Because the System i can run at redline speed all day long . . .
I had the opportunity to chat with one of the System i world's top RPG experts recently -- Paul Tuohy, who is the CEO of ComCon, a midrange consultancy company based in Dublin, Ireland. He's worked in the development of IBM Midrange applications since the 70s, and in previous jobs was the IT manager for Kodak Ireland Ltd. and the technical director of Precision Software Ltd. He's also an author, teacher, and one of the core speakers at the RPG & DB2 Summit conference.
Oh, one more thing: I must admit, we did talk for a bit longer than two minutes.
CM: What are the big issues when it comes to RPG these days? Some people have been saying, for instance, that IBM has left RPG to die, but what are you seeing out there in terms of what IBM's been doing, what developers are doing, and what they need?
PT: I don't think IBM has in any way left RPG to die. The work they've done on RPG since V3R1 with all the RPG IV stuff has been quite phenonmenal. So in terms of IBM letting RPG go, I don't think that's the case. I think the biggest problem has been with us as programmers getting to grips with it and adopting it. It's now that we're starting to see people taking up on things like ILE. It was only four or five years ago that we began to see the uptake on RPG IV. And what we're starting to see now is . . . a lot more interest in the ILE side, so these are people who've been working with RPG IV for a few years and are now starting to move into the new stuff like gro and dip into ILE. To me, that's been the biggest drag for RPG over all these these years, and that is just how slow people have been to adopt it and begin using all the great stuff that's in the language.
CM: That's a double-edged sword for IBM. I mean, the company has done a remarkable job of ensuring that the System i and all of its previous iterations have been compatible, upgradeable, and functional for everything that has come before them. There are few instances in which customers have had to make radical changes to existing applications to get them to run. Do you think with RPG we've had a similar problem?
PT: I agree. Unlike other platforms, the problem that you have is, stuff you had written for System/38 back in 1979 still runs on the system perfectly today, even without recompilation for crying out loud. In terms of other platforms, that's just ridiculous. I don't know any other platform that will give you that. I know that on certain versions of Unix that I've worked on, even when you went from something like version 5 to version 6, you'd have to recompile your applications to get it to work.
CM: So you think it's hard to get developers to move when everything still works?
PT: It's that old adage, 'If it ain't broke, don't fix it.' That's something that I've never subscribed to -- that's not what maintenance is about. If something comes along that's easy to implement that's going to make life easier for you in the future, then you implement it. It's the same as maintenance with anything. You look at guys who work in plants on the maintenance with machines, they'll tweak them and enhance them as they go along. Unfortunately, if more people had been enhancing slowly over the years, they wouldn't be facing what they're facing now with their applications, which is they have to effectively rewrite half of them. They're looking at a much bigger job in one lump than they were before, which puts them in the position of managers saying, 'Well maybe we should be looking at a different package or a different platform or a different solution altogether.'
But in terms of RPG and progress of RPG during the last 15 years, to me the biggest problem has been the adoption of the changes that IBM has brought out -- not that IBM isn't doing anything. I think they've been doing a phenomenal job with RPG.
CM: What about things like EGL, Java, PHP, Groovy, Grails -- what about RPG alternatives? How are they fitting these days?
PT: It's a slow adoption that is really the problem here, and part of the issue that I would have had with things like Java is that they've been seen as a replacement for RPG as opposed to something played along with RPG. If you're looking at the modernization of any application on System i at the moment, which is more than likely going to be toward the web or in a couple of years toward these rich client things we're going to be developing through EGL or using Eclipse as a base for the client. But if you're looking at the modernization of a pure RPG application, half of the RPG is going to disappear anyway. Because if you look at any green-screen application, probably half the code there is just for dealing with the interface. Regardless of whether you use something like CGIDEV2 or you use something like Java or whatever, more than half that code will go away. In a true modernization process what you'll want to keep is that good business logic, which is written in RPG and handles all the database stuff so well, etc., that's the core that will have to remain there. So in terms of things like EGL and Java, to me they are just tools for dealing with an interface, and that's the way they should have always been approached, depending on the size of company you are and what your needs are.
Java for multiple reasons has failed to deliver. I think the big question mark over EGL is how much IBM will decide to charge for it -- because I've got a funny feeling that EGL is not going to be free . . . and I think that will basically exclude it from 70 percent of the base, because if you're going to be paying for it, there are 50 other vendors who will gladly sell you something that will, for all intents and purposes, do the exact same job.
CM: Is that idea of having it free, do you think it's important to the health of the System i in terms of the market and in retaining customers?
PT: I think there's going to be an interesting year ahead because with this whole move of the Toronto Labs now effectively becoming a section within Rational, and Rational's background is as a software company, and obviously any software they develop they are going to want to charge for it. And their model doesn't really fit with the traditional System i model where you buy the machine and the software comes with it and it's a big bang -- it all comes in the box and you order it in one go. If it's a thing that, for example, you're going to get a base WDSc that will do your RPG stuff as standard, and then, oh, if you want to do EGL, well that's another $200 per programmer or $2,000 or whatever they decide is the figure, I think System i people are going to balk at that because it's not the way they are used to buying software. Of course, if they were PC or Unix developers, that's exactly how it works. It's every little plug-in that you want, you buy. We've been out of whack with the rest of the industry for so long, so it's going to be really interesting to see if IBM can convince people to buy software when they need it.
Posted by cmaxcer at October 30, 2007 11:38 AM

| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | 29 | 30 |
We welcome your comments and opinions and encourage lively debate on the issues. However, Penton Media reserves the right to delete or move any content that it may determine, in its sole discretion, violates or may violate its Terms of Use or is otherwise unacceptable. For more information, see Penton Media's Terms of Use.