did anyone ever program in RPG?

[Don Shepherd]

When I graduated from high school in 1968 I attended a computer programming school called ECPI (Electronic Computer Programming Institute). In that year-long course we learned about IBM unit records equipment (which was being phased out by then) and three programming languages for the IBM 360: BAL (assembler), COBOL, and RPG (Report Program Generator). I then attended college for 4 years and got my BA and another year and got my MA, then I went to work as a programmer, writing FORTRAN programs for the Univac 1108. I eventually wrote COBOL and FORTRAN programs for the DEC-20 and VAX. I never actually programmed for the IBM 360.

I'm wondering if anyone here ever programmed in RPG. Of the three languages I studied in 1968 I remember RPG was my least favorite; assembler was my most favorite. You couldn't write an RPG program on just regular paper, it required about 4 or 5 specific types of coding forms as I recall. It was not a procedural programming language like FORTRAN and COBOL. I could see that it might be useful in certain simple file-processing applications, but looking back on all the programs I wrote in my career, I doubt that RPG would have been especially useful in any of them.

I think I consider RPG similar to HTML in that if you use them, you're not a "real" programmer.

Any thoughts?

[Duane Hess]

I programmed in RPG-II a small amount in the early (mid?) 80's.

BAL was my favorite (on a S/360 model 30) and on S/370 & 9370 systems under VM. Most all programs written by me were in S/360 Assembly or FORTRAN. I have programmed in COBOL also.

A former student offered me a limited-term, part-time (weekend) job writing RPG-II on a System/36 (S/34 upgrade) for some "catch-up" programs their company wanted. Basic database extract/report & data entry appliations.

RPG-II was surprisingly easy, quick & effective. A first review of the language makes it appear "twinky" and difficult. Once you used it & learned how to use and when to "ignore" (control) the program-cycle the "lights will go on."

The System/36, although a general purpose computer, had the operating system/development environment closely tied with writing RPG. Applications with a moderate amount of conditional testing of data and simpler programs are a breeze in RPG-II.

Such programs could be written an about a 1/4th the time as COBOL and developed & running in 1/3-1/5th the time of COBOL, especially on a System/36. As the S/36 had the editor, screen development and generally the OS, in appearance terms, could much be described as a RPG program development environment.

I never used a System/38, but am somewhat familiar with the product. It used RPG-III, which added IF-THEN-ELSE directly as a language construct as opposed to the "tricky" and often convoluted flag-structure of RPG/RPG-II to accomplish the same. (been too long to remember details). I'm pretty sure RPG-III also had the COBOL equivalent of procedures (PERFORM).

Man, haven't thought about all that in years!

[Don Shepherd]

Thanks Duane, that's some great experience you have had, I appreciate your insights.

Although I always worked on the "big iron", it always seemed to me that to have worked on a small system like an IBM System/3 would have been a great way to do programming.

It's good to hear that RPG was not as funky as I had assumed.

[Tugdual]

(07-02-2015 06:03 PM)Don Shepherd Wrote:  I think I consider RPG similar to HTML in that if you use them, you're not a "real" programmer.

Interesting comment. I'm working with Qt Qml which I found somehow similar to HTML5; this is the UI part with some dynamic aspects.

How would you define what is coding and what is not?
To me this question is the same as segregating brain and guts; they are both damned important.
As a matter of fact an application includes a great number of aspects and like it or hate it, you'll find that the time you spent building interfaces, storing and retriveing data is equally if not more important than the writing of the core application that performs the expected features that justified an IT project at the first place.

[Don Shepherd]

(07-03-2015 11:01 AM)Tugdual Wrote:  How would you define what is coding and what is not?

Well, I'm old-fashioned, in addition to being just "old." I cut my programming teeth in the 1970's on procedural programming languages like FORTRAN and COBOL and BASIC. Years ago I setup my own website that includes links to all of my favorite places. It's just a simple text file I update occasionally with Notepad when my favorites change. I don't consider that real "programming" and I don't consider people who maintain simple html files "programmers."

If you write "IF....THEN....ELSE...." then you are a programmer.

If you write

<td><a href=http://www.speakeasy.net/speedtest/><font size = 7>Internet speed</font></a></td>

then you are not a programmer.

That's just my opinion.

[Joe Horn]

(07-03-2015 01:55 PM)Don Shepherd Wrote:  ...I don't consider people who maintain simple html files "programmers." ...

I totally agree. Even the name of HTML implies this; it's a "mark-up language", not a programming language.

[Tugdual]

The good news is that there is a lot of material to reheat your IT flame.
First on a conceptual standpoint I believe that IT activities stop when a comprehensive solution is delivered and this goes way above writing lines of code for the core features.
On a technical standpoint, there is more and more a convergence of many concepts into new solutions.
Do you know that most of the HTML code you see is actually generated by ASP and/or PHP servers? Have you seen how Java is embedded into HTML; google apps are the pinnacle of Java/HTML.
Are you aware of the possibilities of XML technologies? XML is a markup language that you can use to store structured data, control format with schema, transform data from XML to anything else, produce HTML or DOCx reports?
Trust me, the era when HTML was a markup language and Basic the real thing for programmers is far far away. I am myself amazed by the new possibilities and I have invested a lot of time into Qt which is free and available for anybody curious to explore solutions based on parallel multi-tasking, with their new QML interface definition language which is close to HTML.
By the way, I'm not an enthusiastic teenager or a young adult, there is no age to enjoy new things :-)

[Don Shepherd]

(07-03-2015 05:50 PM)Tugdual Wrote:  The good news is that there is a lot of material to reheat your IT flame.
First on a conceptual standpoint I believe that IT activities stop when a comprehensive solution is delivered and this goes way above writing lines of code for the core features.
On a technical standpoint, there is more and more a convergence of many concepts into new solutions.
Do you know that most of the HTML code you see is actually generated by ASP and/or PHP servers? Have you seen how Java is embedded into HTML; google apps are the pinnacle of Java/HTML.
Are you aware of the possibilities of XML technologies? XML is a markup language that you can use to store structured data, control format with schema, transform data from XML to anything else, produce HTML or DOCx reports?
Trust me, the era when HTML was a markup language and Basic the real thing for programmers is far far away. I am myself amazed by the new possibilities and I have invested a lot of time into Qt which is free and available for anybody curious to explore solutions based on parallel multi-tasking, with their new QML interface definition language which is close to HTML.
By the way, I'm not an enthusiastic teenager or a young adult, there is no age to enjoy new things :-)

Tugdual, yes, I am vaguely aware of most of the things you mentioned in your post. I know the world has changed since I programmed for a living and nobody uses IBM punchcards anymore or BASIC on teletypes. But that is my background, and it served me well (not to mention that it currently gives me a generous pension check every month, something the current generation will probably never know). I'm not going to go downtown and stand in front of a building with a sign saying "The people inside this building who call themselves programmers really aren't!". It just doesn't really matter to me.

I still enjoy programming little things on calculators and my Tandy 102. It's just fun, the way programming ought to be (but frequently is not, as I know from working in the field for many years).

Don

[BobVA]

(07-02-2015 07:27 PM)Duane Hess Wrote:  ...A former student offered me a limited-term, part-time (weekend) job writing RPG-II on a System/36 (S/34 upgrade) for some "catch-up" programs their company wanted. Basic database extract/report & data entry appliations.
....
Man, haven't thought about all that in years!

I had a similar experience; one of my professors had been called by a local bank looking for a hungry student to fill the same type gig.

The hardware was a TI-990 running DX-10, I think. Nice little mini-computer, but it had a 9-track drive, designed for vertical mounting, installed horizontally. This was a continuing source of entertainment, particularly when loading.

I didn't use RPG, but something vaguely similar called "SORTMRG" which was a "programmable" sort / merge utility, sandwiched between a couple of calculation modules I'd written in FORTRAN IV.

I really miss that 132-column printer and the green-bar paper. *That* was programming :-)

[Don Shepherd]

(07-03-2015 07:36 PM)BobVA Wrote:  I really miss that 132-column printer and the green-bar paper. *That* was programming :-)

Me too.

One time I made a one-line change in a COBOL program, compiled the new version and got like 1000 error messages (I'm probably exaggerating a bit). I went to the systems guy and accused him of loading a new COBOL compiler on the system without testing it first, because how could a one-line change cause so many errors. He dismissed me, rightfully so it turned out.

I had unknowingly deleted a period at the end of one of the COBOL division lines.

I apologized. He laughed. Those were the days.

Another time we had a deadline to complete a program on Friday. We were not anywhere near done on Friday afternoon, so my partner and I stayed all night Friday night, all day Saturday and Saturday night, about half the day on Sunday, finishing Sunday afternoon when we finally went home. But we could status the program as complete on Monday morning.

I have to wonder if programmers do that kind of thing today. I suspect not.

[Massimo Gnerucci]

(07-03-2015 08:18 PM)Don Shepherd Wrote:  I had unknowingly deleted a period at the end of one of the COBOL division lines.

The most common error source in COBOL programming... :)

[Paul Berger (Canada)]

One of our first FORTRAN assignments in college was to plot the trajectory of a cannon ball fired from a tower until it struck the earth... it is amazing how far that cannon ball could tunnel into the earth before the output page limiter, that was wisely in effect, kicked the program out. I was spared the embarrassment of receiving a large wad of paper the next morning as I had a crude programmable calculator and discover the danger of setting the exit point to be when the altitude = 0.

[rprosperi]

(07-03-2015 08:18 PM)Don Shepherd Wrote:  One time I made a one-line change in a COBOL program, compiled the new version and got like 1000 error messages (I'm probably exaggerating a bit)....
I had unknowingly deleted a period at the end of one of the COBOL division lines.

When learning Focal (the original DEC PDP-8 version, very much like Basic) and Fortran in High School in the 70's, the Prof (or Math Teacher I suppose is more correct) told us this exact horror story about Cobol (i.e. 1 missing period=pages of errors). The result was exactly the same for the whole group of us in that class, verified many years later - that result being we all avoided Cobol like the plague, and we also tested it to see if it was true. As you all know, it was true.

Funny to see others had the same experience.

Though I hear one can make damn good money today with good Cobol skills... so the last laugh may not be ours.

[Don Shepherd]

I omitted one significant detail from the story above about working all weekend long to meet a schedule date.

When my coworker and I came in Monday morning after that long weekend, we discovered to our horror that the program(s) we had been modifying and testing all weekend had reverted to their version of Friday evening. ALL OF THE CHANGES WE MADE ON SATURDAY AND SUNDAY WERE LOST!!!

I went and found the systems guy and he said that, because of some anomaly, he had to restore the system (it was a VAX 11/780 as I recall) to its state as of Friday evening. He didn't know that we had worked all weekend, obviously. I was ready to kill, all that work down the drain. But then I remembered that we had hard copy code listings of all the changes we made, we just had to make them again Monday morning. I recall it took a couple of hours.

After that incident, I rarely threw away old code listings.

[jebem]

(07-02-2015 06:03 PM)Don Shepherd Wrote:  (...) three programming languages for the IBM 360: BAL (assembler), COBOL, and RPG (Report Program Generator).

During those days I never had to touch an IBM mainframe directly, but I have worked closely with people working on them, when we had to integrate our WANG computers with the IBM systems. Wang VS series did support RPG II despite I really never had to develop on it as we had our own native languages including a scripting Procedure Language.

I worked about 20 years for WANG Laboratories starting in 1981. The WANG VS series computer systems, based on the IBM 360/70 instruction set, had a full set of networking sub-systems to guarantee interoperability with IBM systems at the time.

Our WANG systems could communicate with IBM mainframes to do file transfer by emulating the IBM RJE 2780/3780 protocol.
WANG did implemented the IBM SNA architecture as well at the 3270 Control Unit level to let the WANG systems behave like IBM 3278 Terminals.

Wang VS computer series used an interpreted scripting language called VS Procedure Language (VS/PL), in addition to the other compiled languages (Assembly, BASIC, COBOL74, C, PL/I, FORTRAN4 and RPG II) and 4th generation application development languages (PACE).

Dr. An Wang was the founder of WANG Laboratories and a brilliant inventor, holding a significant number of patents at the time, related to memory (magnetic and later developed the SIMM), calculators (specially for scientific markets), data communications, touch pad and printer technologies, among others. He did work closely with/for/against IBM as well.

In many ways, WANG laboratories were ahead of its time, when they integrated text, voice and image in their terminals and office automation applications during the mid 80's.
Wang FreeStyle was another futuristic product in 1990, when WANG released the first touch pad device to the market, integrated with all Wang others solutions.
I guess the world was not ready for that kind of technology. We had to wait 20 years to see Wang ideas come to mass market.

I was visiting Brussels at the Wang headquarters (close to NATO facilities) in the late 80's when we noticed the Wang flag was half mast, and then we heard about Dr. An Wang death.
It was a sad day, and after that Wang laboratories business went downhill and not even Chapter 11 could save it.

[Steve Simpkin]

A fun article about the early days of Wang Laboratories. It is really amazing what they were able to do with the technology that was available to them at the time.
http://www.oldcalculatormuseum.com/d-wangcustom.html

[jebem]

(07-09-2015 03:49 PM)Steve Simpkin Wrote:  A fun article about the early days of Wang Laboratories. It is really amazing what they were able to do with the technology that was available to them at the time.
http://www.oldcalculatormuseum.com/d-wangcustom.html

Thanks for bringing this information to our attention. That was an interesting reading.
From the Wang systems described in the document, I have actually worked with the Wang 2200 systems.
While there were a huge range of 2200 mini-computer models, most of them shared the same basic concepts:
- An operating system shell working in BASIC language (most of the models used ROM based OS plus optional ROM scientific libraries; other models,like the MVP, loaded the OS from a Hard Disk partition);
- Single or multi-user terminal systems;
- Storage I/O devices on magnetic tape (cassette) like the PCS series; reel tape was available as well;
- Storage I/O devices on magnetic hard drives (the CDC 20MByte using a combination of fixed and removable 10MB cartridges were very popular);
- Printers and Telecommunication peripherals.

[Gilles59]

(07-02-2015 06:03 PM)Don Shepherd Wrote:  I'm wondering if anyone here ever programmed in RPG. Of the three languages I studied in 1968 I remember RPG was my least favorite; assembler was my most favorite. You couldn't write an RPG program on just regular paper, it required about 4 or 5 specific types of coding forms as I recall. It was not a procedural programming language like FORTRAN and COBOL. I could see that it might be useful in certain simple file-processing applications, but looking back on all the programs I wrote in my career, I doubt that RPG would have been especially useful in any of them.

I think I consider RPG similar to HTML in that if you use them, you're not a "real" programmer.

Any thoughts?

Up... Oh Yes ! I programmed in RPG in my first job en 1985 on a IBM36. In France the langage was named"GAP". I never studied this at school but it was very easy to learn. I must say I hated this at the beginning, it looks for me like an incredibly archaic thing ;D. I continued with RPL on the IBM AS400 but the new versions of RPL on AS00 had not many things in common with the older versions. In the 1990's I remenber that I wroted some programs to generate automatically the RPG code because I was boried to write always the same kinds of program

[badaze]

I program in a language that generates RPG. In my 25 years carreer, I did only a few programs in RPG but I can read it. The last program I made,this year, creates dynamic screens and was not that simple to make mainly because there were no resources nor examples on the internet.

[Mark Hardman]

(12-24-2017 01:46 PM)badaze Wrote:  I program in a language that generates RPG. In my 25 years carreer, I did only a few programs in RPG but I can read it. The last program I made,this year, creates dynamic screens and was not that simple to make mainly because there were no resources nor examples on the internet.

Would the code generator in question be Synon/2E or COOL:Plex (now CA 2E and CA Plex)? Between 1987 and 1994 I worked with both of these software generators. Like you, I never touched a single line of native RPG code. At last count, I had generated over 10,000,000 lines of RPG code. I have a plaque from Computer Associates (CA) commemorating this "accomplishment."

Good times.

Mark Hardman