What Does IBM Need to Do with RPG?

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
** This thread discusses the Content article: What Does IBM Need to Do with RPG? **
0

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
I agree that RPG IV is moving too fast, although I'm not sure it has to be slowed down the same way as suggested here. I like the compiler directives on the H specs and the more BIFs the merrier, but the free format stuff is just overkill. IBM seems in a big rush to turn RPG into JAVA and I don't think its necessary. Mr. Cozzi is right on when he talks about technical expertise coming through tech newsletters. I find all the free format RPG stuff worthless and not worth my time to learn. Mind you I'm and ardent user both of RPG IVs new features as well as the ILE concepts, procedures, service programs, etc.

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
I agree with the article almost completely. As much as I love working on the 400 and love working with the latest and greatest techniques in RPG, it (RPG II/III/400/LE/Free) has become fragmented with the various OS's, which hurts all of us in the long run. Since I love new programming techniques/capability I would be disappointed if I knew there weren't going to be changes/enhancements for 3 or more years. When I hear about a new release of OS/400 the first thing I'm excited to check out are the RPG enhancements and what cool new things I can do. I would hate to lose that... Perhaps IBM is using that as a marketing ploy for the programmers to push for the latest and greatest upgrade? Gotta have it! But, as Bob points out, the programmer community does end up becoming fragmented. There isn't a good 'college' major for RPG/IBM systems analysis studies readily available for everyone, and when you can't get the expensive training that is available by the third party people, you either buy a book written by Mr. Cozzi, or you come to these web sites for assistance. Now Robert has to have multiple releases of his books for the OS releases, and it becomes too difficult for users to help each other because we are at different levels. RPG staff starts fading, the language gets a bad name, corporations don't want to use it anymore, and we are all out of luck. In closing I will say that RPG has grown so fast recently that, despite my feeling of being a proficient programmer, it has grown beyond me with the time I have available to experiment with the new capabilities/techniques that are out there. It will probably take me three more years to try out the new stuff. So I'm okay with putting a freeze on the current free-form RPG. What I think should happen however is give it a new name. RPG after all stands for "Report Program Generator" right? It is certainly WAAAAY more than that and has been for over a decade. Despite where it came from it would be hard for a new programmer to go back and look at the original code (RPG II) and even be able to tell what it is. Let's give it a new launch name I can be proud of spewing to my 'C++' and 'Java' friends and then work on this new staggered implementation schedule that Bob is talking about! jefferson

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
I think it's great that IBM is constantly updating RPG IV. For those who haven't made the jump to V5, they only have a few months of support left, and incompatibility is the price they have to pay for lagging behind. You know what they say, fail to evolve, and you'll end up like the dinosaurs. As far as the complaint about the three different ways to chain to a record, I think I'm safe in saying that the new methods are going to be considered "preferred" over the old. The old method is hard to read (don't you love searching through the code to find which fields are in the key list, and then having to relocate the chain command?), and is probably only supported for backward compatibility. Same thing goes for BIFs vs. indicators. I found it interesting last week at COMMON that I saw a speaker, obviously proud of his code, remark about how much of his subfile program flow he could control with one indicator. I'm sure since he writes most of his RPG this way, it makes perfect sense to him, but to me, it really would be unreadable without comments. Sure a good programmer will thoroughly comment their code, but it makes it much easier if they don't have to explain what's going on each time they use an indicator. God forbid if someone else comes along with a modification and doesn't bother to update the comments. That's a great segue into my next point. Another post complained that Toronto isn't adding much new functionality, but "... having a programming language do the same thing in a different way." I would argue that most of the enhancements made in the last two releases either 1) made it easier to do something somewhat complex in RPG (i.e. data area handling), or 2) made the code much easier to read (i.e. the new chain syntax). The latter being much more important (to me) than the former. Who want's to rely on comments, when the code automatically documents what's going on (i.e. IF *IN01 vs. IF %FOUND(my_file)). In conclusion, stay at the lowest supported release if you want, but don't complain about the features that you miss. I've actually published utilities in the trade press, and I've only had one person complain that my code wouldn't compile due to the version/release issue. Thankfully, that reader had an upgrade scheduled for the next month, and wrote back that things worked great after they finally got back on a supported version of OS/400. Finally, for the poster who complained about not being able to keep up, do what Mr. Cozzi says and read the trade press. According to him, there should be plenty examples of the new features put to practical use there...

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
I've been programming in RPG since the System/34. I've gone thru the System/36 phase and on into native AS/400 (oh excuse me...iSeries if you must). Anyway, it's been my experience that you use what produces the desired results in an efficent, timely manner for those who use your results (programs) to acheve their results (data manipulation, lists, etc.) and if you can do it quicker and more accurately using the PROVEN "old" techniques and you also can understand your code better, then so be it. Just do it. In my years I have found that there is no one person in upper management who gives a damn about what your code looks like or what techniques you use when writing a program. They are just concerned that their bottom line results are accurate. And after all, those are the people who signs my cheques and as far as I'm concerned, they're the ones I have to please. And yes, I still use the RPG cycle, indicators and such and proud of it. To hell with this free-form shi*.

[buck.calabro@commsoft.net]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
While I agree that it would be nice to have ONE stable version of RPG that I could code to, and I agree that a public comment period would be tremendous, I'm not at all convinved that the additions to the language are crippling. Bob's remark that if he publishes /free code some 85% of the iSeries population can't use it is is dead-on. Only a tiny fraction of RPG customers are up to date. I suggest that this inertia to "keep doing the same thing" is what the problem is, not the pace of change in the compiler. --buck

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
Bob Cozzi is one of the last real representatives of our AS/400 world. Yes I said AS/400 because I find trying to say e(logo)server iseries is ridiculous. He is also one of the few people I will take the time to read these days. I remember wishing there was a high level language when I was an Assembler coder. Then COBOL and RPG came along and it was wonderful. With RPG you could simply fill in a couple of columns and forget about coding Level break and Matching record logic. Not having to keep track of register addresses and pointers was also wonderful. Now it seems like everyone wants to be an assembler coder. Personally, I haven't found the need to use most of the new features of ILE RPG. I also have to laugh at many of the BIs(%), many of which make the code much more difficult to read, if you didn't write it. What IBM needs to do is to slow down the C++/JAVA coding style and focus more on direct access to the IFS and add graphical op codes. They also need to lose Lou's anything for a buck mentality. He's gone now, so IBM can get back to re-building the company instead of tearing it apart. Bob thanks for saying the things that need to be said, the quiet majority agrees.

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
1) Rename it! (You're not scared to change the AS400 to eServer iSeries, why stick to RPG? I had a PC biggot bash RPG yesterday based on the name and his view based on RPGII code exposure.) Application Program Generator - APG 2) Don't Screw it up. Making it look like C sucks period. When you add new functionality make it intuitive. 3) Give us the ability to natively access DB2 Universal, MYsql, etc.. so we can use it on a PC. 4) Fixed format RPG to me brings more to the table than does FREE format. I can write pgms to fix thousands of programs if it has a structure. Also within columns/fixed format you have cursor sensitive help text.

[Guest.Visitor]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
Let us remind ourselves that RPGIII in the 1980’s gave us IF/THEN/ELSE, READE, DOxx and AND/ORxx. Without it, our brains would have become completely numbed by simulating structured constructs with redundant indicator manipulation. So who knows, maybe we really do want or need free-format RPG, where we can continue with the RPG tradition of conciseness, combine statements, lop off the EVAL and the Tab, Tab, Tab. That might be nice. Ironically, Bob Cozzi has been and still is, I suspect, a great proponent of all sorts of iSeries improvements (don't get tired Bob, you can really get on a good roll at times!). So I think we would want to keep Mr. Cozzi busy with all sorts of topics for advancement (did I read you can’t use MOVE with free format?), pressing forward on a timely basis, even if somewhat ad-hoc. Plus at any time, there is something to grumble about (who wants to wait 2 or 3 years to grumble?). This article is a refreshing twist because it provides an opportunity to comment on the trade press’s self-imposed pressure to discuss that very latest technique. I know I have a tendency to feel I don’t have time to study something now if I won’t be able to use for 9 months (off it goes to the reference library?). So my preference goes for trade press articles that help us more fully utilize what is relatively new (with a little what if?/grumble for IBM sprinkled in), while IBM keeps up that unique and proactive iSeries upgrade strategy.

[S.Mildenberger]

** This thread discusses the article: What Does IBM Need to Do with RPG? **
I feel the evolution of the RPG compiler is in the right direction. The newer versions (with free format) allow code to be written that is much more readable and I know that I can be much more productive in the world of ILE than without it. I see many of the complaints about too much changing are from people who aren't keeping up and they just don't realize what they are missing. If some people are satisfied with what they can do today then fine, they can stay right where they are. But, I am going to continue to expand my skills and productivity and the enhancements to the RPG compiler are a big part of that, as are the client side tools like CODE. Scott Mildenberger