Tweet
jacobus wrote: It's an arcane, inflexible and very limited technology and it's programming interface (subfiles etc) is horrible... For example, one could build an even better data-entry screen because you could instruct the cursor to go to a specific field after pressing the tab. One can't do this with 5250. 5250 would probably be more flexible for you if you knew something about it. Try FLDCSRPRG. rd
-
-
It's the DEVELOPERS, stupid!
jacobus wrote: My point is this. If we are so in need of a GUI, then why don't we build one? Or better, port the X Windows System to the i5...Especially because our own future depends upon initiatives like this. Don't count on IBM to do it for you. We have to do it ourselves. I really believe this is the only strategie which is going to work. I agree with your overall point. However, from what I've read XWindows is included with PASE. I've never seen a post from anyone who talks about it, so I was quite surprised to see it's there. I have posted many times that Unix, for example, is not cited as an OS without a native GUI like OS/400 is due to Unix having XWindows, yet we also have XWindows. On the other hand, if AJAX can be chatty, XWindows is a chatterbox. And people really do have work to get done. Lots of people, with lots of work. 5250 is what gets it done. However, I believe that a more powerful desktop interface that integrates desktop apps, and an enhanced iseries interface that incorporates desktop paradigms, such as a spreadsheet subfile, can take iseries integration to the next level for the user. In that, I agree with your thoughts completely. rd -
It's the DEVELOPERS, stupid!
Thats the point, the whole discussion about OS/400 not having a native GUI is moot. Even without XWindows on OS/400 it's moot. XWindows is also "old technology". Modern apps consist of a thin client presenting the GUI which communicates with the server, being Unix, Windows or OS/400. Because OS/400 supports the small-bandwidth 5250 protocol and all RPG programmers are used programming for it, it's still widely used. It's no option to replace 5250 with XWindows because this is not scalable. Although it certainly would be a good idea to have the ability to extend existing green-screen programs with XWindows functionality. For example to pop-up an entry-window or whatever. Anyway, people tend to compare OS/400 with Windows which has an "intergrated" GUI, adding much to the instability of windows. OS/400 and other Unixes have their GUI (XWindows) separated. The reason OS/400 programs present green-screens and therefore add to the confusion that the AS/400 is "old" and must be replaced is not something you can blame the managers but the RPG programmers who seem to simply refuse to learn anything else than RPG and DDS. JohnComment
-
It's the DEVELOPERS, stupid!
5250 is the tip of an iceberg called a session. What you advocate learning and using is that which is not a session, stateless web pages which much somehow be strung together to simulate an interactive session. It is done, with Websphere Application Server (hence the need for a web application server), with Java session save and restores (write the "object" to disk and reload the session each page) and similar support from PHP. No one has begun to make a coherent case for the programming and execution hoops that must be traversed to fake a session out of web pages so you can display a screen in a web browser instead of a 5250 emulator (which also are available in web browsers as Java or ActiveX). In fact, none of you can even coherently describe the programming and execution hoops that must be traversed, much less describe the traversal. Should a coherent argument ever be made, then it could be considered. rdComment
-
It's the DEVELOPERS, stupid!
I was not talking about session-less connections. There are several options to implement an interface. Using a browser, with or without sessions, or 5250 (with sessions), X Windows (with sessions), or client/server with or without sessions. I was merely talking about the fact the most RPG programmers dont know about these choices.Comment
-
It's the DEVELOPERS, stupid!
Choices we have. Coherent reasoning about them we don't. I will add this to my previous comments. Bob writes today about ILE RPG as a language and the underutilization of its features. I agree with him. He gives a good example of generating a pdf file from some invoice data to display to a customer in a browser. This is a reason to use a browser. Again in the hype and hysteria of web browser usage not even a passing nod is given to the different design and usage factors of interfaces for the consumer public, business partners, and internal business users. It is the internal business user to which the majority of our programming goes, and it is for them that 5250 is a superior solution. The fundamental function that Bob describes should have already been done by a web advocate as an open source solution demonstration. There shouldn't even be a question of doing this from the iseries, it should be cut and dried by now for best practices. As I wrote in an earlier post in another thread ("I'm not happy anyway with available option tradeoffs of language, server, performance and cost for personal development work I want to do."), I will instead be spending time on /free RPG web development projects which I will open source. Some of what I will be basing my work on is open source anyway. I spend my personal time on the web. I know when the web is needed, and I know when 5250 is needed. I know when SQL is appropriate, and I know when native IO is needed. Others don't have the choice. We do. Let's actually think about the solutions we're providing, and provide better solutions on the iseries than anybody else, instead of trying to grab hold of some magic web ring that doesn't exist. rdComment
-
It's the DEVELOPERS, stupid!
1. FLDCSRPRG only works with IBM's Client Access which sucks for throughput. FLDCSRPRG does not work for most third-party Screen emulators. Rochester still says that major through- put improvements have not yet happened for IBM's Client- Access. 2. VB.NET forms and ASP.NET 2.0 skate circles around subfiles. 3. WDSc 7.0 IS lightyears behind VS2005. 4. RPG to the Web involves 10 times more coding than VB.NET/ ASP.NET 2.0 or 3.0 to the web. PHP might offer hope. 5. AJAX is Over the Top Chatty. Help and Forums are sometimes insulting, elitist, clickish and many times withholding. 6. PHP is the new white horse for IBM. We'll see about PHP. I am WOW! Just so impressed that IBM had 5,000 downloads. 7. The below choices are balky and "Tack-on" type solutions compared to the MS-based Web Solutions: "I was not talking about session-less connections. There are several options to implement an interface. Using a browser, with or without sessions, or 5250 (with sessions), X Windows (with sessions), or client/server with or without sessions. I was merely talking about the fact the most RPG programmers dont know about these choices." RPG programmers have been lazy and are, for the most part, waiting out the clock. Their conservatism is one of the BIG reasons we are in the End-Game for the AS/400. I can be programming in the Green-Screen world and even pretend I am still on one of those demo System 38's I first signed onto in 1979. 8. JAVA sucks and is now a "Has-Been" in the general community. 9. the i5xx is like an archeological site with many layers and few remaining inhabitants. COMMON is the tour-guide to this ancient site. I can just imagine the back-flips that IBM is going through to keep this Rube-Goldberg contraption alive. This is a major Rochester-Accomplishment. 10 I really liked your answers Jacobus. Your view in the big and medium picture is right on. Ralph is a blow-heart who barrages people with facts but doesn't have the right awnwer.Comment
-
It's the DEVELOPERS, stupid!
Ralph: With your comments everywhere: Are you actually working for your employer? I don't post from work. I post from home on my own time. Client Access is slow? What a troll. What, are you playing hooky from naptime again? rdComment
-
It's the DEVELOPERS, stupid!
Oh, I see another modernization miracle again. From none other than IBM, who of course wouldn't be telling us we should be "modernizing" our business applications if they had one "modernization" failure after another. Oh wait. I guess they would. I'll give the drive by summary. Should be familiar to anyone who has seen all the other "modernization" failures I've posted. http://www.computerworld.com/action/...cleId=9019687+ &intsrc=hm_list Texas again, sheesh, they must have a terrific relationship with IBM. TEAM Texas Election Administration Management system is supposed to allow election workers state wide to check voter registrations against driver's license records and Social Security numbers, results sent back. That's simple enough, just the thing an iseries excels in with 5250 sessions and native IO RPG programs. Multi-billion dollar companies run their business doing thousands of things like this every day. But IBM comes in and for $14 million ($14 million? Oh, it must be "modern") writes a Java-based registration system running against Oracle on Sun Solaris. So who's the J2EE culprit? Answer is... IBM's Websphere. No surprise there. Here's the operative statement, so to speak: "Since January, county elections officials have flooded the state with complaints about the system, particularly its poor performance. The complaints have increased markedly since local elections were held throughout the state on May 12." "Technology should improve our productivity, not create unnecessary additional burdens," said Cheryl Johnson, the Galveston County tax assessor collector, who has issued a number of formal complaints about TEAM's performance and its problems in retrieving historical records. "TEAM has adversely impacted productivity in the Galveston County Voter Registrar's Office as well as negatively impacted many of our voters," she said. The IBM spokesman acknowledged that "as with any new application deployment, work was required to optimize the solution based on actual usage patterns." Nevertheless, four county officials this week contended that the TEAM system still has serious flaws. "We're frustrated," said Candy Arth, president of the Tax Assessor-Collector Association of Texas and the tax assessor-collector for Washington County. "Everyone has been." Arth said members of her staff in Washington County say that the process of entering names into the TEAM system can take minutes when it should only take "nanoseconds." Officials said the TEAM system promised to allow workers to enter 40 voter names into the system at a time, but the sessions often time out, forcing the process to begin anew. Therefore, she said, workers have to limit the number of names entered, costing the county thousands of dollars in overtime to complete the work. end quote There's so many "modern" features here to report, I have to draw a line somewhere, but let me finish with the tax assessor for the Houston area: "With its current performance, the system is unusable in a major county," he said. In some cases, he said, Harris County has had to resubmit voter applications three or four times before they are validated by the TEAM system, Bettencourt said. end quote Yeah, but don't those timed out web pages look pretty? Well, I'm off to try to forget about stuff that works like 5250 and go play with pretty web pages. They're "modern". rdComment
-
It's the DEVELOPERS, stupid!
"It is the internal business user to which the majority of our programming goes, and it is for them that 5250 is a superior solution." Besides OS/400's ubiqitous support for it and that each RPG programmer has the knowledge to use it, what really is the added value of 5250 if compared with a well-designed web or GUI interface (the latter being a thin client). I don't see much added value in being constrained within a 25x80 character screen. Constrained that is... Making a GUI for the same task requires a keyboard- and character based GUI to handle data-entry tasks. But instead of being constrained one has the opportunity to expand beyond that if needed. Such as other fonts, more characters/lines, an user-controlled (event-driven) interface instead of a program-driven (hierarchical) etc. While 5250 is quite simple, it's RPG programming interface (subfiles and 2000 DDS keywords where most of them are obsolete etc) is not. Using Swing in Java is conceptually simpler that using DDS and subfiles to do simple things like putting some labels, entr-fields and lists (subfiles) onto a character screen. JohnComment
-
It's the DEVELOPERS, stupid!
Pardon me....? Java was and is very succesful, on servers and on (mobile) devices. But now after 12 years its proven technology, and therefore it's less interesting. But mind you, Java rules.... Check out JavaFX for example. Keep informed.Comment
-
It's the DEVELOPERS, stupid!
5250 interactive session vs web page stateless re: "event driven", 5250 cursor sensitive and other feedback info is either just as much or more "event driven" subfiles are much more powerful than drop down lists, and it would require AJAX to have the fast scrolling within a subfile window that 5250 does, except AJAX wouldn't be fast for a non-trivial number of users 80x25 text sounds limited, but 5250 screens are much more jam packed with data and functions than web pages in screen area. Programming to DDS screen object is light years more advanced than with HTML 5250 interactive sessions is high speed business processing. I have never seen a user willingly select a web page rearchitecting of functionality they have with 5250. RPG programs with 5250 and native IO have been performing the kind of functionality that IBM and other people with web products to sell have been failing to simulate, large, unalderated multi tens and hundreds of millions of dollars failures. The abused users even plead for their working green screen interface to be given back to them. Even when working decently, web page design has them going through tabs and multiple pages to do same functionality. "More integration with web pages" than 5250 is a completely false, unfounded, and baseless claim, but it's always made by web types and their web vendors. Nevertheless, I agree, advanced architecting and handling of web pages for multiple purposes over the internet is essential to have, and there shouldn't be the slightest hesitation to use iseries for it. rdComment
-
It's the DEVELOPERS, stupid!
Dude, that's SOOOOOO "Legacy" :-) Actually, sometimes I think a lot of you are off the charts in la-la land with this pointless arguing over technology. Don't you know that, regardless of what technology you are using, if you are employed in the United States, your position is at risk?! Some seven or so years ago, "getting certified", was all the rage. The PC boys were beating their chests as they hung those gold-mine MS certs on their walls. "I can write my own ticket with this", they proclaimed. Today all the network gurus are GONE. They are contracted on an as-needed basis. The glorious "certification wall" is gone. Yet we are still here. We are all doing as much as we can trying to update our skills, for both our benefit and the company's. In the end though, is it really going to make me less vulnerable to getting whacked? I don't know---maybe, maybe not. Can you say Wal*Mart greeter? I reckon one could get into real estate huh? That's really booming right now :-|Comment
-
It's the DEVELOPERS, stupid!
I don't know, even IBM's LEAN layoffs is said not to be affecting software people that much. All speculation though. I think people who can produce good solutions as we discuss doing will for the most part still be in demand, as long as there still is a demand, barring another major downturn. rdComment
-
It's the DEVELOPERS, stupid!
There's a big need for truck drivers right now.Comment
Comment