The Great RPG MOVE Debate

[coteijgeler@chello.nl]

I think Joe's arguments have more value than IBM's: those do not exist. BEGSR can be used in free-format and for some reasons (we can use subproceduresnow, ain't it?); IBM should use simular arguments for the MOVE opcode. But I think there should not be a MOVE-BIFs (%move, %movel). IBM should port MOVE and MOVEL to the free format, instead of spending time and money on creating all kind of new BIFs to convert from character to numeric. It is appalling that it took IBM three or more OS releases to get the %dec BIF (and some of those others as well) accepting (database) variables, and not hardcoded literals, constants at compile time. Besides that, MOVE is a much easier to read opcode. Just my two Euro cents. Regards, Carel Teijgeler.

[David Abramowitz]

The Great RPG MOVE Debate

Joe Pluta wrote: About 20% have already converted (split 50/50 between /free and mixed mode) About 10% flat out won't go to /free Joe, please keep in mind that these polls are highly unscientific. Readers only represent a tiny fraction of a percent of all AS/400 shops out there. I will go out on a limb, and surmise that the readers represent a more sophisticated strata than the typical AS/400 shop. You're in good company though, IBM makes the same assumptions when determining what enhancements will be applied. I think if you could actually poll every AS/400 shop you might find results diametrically opposed to current results. Dave

[csmith@baycorr.com]

The Great RPG MOVE Debate

I agree with you Jim. I'm an old RPGLE programmer myself and haven't delved into the free format arena yet. I find it difficult to believe that IBM has gotten rid of the MOVE opcode for the reasons mentioned within Joe's article. It just gives me another reason not to get involved with free format just yet. I know I will eventually, but I haven't read enough about it yet to make me jump in. I work in a manufacturing environment where pretty much everything is written in-house in RPGLE. I converted everything to ILE back in 1998 while combing through date routines for Y2K. Even had the pleasure of converting some old S/36 programs to ILE. When I feel the need to embrace free format, I won't be looking forward to removing all the MOVE statements within our legacy code. I hope the IBM compilers spend the 10 minutes putting back the MOVE opcode in free format so the world doesn't have to spend weeks removing it to enjoy all the future benefits of free format. C.M.Smith

[J.Wells]

The Great RPG MOVE Debate

How many people have voted?

[J.Pluta]

The Great RPG MOVE Debate

David, of course these polls are unscientific, but they're the best we've got. These polls may indeed represent only a fraction of the shops, but my guess is that the folks in production are even LESS likely to move to new stuff at the sake of losing the old. What sort of reasoning did you use to come up with the assertion that the results would be "diametrically opposed"? Or are you just saying the results might be different? Because based on common sense, my feeling is the folks that don't bother to read these magazines would vote even more highly in favor of not moving to /free. Joe

[J.Pluta]

The Great RPG MOVE Debate

I'm not sure, Joe, but I believe it's well over 50. Perhaps someone from MCPress can tell us.

[David Abramowitz]

The Great RPG MOVE Debate

Joe Pluta wrote: Because based on common sense, my feeling is the folks that don't bother to read these magazines would vote even more highly in favor of not moving to /free. I guess I haven't made myself clear. We are in agreement. Dave

[buck.calabro@commsoft.net]

The Great RPG MOVE Debate

Joe Pluta wrote: > Seriously, Buck, there's a difference between saying "I > am happy to use /free without the MOVE isntruction" and > "I don't use /free, so I don't care about MOVE". Sure is! To be truthful, I never really thought about it much until you brought it up. To re-iterate my position, I am personally happy to use /free without MOVE. That's beacuse I never convert to /free: I only write new logic in it. > So of the ones who are currently on the fence, > only SIX PERCENT have plans to move to either > /free or mixed mode. A rather astonishing 60% want > either MOVE or the extensions I suggested in the >article (they're split about 50/50 as well). As you say, an interesting result. I wonder how many of those people are using RPG IV in fixed format today? My guess is 'not many.' RPG programmers are almost unique (in the programming field) in their desire to avoid changing 'what works.' I'd love to see the same poll but limited only to RPG IV programmers who have been using RPG IV for 5 years. The idea being to gauge the opinion of people who are comfortable with change, AND intimately understand subprocedures and BIFs. That sounds elitist, but I am a far cry from elite. I have no university schooling, started with cards on S/3. I'm basically a transplanted viola player... All I'm saying is that I honestly think that if someone with my limited skill set and background can adapt to /free, really anybody can. Interesting thread... Warmest regards! --buck

[J.Pluta]

The Great RPG MOVE Debate

"I'd love to see the same poll but limited only to RPG IV programmers who have been using RPG IV for 5 years. The idea being to gauge the opinion of people who are comfortable with change, AND intimately understand subprocedures and BIFs." See, that's the point. Now that the results are in, people like you and Jon are yelling "Fix!" . Every time a poll shows that legacy programmers are the largest percentage of the iSeries users, bleeding edgers want a recount. Why is that? And why is it that mailing lists seem to always skew towards the new technologies? One reason may be that newfangled users (like you, Buck, whether you admit it or not) want neat new features. You have the time to learn them and use them, maybe because you don't have the same issues other shops have. For example, you, Buck, seem perfectly willing to have different flavors of RPG in your shop, and so you don't worry about the effects of /free on legacy code. Many people don't have that luxury, whether it's because of a corporate mandate, or for purely pragmatic reasons, such as at an ISV. My point was, and still is, that the folks who are pushing for all the neat new features are the people who LIKE neat new features, and that those folks are more vocal than the rank and file programmer, and thus tilt the scales. Maybe it's because of job demands. If you have time to play with new bells and whistles, then you have time to hang around on mailing lists and even more time to actually type responses. My guess is a lot of the rank and file folks - the ones who don't want to lose the MOVE instruction - are people who spend too much time doing their job to become active participants on mailing lists. So, no, I don't want to create a new poll for bleeding edge developers. You guys already have that outlet - they're called mailing lists. I wanted to create a simple way for the rank and file programmer to make themselves be heard, and from what I've seen in the last couple of days, they've done just that. Loud and clear. Joe

[JonFParis]

The Great RPG MOVE Debate

> The ideal solution would be to be able to convert the whole program to free-format with absolutely no fixed-format

Take a look at the latest version of Linoma's RPG Toolkit - it is capable of doing an awful lot of the conversion. You can download a trial at their web site - www.linomasoftware.com. The CODE tool is intended more for converting code fragments than wholesale program conversion and it doesn't attempt to convert many things like MOVEs

[J.Pluta]

The Great RPG MOVE Debate

Disclaimer: Linoma is a partner; they sell PSC/400 (rebranded as Envoy). But that does not detract from the fac that the RPG Toolbox is one of the best utilities on the market (although their Surveyor product is also quite nice!). RPG Toolbox is a lot more than a conversion utility, though. The Toolbox is Bob Luebbe's brainchild, and it is a real productivity aid for programmers. Think of it as Textpad for SEU; you get not only the conversion capabilities, but also a myriad of command line functions, including "snippets", which are like macros for SEU. Definitely a five-star utility, and it goes a long way towards addressing the issues of /free. I think it's a telling issue, though, that CODE/400 doesn't even try to convert MOVE instructions. They're what? Left as an exercise for the student? Joe

[buck.calabro@commsoft.net]

The Great RPG MOVE Debate

Joe wrote: > See, that's the point. Now that the results are in, people like > you and Jon are yelling "Fix!" . With the greatest of respect, isn't that exactly what you did by taking a poll _here_ and compared it to the mailing list? > Every time a poll shows that legacy programmers are > the largest percentage of the iSeries users, bleeding > edgers want a recount. Not me. _I_believe_ that legacy programmers are the mainstream of iSeries customers. I also believe there are many more S/36-style programmers than any other RPG folks, and for them this /free question is incomprehensible. > And why is it that mailing lists seem to always > skew towards the new technologies? Because the people likely to subscribe to mailing lists are less insulated from new technologies like list servers. I know I was insulated like that for many years. > One reason may be that newfangled users (like you, > Buck, whether you admit it or not) want neat new features. Guilty as charged. What programmer doesn't want to do better, and use better tools? > You have the time to learn them and use them, > maybe because you don't have the same issues > other shops have. I don't think so. In fact, virtually everything I have learnt was done so during lunch, before or after work hours. I most certainly have no employer training programme to send me on my way to neat new feature nirvana. I probably DO have different issues than other shops because I work for a software vendor, and have to meet a very different set of criteria for my code than I did when I worked strictly in-house. But I think that makes it harder for me, not easier, because I can't roll out /free until the last of our customers moves to a /free release. > For example, you, Buck, seem perfectly willing to have > different flavors of RPG in your shop, But, but, but we already have different flavours of RPG in every shop! Unless one stopped at S/36 RPG II, that is. Certainly S/38 RPG III was different from S/36 RPG II, but that didn't stop people from converting RPG II to RPG III format. Now we have RPG III programs with the style and logic of S/36 programs from a decade before. Then came AS/400 RPG/400, with it's small differences and fun with SELEC and ten or fifteen deep nested IF structures... blech! Then that not-quite-homogeneous mess got converted to RPG IV, but please for the love of all that is beautiful don't try to tell me that it still doesn't smell just like RPG II (now 15 years old). I haven't been to a shop with a single coding style, and I've been to a lot. > and so you don't worry about the effects of /free > on legacy code. I worry about it constantly, which is why I never convert legacy code to /free. I only write new logic in /free and strongly advise others to do the same. > My point was, and still is, that the folks who are pushing > for all the neat new features are the people who LIKE > neat new features, and that those folks are more vocal > than the rank and file programmer, and thus tilt the scales. Agree with you 100%. > My guess is a lot of the rank and file folks - the ones > who don't want to lose the MOVE instruction - are > people who spend too much time doing their job to > become active participants on mailing lists. I'd hesitate to cast aspersions on the motives of people who are active on mailing lists. Some might just be trying to help the community at large. Before I became 'active', I had no PC at work and got to read my email once a night via a Fidonet gateway. Such an arrangement effectively precludes much posting. I suspect that many people remain in a similar situation today; unaware of news groups or list servers in general, and deprived of an immediate connexion to the internet. But please remember that the people who don't want to lose the MOVE instruction are probably the same ones who don't want to 'learn' RPG IV or Code/400 or SQL or triggers or relational integrity or commitment control or in many cases even that newfangled /COPY thing (introduced back in what, 1976?) That would be the SEU-using S/36 crowd who go on plugging away under the internet's Cone of Silence day after day. It's not their fault that they are insulated, but that's how it is. That's how it was for me. And that's why we don't hear from them very much. > So, no, I don't want to create a new poll for bleeding edge > developers. You guys already have that outlet - they're called > mailing lists. I work on telephone rating and billing; the salt of the earth as far as applications go. Not nearly as sexy as MRP. I _hope_ it's a compliment that you think of me as bleeding edge, or somehow 'elite'. It's funny how a formerly useful term became pejorative. It used to be that the elite were like the SAS or Green Berets; something to aspire to, and to look up to: to emulate. You know, to become the best that you can become. Now, to be called 'elite' or 'bleeding edge' has become a politically correct euphemism for 'out of touch with reality.' I deeply apologise to all whom I've offended with this. I'm a dinosaur who believes that doing a better job is a good thing, and that better tools can help me do that. If that makes me elite then I shall wear that badge with honour; especially considering my very humble beginnings in this field. The truth is that I agree that the majority of RPG people want a MOVE BIF/equivalent in /free, and all I was ever trying to say was that if a retard like me has no problem with a MOVE-less /free, then it must not be much of a problem. I am clearly wrong, and my backwardness is evident to all. Finally, some practical advice to all the good readers who want MOVE added to /free: Submit a Design Change Request (DCR) to IBM and explain how you will be impacted by the lack of it. A link to the form can be found in the Midrange FAQ at http://faq.midrange.com. Very best regards to all who took their valuable time to contribute to this thread, --buck

[rfillmore@gmail.com]

The Great RPG MOVE Debate

It seems that when there appear to be roughly equal arguments on both sides that the issue of compatibility should be the the overriding factor. While the new features of the free form spec are attractive, as long as the majority of programmers in my shop have to support RPG II, III and IV (fixed format) code, there is a strong need to be able to transfer their existing knowledge if there is to be _any_ use of the /free features. I fully agree with previous statements that MOVE is an integral part of RPG, and if someone can't be bothered to read the manual for that, how can they be expected to understand %EDITC(myNum:'X');? Granted, the compiler development team should be looking at new features and functionality, but if the goal is to create a new language, then they will lose a significant portion of their user base. New languages may be great for new development, but how many companies do you know that are starting a completely new project in RPG IV Free? What is that market share as compared to the existing "legacy" users that deal with a mix? If I lose functionality or transferability by switching to a new language, can you guess how often I will use the new language? Only when I absolutely have to to get the job done. Is this really the environment we want to have for new RPG development by the majority of users? Oh, and by the way, I have been coding RPG IV for over 5 years, Java for 4 and CGI in RPG for over 5 years, all on the AS/400/iSeries. Backward compatability is what keeps RPG a viable language for the vast majority of users, not the latest collection of opcodes that can only be used in the beta release of the latest OS version with 5 PTFs installed.

[J.Pluta]

The Great RPG MOVE Debate

"With the greatest of respect, isn't that exactly what you did by taking a poll _here_ and compared it to the mailing list?" I don't think so, Buck. On the mailing list, there were perhaps a half a dozen diehard "get rid of MOVE" folks, and another handful of people who thought I was complaining over nothing. My thought was to get a larger audience and see what they thought, as well as make it easy for people to cast their vote without having to wage a battle of words with some of our more, shall we say, assertive mailing list pundits. Let's face it, in order to have an unpopular position on the mailing list, you better be willing to take some serious abuse. An anonymous poll removes that particular bugaboo. So here, there are over 120 anonymous votes. That's a far larger statistical universe, and of those people, nearly 80 have voted that they want some support of MOVE. Only a dozen have moved to /free, only 8 or 9 to mixed mode. If you chart the trend, that indicates that the larger the audience, the more folks want any changes to be at least somewhat backwards compatible. But here's the REALLY frightening thing, to me. Of all the people voting, only SEVEN are currently considering moving to /free or mixed mode. SEVEN, Buck. As opposed to nearly 80 who say they would move if IBM made some changes. That's about as clear a mandate as I've seen in some time. "But please remember that the people who don't want to lose the MOVE instruction are probably the same ones who don't want to 'learn' RPG IV or Code/400 or SQL or triggers or relational integrity or commitment control or in many cases even that newfangled /COPY thing" Buck, I think this *IS* an elitist sentiment, in the perjorative sense of the word. Did you do any sort of research at all to come up with this statement, or aer you just trying to extrapolate your own views onto a group of people you know nothing about? Because from the numbers in the poll, the majority of programmers are ready and willing to go to /free, as soon as IBM fixes the MOVE statement. This sort of deflates your argument. The conclusion I draw from the poll numbers is that most people think removing the MOVE istruction was an unwise choice. And no, you're not backwards, Buck, I just think you're a little unusual in both your skill set and your attitude. That being said, I appreciate your input and especially your suggestion to submit the DCR to IBM. Joe

[buck.calabro@commsoft.net]

The Great RPG MOVE Debate

Joe said: > So here, there are over 120 anonymous votes. That's > a far larger statistical universe, and of those people, > nearly 80 have voted that they want some support > of MOVE. Like I said, I totally agree that given their druthers, most folks would prever MOVE in /free. I never doubted that for a moment. >> "But please remember that the people who don't >> want to lose the MOVE instruction are probably >> the same ones who don't want to 'learn' RPG IV >> or Code/400 or SQL or triggers or relational integrity >> or commitment control or in many cases even that >> newfangled /COPY thing" > > Buck, I think this *IS* an elitist sentiment, in the perjorative > sense of the word. Did you do any sort of research at all to > come up with this statement, or aer you just trying to extrapolate > your own views onto a group of people you know nothing about? Seat of my pants only. I've been in something like a dozen shops and done programming work for a few beyond that. I work for a software company with over 50 customers, and I've experienced their shop standard practices (having to code/design to their requirements.) A previous owner of this company (much larger than we are) specifically mandated NO RPG IV, NO commitment control for a brand new project in the 1999-2000 time frame. I've been interacting with midrange programmers since 1978 and I like to think I have a fair grasp of the culture. I freely admit that I might be very wrong. But how many shops have adopted RPG IV in any form for ALL production work? >I appreciate your input and especially your suggestion >to submit the DCR to IBM. Not a problem! I figured I might as well tell folks how to 'do something' rather than just hope for the better. Just to clarify: 1) Joe and I are friends, but we're not clones 2) Joe and I have a difference of opinion on how IBM should spend their development effort 3) There's plenty of room in the world for different opinions. 4) I don't use MOVE in NEW fixed-format RPG IV code, and that's a big reason I don't miss it in /free. 5) Other people code differently than I do: That's A Good Thing! 6) I don't object to the existence of MOVE in /free; if IBM puts it in, I probably won't use it except for very rare circumstances. I seriously urge all the folks who want MOVE in /free to submit the DCR. If you don't tell IBM, they won't know you want it. Again, it's in the FAQ at http://faq.midrange.com Good luck! --buck