System iNetwork

I fully agree with Max views regarding the modernization of iSeries applications. Infact we should not be calling it as Modernization rather I would prefer it to call as revitalization, since we are not going to change the system as such but revitalizing with more features to extend the life span of the application system.

The main goal of any application modernization is a marked improvement in all fronts compared to the currently running application and on top of all this should be the major goal, that is to extend the life of the current application.

The main challenge as far as I am concerned is to come up with a modernization strategy for the iSeries legacy systems. Please do note the point that there is no generic modernization plan which will be applicable to all the iseries shops. Modernization requirements for every iSeries shops will be unique and the priorities for modernization is determined by the shops business requirement.

The most concise but insightful comments I have read on this subject from anyone since modernisation became an issue. I look forward eagerly to hearing of your progress as someone in a similar position but with a company about 1% of your size.

a commendable blog, Max. I'm sure we'll all learn from it, and looking forward to everyone's thoughts.

rd

Will this modernization provide support for web applications, reduce support and maintenance costs, reduce duplication (those "cloned" programs and "cloned" tables), update IO to use SQL, update the design of the database, add commitment control, rule engines, workflow, auditing, ILE, activation groups, imaging? You may want to choose some low hanging fruit for a quick first success that makes sense to business people.

Are all the "system i" ISV's or webblog initiators here on this list heavy sponsored (like 100%) by IBM right now?

Just a question.

PS: IBM, I need to have a customer telephone call registration and call search system for a country wide bank network. Calls are registrated as "*.wav" formats, directories are automatically assigned per local bank, date and callee. I could use the ifs file system to store the calls, growth rate estimated 2GB/day. Pls can you supply 100% funds for this "innovative" project?

I'm just curious. In the past, 'modernization' was coder-speak for arbirtarily rewriting existing and functional applications so that columns are replaced by free format or subroutines are replaced by procedures. In other words, busywork ... adding no value but keeping coders busy. Is this what you are writing about?

Also, is any cost - benefit analysis being applied or are just cosmetic appearances being considered? Is replacing an existing application with a lower cost but excellent purchased application that might work on another platform being considered? Pardon my cynicism, but so many times this type of exercise is more an effort to perpetuate the IT department as it currently is and not improve the company's IT effort.

Is any type of analysis that attempts to compare and contrast Where you Are, Where you Want to Be and Why, and How much will it cost to get there being done? Are user departments being interviewed and are they a vital part of the process ... perhaps even leading it or is this a stealth effort that ignores all users except in a paternalistic and patronizing way?

Decide on a future architecture build a framework then start building/replacing applications (start small first) to the new architecture and bridge back to the old until the old has been completely replaced. Replace the bridges as the new applications are integrated.

Screen scrapping or webfacing is like putting lipstick on a pig.

The biggest challenge to developing modern applications on the iSeries is building the business case and justifying a ROI for the average company. I think most will let the green screen legacy ERP applications die on the vine and eventually replace them with new software and/or hardware. Hopefully more vendors will step up and develop new ERP software for the platform.

See Don Denoncourt's February article "Developing Reusable RPG APIs" if you want to truly modernize your RPG. All the things that are better in Java are easily achieved in RPG and that article shows how.

Where I work, we RPG developers failed in an attempt to modernize the RPG and we now have a mandate prohibiting new RPG development. New development here is going to be .Net, Java and Oracle. We hired Java architects and everything they tell us boils down to what is in Denoncourt's article.

You could also attend the Dallas area Java UG, the Plano Java SIG and the Pragmatic Programmers meetings.

Re: "Screen scrapping or webfacing is like putting lipstick on a pig."
Don't forget that the original Windows was lipstick on the DOS pig ... and dominated the PC world! While a nice face does not replace good architecture, don't underestimate the power of something that is both familiar to the user (GUI) and attractive. Oink! for the iSeries.