The Slow Demise of 4GLs

On a regular basis, web searchers come to my site looking for advice on how to eliminate legacy 4GL applications.

In the late-1970s, some innovative software vendors came out with “fourth-generation computer languages” with the goal of eliminating COBOL, the standard “3GL” of the day. Needless to say, these companies missed their marks.

Three decades later, COBOL is still alive and kicking. Funny thing is, so are some of the 4GLs.

At the time, it seemed pretty obvious that data processing professionals would love to reduce the complexity of building software applications. The 4GL vendors offered programmers the same capabilities as COBOL but in a more succinct and easy language. But COBOL coders gave the vendors the cold shoulder.

When you have a product nobody wants to buy, you become creative.

“Okay,” these 4GL vendors said, “we understand that you want to stick with COBOL. But how about your business users? Users can’t handle power tools like you can. Let’s give them something safe and easy to use — just imagine, they could build their own report programs and leave you alone!”

And so, companies all over the world started giving business users 4GL application development tools for DI…

On a regular basis, web searchers come to my site looking for advice on how to eliminate legacy 4GL applications.

In the late-1970s, some innovative software vendors came out with “fourth-generation computer languages” with the goal of eliminating COBOL, the standard “3GL” of the day. Needless to say, these companies missed their marks.

Three decades later, COBOL is still alive and kicking. Funny thing is, so are some of the 4GLs.

At the time, it seemed pretty obvious that data processing professionals would love to reduce the complexity of building software applications. The 4GL vendors offered programmers the same capabilities as COBOL but in a more succinct and easy language. But COBOL coders gave the vendors the cold shoulder.

When you have a product nobody wants to buy, you become creative.

“Okay,” these 4GL vendors said, “we understand that you want to stick with COBOL. But how about your business users? Users can’t handle power tools like you can. Let’s give them something safe and easy to use — just imagine, they could build their own report programs and leave you alone!”

And so, companies all over the world started giving business users 4GL application development tools for DIY reporting purposes. In the 1980s and 1990s, many organizations researched and selected one of the three major 4GLs: RAMIS, FOCUS, and NOMAD.

Today, firms are trying to figure out how to clean up decades of unmanaged end-user 4GL application development. One of my clients estimated that rewriting the hundreds of 4GL libraries used by their business groups might cost over $4 million USD. Another was shocked to discover somebody had used what management considered to be just a reporting tool to develop a mission-critical shop-floor scheduling application.

I know I personally used 4GLs to write many critical applications that were downright difficult to replace. I am willing to bet that 80% of my American readers have used consumer product goods coupons generated by one of my 4GL applications. Another easy bet is that almost all of the readers have drunk a brand of coffee whose purchasing, quality testing, blending, and inventory processes were managed by 4GL applications.

There are problems with these 4GL applications. First, they were most likely written by somebody who left the company years ago. The code is handed down from one poor soul to another until it becomes unsupportable. If the IT organization turned a blind eye to end-user 4GL development, no standard app dev methodologies were ever followed. To make matters worse, not a single college today teaches 4GL application development to its students. But the major problem today is that business users still depend on 4GL output.

If you are in this situation, I can help by providing software that can inventory and scan your 4GL applications, automatically analyze them for functionality, and even convert them to other BI products. A spreadsheet generated by the scanning process will give us the first step in planning a cost-effective 4GL elimination strategy.

If you are interested in learning how I have helped other companies replace their 4GLs, contact me.