback. E-mail him at valuemgr@yahoo.com.

==================================
MORE INFORMATION
==================================


Dig Deeper on iSeries COBOL programming

Q&A: Mason calls WDSC the tool of choice for Web development beginners

Read feedback to this Q&A.

iSeries shops are placing Web development and Web enablement at the top of their priority lists. But if you're just starting out, where do you begin? There are so many options that choosing the one right may seem like an impossible task.

Jim Mason, president of ebt-now, talked with assistant editor Deb Tart recently at the Northeast IBM User Conference in Framingham, Mass., about tools and techniques for developing for the Web and what his Web tool preferences are.

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

-ADS BY GOOGLE

SearchDataCenter

Close