Here is a link to an article on RPG
Now as to staying with the current application.
First question is what is the driving force behind the thought of changing?
If it is something that can be replaced with a new package, you need to weigh the total cost of replacement: Software, Hardware, Data Conversion, Parallel processing, Training, etc.
If this is something you just want to rewrite from scratch, then I ask WHY since you say the current software is still effective.
Unlike applications on other boxes, RPG code written over 20 years ago still can run fine. Granted there are new techniques that may make it run faster, but the end result would be the same.
Are you looking for new features that your current software does not provide?
That is a different question.
You may chose to buy or write something in RPG or another language on this box of a different one. Intergration should be relatively easy.
So it it not just what to do, but more important is WHY do you want to do it.
Just because it is "OLD" is not an acceptable reason.