Unconfigured Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

It's Time to Get out of the Stone Age

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • It's Time to Get out of the Stone Age

    This may be a first, but there's not a single point in Bob's article that I would disagree with. The overal negative response is surprising, and telling. Here's my own take on the issues brought up in this discussion: 1) First, for IT professionals, you know how fast things change in this profession. The need to keep up with current trends is perhaps greater for computer people than in any other line of work. And so, continuing education is an absolute necessity. If your employer won't give you the resources to keep up to date, then you must take it on yourself, no matter what it takes. As Bob points out, learn something, anything. It doesn't really matter. Demonstrating to a potential employer that you can learn new things is just as important as what you already know. (If you don't know where to start, I recommend learning Python. It's an easy language to learn, fun to use, and includes a robust, powerful, and extensive class library. The language and class library give you a gateway to other technologies, such as CGI, XML, and OOP.) 2) For IT managers, you know how fast things change in this field. You have a responsibility to your company to understand, not just current day to day operations of your shop, but also to anticipate future needs and make your bosses aware of the choices. Even if the suits don't see a need for modernization, they may well someday. If or when that day comes, you know that the last thing you want is for the suits to bring in some outside consultants making the recommendations. You know your company's IT needs best. You need to be the one making the choices. And so you need to be sure that your bosses know that you can do the job when the time comes. 3) Regarding green screen versus GUI, it is always important to know who the audience is for your user interface designs. The needs of the heads-down data entry clerk are different than the needs of the casual user. If one requirement is keyboard entry, then you may need to forego fancy widgets like radio buttons and check boxes, and have all data entry through text fields. Cheers! Hans

  • #2
    It's Time to Get out of the Stone Age

    knowing where to start. Being brought up on IBM/Midrange systems, my mindset was that the "system" resides on one machine, that is, the programs, data, interfaces, etc.; all created using IBM's toolset, with any unanswered questions residing in their manuals. IBM made it too easy on us you know. Naturally that isn't so anymore, but there are so many options, each having a sizeable learning curve. I don't mind investing the effort at all, and I am afraid of none of it, but I certainly would hate to spend the effort going down a path that led to nowhere. It's fairly easy (I think) to pick out those of you having the most knowledge & wisdom, but even so, each of your recommendations are different, yet all seem to work for you. So which direction does one take? Maybe I should just google up the nearest museum and reserve a spot now...before the other fossils show up. These places can only afford to have one brachiosaurus on display you know.

    Comment


    • #3
      It's Time to Get out of the Stone Age

      The only people who suddenly face the learning curve, and sank, are the people who were not in the process of continous learning. Yes, a lot of what you learn, will go down the drain, yet the ones you will be able to use, will be the life saver. I am among the people who were not in this continous learning process. Simply because I was considered a Guru. As if the Gurus are not only suppose to know what is out there, but also know what will happen in future. We sink because we depend for life on the one time learning we had. I went back to school and came back to pick and am still behind in many many area.

      Comment


      • #4
        It's Time to Get out of the Stone Age

        Os wrote: Naturally that isn't so anymore, but there are so many options, each having a sizeable learning curve. You've hit it right on the head, Os. iseries development and interface options are totally fragmented and all over the map. Gurus say just do something, anything, as long as a web page is involved. The advice is often couched in implicit terms that worst case this is also about preparing for a job on something else that will survive. I rarely see coherent business requirements for any of this. It's all about the programmer and can they have their cake and eat it too. I have long advocated IBM making some reasonable enhancements to our development environment to use 5250 and all of its capabilities as the base for interface independent program development. Our previous synergy that made the AS/400 and its development community so valuable would return when there is once again a built in programming standard as we had with 5250, only now with output configured for the connection type. This is saying in yet another way that IBM must supply standard GUI interface output from our RPG programs for the iseries and we iseries programmers to survive. The third party options are welcome and will always be there, but divided we fall. rd

        Comment


        • #5
          It's Time to Get out of the Stone Age

          What kind of place has a person who can read RPG II and RPG III but can't read RPG IV? And if such a person exists, why would anyone care? rd

          Comment


          • #6
            It's Time to Get out of the Stone Age

            Most business people at companies run on the AS/400 type into fields on the screen. I would call very little of it data entry. It runs the gamut of typing in things to look up to entering or updating data, but it's not the "heads down" data entry of keypunch days in all but the rarest of cases. I cannot recall seeing a business person choosing a web page version of a 5250 screen in any of the situations I was involved in when we bought web overlays for ERP packages, for reasons given elsewhere about data entry. Now, that is comparing apples to apples. But if you have a 5250 screen with no prompt for an entry to look up, and a newly programmed web page that does, then of course the business person is going to choose that which helps them more. But same capabilities, apples to apples, and quite frankly that should include 5250 mouse, menuing, and windowing programming, then why would people want to take their hands off the keyboard? They don't. I don't. These web page gurus running around in here probably don't either if they had to admit it. So when the reference to people preferring web pages is made, what the people think of is better data integration and smarter screens to work with, not the goofy logo webheads put in the corner to prove it's GUI you're looking at. To the people that should matter, our business customers, that's just real estate lost for info they wanted to see now, on the screen in front of them, not where they have to hunt for it with a mouse. 5250 screens don't have the goofy logos, but they have everything a business person wants... if you put it there. rd

            Comment


            • #7
              It's Time to Get out of the Stone Age

              But a dinosaur that makes a lot of sense. rd

              Comment


              • #8
                It's Time to Get out of the Stone Age

                I saw a demo for VAI System 2000 in late 90's that, with source code and optional web page overlays, looked like everything an AS/400 SMB could want. I just googled it and link is: http://www.advgroup.com/partners/vormittag.html rd

                Comment


                • #9
                  It's Time to Get out of the Stone Age

                  Bob makes many good points, especially when it comes to IT managers not embracing the new. I also agree with Jonathan. It's also important for the lovers of everything new to learn to embrace some of the old when it make sense to. There has to be a happy medium. In our shop most of the data entry is done in remote sales and/or dispatching offices through DSL and a private frame-relay network. To convert everything to GUI would require upgrading the communications, which would cost a lot of unnecessary $$$. Simply put there's nothing wrong with 5250. (Not that we don't do GUI and some of it has been written for a few internal users, but the bulk is for the B2B website where we need to output HTML for our customers and vendors who are coming into our network on a T1. Some is written in Java, but the bulk is written in Net.Data. In the future we're going to write a lot more stuff using PHP.) Our rule is when an application needs changing and it's CLP with RPG-II or III then we convert it to CLLE and RPG-IV. All new applications are required to be written using ILE and free-format RPG.

                  Comment


                  • #10
                    It's Time to Get out of the Stone Age

                    Ralph, I agree with you, and would like to go one step further. I am of the opinion, that IBM has left work on the 5250 interface incomplete. Years ago, I programmed using GDDM. A 5250 graphic interface that goes all the way back to the System/38. I was one of the first to use 5250 drop downs, choice fields, radio buttons, check boxes, and menu bars when they were introduced as PTFs to V2R3. I know that by using certain APIs, you can go beyond the six color spectrum. I know that you can program a 27x132 screen, although I've seen this dimension only on extremely rare occasions in production environments. I also know what is missing:
                      [*]The ability to select fonts within DDS.[*]The ability to mix fonts within DDS.[*]The ability to mix font sizes within DDS.[*]The ability to retrieve a stored graphic and display it (such as a BLOB field) within DDS.[*]The ability to have all of the above immediately translatable to a web page using HATS, Webfacing, or other product without having to produce additional code.[/list]That's just for starters. I believe that the above list is possible within the confines of 5250, and could be implemented if IBM would only green light such a project. The enhanced capabilities would assist in protecting the existing business investment. You could even use goofy logos and banners - - - and make them mouse capable. Dave

                    Comment


                    • #11
                      It's Time to Get out of the Stone Age

                      ... and you in turn have hit right on the head too. We are told again and again about going to GUI. I have seen the result of third party GUI like Seagull. User simply called it a "Mickey Mouse GUI aimed to fool user into thinking it is a real windows based interface". Furthermore, buying a third party software requires "budgeting" and you know how the process of "cost justification". It was difficult getting a 1G RAM to boost my desktop so that I may get WSDC up and running. However I will disagree with you that going for technology is about all about the programmer. I see a bigger picture here. It is about the AS/400 shop, then it is about the reducing number of AS/400 shops. This ship is sinking. It is easier to convince management to retain AS/400, let the number crunching on it, and put windows based face on it. In this shop I saved an application by putting database on AS/400 and write RPG procedures to process them via procedure calls from VB. They like the speed of number crunching and the graphics of VB.

                      Comment


                      • #12
                        It's Time to Get out of the Stone Age

                        "What kind of place has a person who can read RPG II and RPG III but can't read RPG IV?". Good question. I wanna meet such person It is more of job "security" for the I.T. Manager who wrote the code, than the "obscruity" of the new style of coding. On a serious note, in a Florida based oxygen company the I.T. manager was afraid of the unknown in conversion. Despite assurances by everyone that nothing can go wrong with CVTRPGSRC, things went wrong unfortunately. I converted a program while I was making a minor change. The QA tested it OK. Unfortunately another RPGIII sometimes called it, and the CALL was followed by a FREE. When the call was made, it ran fine but at the end of execution, the calling program broke after the call! This one problem did it, and I had to do a lot of explaination to the I.T. manager as to why did I convert program from RPGIII to RPGIV when it was not need. "Did I not tell you to not fix what was not broken". Fortunately she lost the job before me and the new I.T. Manager gave a free hand for conversion. Strangely, that was the only hiccup!!!

                        Comment


                        • #13
                          It's Time to Get out of the Stone Age

                          Ralph wrote: "This is saying in yet another way that IBM must supply standard GUI interface output from our RPG programs for the iseries and we iseries programmers to survive." I disagree. Adding support for a specific user interface methodology will only result in the same kinds of problems. Currently, there are a lot of RPG programs written to the DDS/5250 model of UI. For these apps, conversion to a web model is very difficult. That is, DDS UI is driven by the program, but the web UI is transaction driven. If RPG apps are now written to a different UI model, what happens on the next paradigm shift? What's needed is a different mindset. Application developers using RPG need to understand MVC design. That is, separation of application into "model", "view", and "controller". RPG has a place in the "model", along with the database. The "controller" handles user input, such as GET and POST requests from a browser, or reads from a 5250 device. The "view" handles user output, such as HTML or XML or 5250. If different forms of input and output are needed, the requirement can be met without touching the business logic. In this model, you can choose the appropriate technology for each part. For example, Java servlets for the controller, JSP for the view, and RPG/SQL for the model. Today, RPG already has pretty much everything it needs to participate in this world. If you were a masochist, you could program the controller and view using RPG too. But again, you already have all the pieces needed for that as well. Cheers! Hans

                          Comment


                          • #14
                            It's Time to Get out of the Stone Age

                            As still another dinosaur in a one man shop for over twenty years, it's difficult to figure out what to do next. Especially when you must keep everything running smoothly, ajust to changing business conditions and all the other things involved in a small business (and live a home life as well). I dabble in RPG IV. But when I needed to debug the first one I wrote, imagine my surprise that the intuitive, easy to use STRISDB command does not work. Now I must learn the STRDBG command, which sure seems like two steps backward for the one step forward. I would love it if IBM enhanced DDS with font keyword and otherwise enhanced the 5250 stream. But I don't believe they ever will. So what should I consider instead? Webfacing is kinda nice, but (fortunately for me) in our small company, it's not a big consideration. How quickly it takes me to create programs is the bigger consideration. No matter what I want to learn, I'm going to take big hit on my productivity. Should it be RPG IV subfiles? CGI-DEV2? What about third party stuff like Lansa, PlanetJ WOW, RPGSP? Just deciding what to pursue is a time consuming business. Not to mention I will have to sell it to the boss if I want to buy a third party product. When I evaluated lansa, I found that it does not lend itself to being called from traditional AS400 menus, or calling it from CL. So that is another new method of operation that must be evaluated to decide how the new stuff will work in conjuntion with the old stuff. My users will need to understand the some things are run from a menu, some from a browser, some from yet another interface. I'm sure they could probably get used to it, but that is yet another consideration for me to think about and worry about. What about WEb hosting? I want to learn more about HTML, but I sure don't want to write raw code. I got Dreamweaver, and a high school intern to show me how to use it. But after initial set up, other tasks take priority, it's been several weeks since I last used Dreamweaver and I am forgetting how it works. Even before, I have difficulty understanding how to use the internal web site for real business usage instead of just displaying staic pages. I would like to archive reports in PDF format so users can review them. While I have IBM's clumsy Infoprint (don't get me started on that) I need to create a file name and link in dreamweaver for every report I want to archive. That's a lot more cumbersome than looking at an outq. As is typical, I start playing around with something new, and then a business change comes up. We revised how we calculate our cost of goods this year. I didn't have time to try it in something new, I had to get it working right away. I used the tools I had. ---Dale

                            Comment


                            • #15
                              It's Time to Get out of the Stone Age

                              What's a dinosaur to do? Many species of dinosaurs became extinct. However, many scientists believe that one species evolved into what are now birds. So it is possible for a dinosaur to evolve and rise to great heights! What to do? The first step is to decide what your own career goals are, and separate them from your employers needs. Second, realize that you alone can't solve your company's problems. That's you boss's problem, not yours. You can only do so much. In terms of getting up to speed on current technologies, don't look at your company's specific needs, since that can only bog you down in endless alternatives. As you've seen, one person alone can't both maintain a shop and modernize it. IMO, you need to step back and learn the basics of things like OOP and CGI and UI design and HTML and so on. As I mentioned before, learning Python is a great way to start. You probably won't be able use Python in your current job, but the capabilities of the language and its included class library are an easy way to learn much of what you need. Once you understand the basic concepts, you can apply them in other languages like C, Java, and even RPG. Good luck! Cheers! Hans

                              Comment

                              Working...
                              X