Unconfigured Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

IMHO: What Will It Take to Turn the System i Around?

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

  • #76
    IMHO: What Will It Take to Turn the System i Around?

    Gary said: No, I didn't read what you got out of that other thread...not at all. Really? You didn't see this: Kjack wrote: "We were essentionally rewriting our green screen validations in the VB rules layer, and using regular RPG chains ands reads for the database access layer using ASNA's DataGate technology. We found that the result of going after the data directly using VRPG was ssllooooowwww. We found that limiting the bandwidth to just what you are presenting to the user by keeping the validations and database access on the iSeries was way faster. We did this by scrapping the VB rules layer and VRPG database access altogether and just calling programs on our iSeries." And you accuse ME of not reading for the purposes of understanding! Joe

    Comment


    • #77
      IMHO: What Will It Take to Turn the System i Around?

      Well sure enough. The link I received didn't start at the top and kjack's message you quote was out of my view. We searched everywhere below that point and couldn't find the messages you found. Unless the client's access to the iSeries is through communication lines, we don't see performance differences between the scenarios of accessing data directly from VRPG or through calling an iSeries pgm to do it for the VRPG app.

      Comment


      • #78
        IMHO: What Will It Take to Turn the System i Around?

        Hi boomer, I certainly agree with you and Timothy. What I was saying was that TMP was trying to salvage a typical IBM marketing plan that confirms yet again what I've been saying: IBM has niched the iseries~AS/400 as something to force Websphere on 20 people or so. Anything related to iseries is secondary to trying to force sales of Websphere. That's what they've staked their business on, that's all they care about. Sure, they'd like to sell iseries, but they run ads of a little box spinning around on a merry go round with some deliriously happy small business tech guy spinning around with it, "Geeeee this is easy!!!" That's the system i~i5~iseries~AS/400 that IBM has carved out to not intrude on their real computers, the AIX and Z/OS boxes. So, no, there's no hope in marketing orders to sell "What goes up, must come down, spinning wheel, going round and round" for 20 people. Here's the reason. It's not that simple. It's an OS that runs multi-billion dollar companies. We are career people on it learning new things every day. I for example don't have time to learn Linux enough to run a business with it, and I've installed it a couple of times. Sure you can pick up a few things a little at a time, but you need to be a pro to run a business with it. The same thing is true for Unix and OS/400. For that matter, you're in deep if you put your company's Windows servers out there exposed to the internet unless you're a pro, and often even that's not enough. But Windows admins are a dime a dozen, everybody and their mother has a cert. A small business with 20 people isn't going to be able to justify ramping up on staffing for OS/400. But here's where we get to the real disconnect. IBM isn't even trying to sell i5/OS~OS/400 to 20 people. They are selling system i to 20 people, which to them is like we don't care, operating systems, shmoperating systems, who cares, here buy this computer that makes it all easy, one backup command. That's how they are trying to sell the iseries. They've never taken TMP's advice before, they wouldn't be in this boat if they had, but if they ever get desperate enough to change their ways before they abandon ship, they should go with user pricing along the lines of his suggestion. Scaled as high as the iseries goes, with all hardware on board maxed out, enough memory to run Websphere, and a solid IBM integrated infrastructure built in for those seats, with only advanced features like J2EE at extra cost. Websphere with Notes Domino and a native iseries Eclipse desktop interface with their new Rich Client Markup Language for us to program to for starters. Think of the FBI Case File system that's a multi-hundred million dollar disaster. Every iseries should come out of the box ready to program and run that kind of software. That's what's every business wants, and a standard infrastructure with IBM middleware for a seat or two of every product would provide similar but more capabilities that enabled the success we had developing and selling software systems for the AS/400. It should be a win win for everyone. We would have a standard interface to program to and a standard case file level infrastructure to call upon for our software. IBM can showcase an integrated knowledge worker environment computer that can run the software on any mix of the major operating systems. Seats would be added up from all communication sessions, with concurrent CGI jobs counting web activity. This is nothing that couldn't be done for Christmas. It's just marketing. The way it should be done. rd

        Comment


        • #79
          IMHO: What Will It Take to Turn the System i Around?

          My 2 cents worth for Websphere: MS gives away a development environment so that people can learn it easily, and it runs on their PC. IBM makes you purchase an AS400 in order to get Websphere Development Client. I have wanted to learn Websphere for the past 5 years, but had no way of doing so, as an independant contractor, or as an employee. I now have a full-time job, but Websphere is not even spoken about here for development of green screen applications OR web based ones. I could steal the Websphere disks here from my company and load them on my PC at home to try to learn it, but would have no AS400 to connect to. IBM should produce a Websphere Learning edition that is self-contained and runs on a PC, and GIVE IT AWAY. I AM learning MS VB/C# on my own, as I see the writing on the wall: My days as an AS400 programmer ar numbered. I was unemployed for nearly 2 years because I could not find an AS400 job within a 75 mile radius of where I live, and could not move. However there were literally hundreds of VB/C# jobs posted on the job boards within that radius. So, do I learn Websphere for the AS400, or VB/C# for the Wintel platform? My answer is to learn VB/C#, as Websphere will die the same death as the AS400. IBM just is not marketing the machine enough, not making their application development platform of choice available for people to learn it easily, and the AS400 is considered just plain "Legacy" by the people with the purse strings. Please don't get me wrong. I love the AS400, and I've been developing RPG applications for close to 20 years. It's the best box out there for reliability, stability and businesses to run on. But the OS/2 treatment that IBM is giving its marketing (or lack thereof) really makes me think that IBM just wishes that the machine would die so that they could forget it and move on to Linux/Unix/AIX. I'm learning VB/C# so that I will be able to work when I lose my job, when the AS400 finally gets its plug pulled for being a legacy box that has outlived its time.

          Comment


          • #80
            IMHO: What Will It Take to Turn the System i Around?

            Interestingly, I've never had the feedback that MS is an "Open Systems". It was/is well understood that it is proprietary. It's market dominance makes it the defacto solution - but no more than the ride IBM had right through to the s/36 boxes in most markets. I was there. We sold some crumby performing boxes with very crumby programs to a lot of people. I did love that little O/S though. I do agree my "superior" tag was a throwoff. I'm a coder and look to be right in the mainstream of the current flow on thinking. The reference was to code production as it is dear to my heart. Ultimately it is dear to the survival of the Series I. I want the best way to generate code in the shortest amount of time. There was another post on the quality of the database structure in SQL Server environments. With no disrespect to the author, this is not a SQL Server problem. The crap code and database structures I used to support on the s/36 and AS/400 are nothing to brag about. The "Superior" product are definitely MS's product development tools. I'm sorry but the infinite number of Java Frameworks out there is just clutter. Distracting at best. The IDEs - whilst competent - are no where near VS for productivity. I actually have wished for SEU screens instead of some of the Java IDEs I worked on. Back on topic, build a better development tool for your proprietary box and you'll get the bees in. JAVA is open to any box (including windows) so it won't encourage the bees to the Series I like VS (.NET languages) will to MS platforms. This is not to say there wouldn't be a problem attracting ISVs to the platform given it's small market footprint versus, say, windows. Historically, I wrote one of the first GUI apps to be released targettng a particular vertical in Aussie hospitals. It was developed on OS2 but was a Win31 app (as the market demanded). At the time a huge chunk of the Australasian hospitals had s/36s and AS/400s. Now there are none. Their replacement is THE "5 seat solution" you referred to. There is 1 vendor left peddling a Unix solution with Informix/Oracle database backend and they are struggling with the "legacy" tag even though their current release is Web based. The rest are Windows served solutions with 1 of the smaller ones having the Cache database rather than a SQL Server backend I also work in the Tourism industry. Every Hotel had a s/36 or AS/400 in the early 90s. There used to be 30+ boxes in my city alone. Now there is one left in this industry that I know of as I am it's principle support - there is no other experience in the city for the box. All the IBMs have been replaced by the "5 seat solution". As for ERP, you may still have a small vertical remaining. But don't kid yourself that it is safe. You'll need to watch SAP (Windows/Linux) and software vendors for the commodity Linux boxes. Ron

            Comment


            • #81
              IMHO: What Will It Take to Turn the System i Around?

              If you're saying five seats is what's running the business, then I agree, that's not enough to justify moving beyond the familiar. As for all the stuff about Java, there are only two main IDE's, Netbeans and Eclipse, both open source. There are only two GUI interfaces, Swing and SWT, both open source, and there's a bridge for them to work together. Frameworks? Yeah, being open source, there's an active evolution of frameworks, it's not like you just do whatever Microsoft tells you to do. As far as that goes, .NET is a new framework over VB 6 and I saw lots of very upset open letters and gathering of signatures to try to keep their VB 6 platform. Microsoft said nyet. You don't have anybody telling you that your framework is no longer supported with open source. So yeah, they're still around if you want to use them. rd

              Comment


              • #82
                IMHO: What Will It Take to Turn the System i Around?

                My 2 cents on the development language issues on the iBox vs. MS Vx et al. I really think it's actually a matter of accessibility. MS Vx is ok but has it's limitations and issues as well and Unix dev has its' own particulars too. However, for a someone in college or just out, it is virtually free to download, install and run a IDE on their PC. You can't really do that with an iSeries VAx product; you need an iBox to connect to. If it weren't for that MAJOR limitation, we probably wouldn't really be having this discussion. As far as I can see, the only way currently around this problem, is for someone to sell either a iBox VM that works on an Intel 64 bit processor(s) or sell an iBox so inexpensive and so open that young developers feel a need to get one and do stuff on them. So it has to be a box that does a lot of things the current machine is unable to do either due to physically unable to (no IOPs for different hardware) or no software exists for those "new" functions. Does anyone have any other scenarios?

                Comment


                • #83
                  IMHO: What Will It Take to Turn the System i Around?

                  "If you're saying five seats is what's running the business, then I agree, that's not enough to justify moving beyond the familiar." Sorry - I was quoting from Joe Pluto's comments where the MS market dominance consisted of anything less than 5 seats. My post pointed out that there was an enormous scope of enterprises running MS with thousands of connections - not 5 connections or less. The examples given were in my sphere of contact only - how many other verticals have gone to MS? Recently we proposed a software solution in a bid for 2 large state health department clinical contracts. Both stipulated SQL Server as the preferred platform. Oracle was a 3rd choice of database for one of the tenders. DB2 obviously wasn't mentioned on either contract response. This is just the reality out there. Ron

                  Comment


                  • #84
                    IMHO: What Will It Take to Turn the System i Around?

                    SQL Server preferred by who, and why? The answer is the IT people are SQL Server people, which is my point. And if they can engineer a cost effective product solution, then more power to all involved. I spent a few hours last night researching a Texas HHS disaster, TIERS, and a preceding one in Colorado, which amusingly was contrasted with the nascent Texas one at the time with all the reasons the Texas nearly $1 billion software project would be so successful while they scoffed at Colorado's huge failure. But the plug was just pulled on TIERS at half a billion, with $3 million a month still being paid for life support. From my research, it was Oracle J2EE developed by a consortium of consulting companies. I looked at the web pages of the new system. Very simple screens, much simpler than green screen. And what do the Texans who have to use this failed new system say? "State workers swear by SAVERR, an (IBM mainframe DOS based) application that was built in 1985 and only requires a green monochrome monitor to do the work." Just as with the multi-hundred million dollar failed FBI Case system, an IBM green screen system is still chugging away, still doing the job, and no amount of hundreds of millions of dollars of Java and Oracle databases can replace them. Knee jerks always blame the government, but for this TIERS disaster the whole thing was attempted to be privatized and the government agency and employees eliminated, so the operational disaster is all private. The government was hated enough by these types that they aren't there to blame for this one, for once. Instead the plug was pulled, government layoff notices rescinded, and government employees asked not to leave with $1800 pleas for forgiveness. And all this talk about how "new technology" makes everything so easy to change and "legacy technology" is oh so hard to change? "In April, Health and Human Services Commissioner Albert Hawkins indefinitely suspended plans for a statewide rollout of the call centers and asked 1,000 state employees scheduled for layoffs to stay on the job. The State Auditor's Office said that inspectors have been unable to pursue criminal cases for the past two years because of insufficient data coming from the (TIERS) call centers. It stopped looking for fraud last April because information was not "readily accessible," the report said. "Furthermore, TIERS lacks interfaces to the automated systems at the U.S. Internal Revenue Service, U.S. Social Security Administration, and the Texas Workforce Commission" that are necessary to identify potential fraud, the audit added." These "interface" issues are promised to be resolved by September 2007. Of course this two year failed rollout was supposed to be six months, and how many promises like that do you think have been given for the last two years? Mainframe green screens work (and I count the AS/400 as a mainframe), and I haven't seen anything replace them for any amount of money and web pages. And if Microsoft knew how to do it, they wouldn't have put their Project Green ERP in suspended animation. rd

                    Comment


                    • #85
                      IMHO: What Will It Take to Turn the System i Around?

                      Well, some clarifications first for the GP post. Developing in RPG with "Websphere" WDSc is not really "learning" anything, it's just a screen that you need instructions to use, just like you were given instructions to use the SEU screen. Of course pictures are worth a thousand words, and the book(s) on WDSc containing those instructions and pictures are as helpful as with anything else. I used WDSc exclusively for much of last year for a project, but just to contrast with SEU. I still prefer SEU. However, WDSc is much preferable to work with a project with several source files, as in any desktop GUI editor product with a tab per source file. Now, as to "learning" it. It is an Eclipse IDE. Using Eclipse to develop with any language will have the same basics involved. A plus would be learning the language involved in the process, such as Java, but it's just another IDE for a language. The same stuff is involved in Visual Studio. Eye candy. Having done that, what does it get you? The number of RPG jobs out there where saying you have developed with "Websphere" Eclipse WDSc as a significant factor in choice has to be vanishingly small. There are so many more overriding factors involved that it's irrelevant. But you can download Eclipse for free and most any language IDE plugin with it. An interested person can and should do some web development with it for the experience and knowledge. That is unless they've decided to go the Microsoft route. Then they'll drop into the Visual Studio IDE, of course. I have always maintained that given same integration, the customer will choose the richest interface. Often integration and interface richness have been tradeoffs, however. The iseries~AS/400 offers a high level of integration with a powerful 5250 green screen interface. It would be difficult to make it richer without changing the communications to keystroke level, which a PC can support but an integrated server like an iseries~AS/400 mainframe cannot without having the equivalent of a PC for each session. This is readily apparent in the AJAX keystroke level interface in web pages. I do think and advocate the 5250 interface be extensible to a richer standard desktop interface, such as one based on Eclipse, but the keystroke type stuff where data changes on every keystroke or even cursor movement requires a fundamental tradeoff with integration. Customers choose the 5250 interface when they have experienced it. Dazzling effects are nice but they actually are trying to get their work done, and the iseries~AS/400 provides the most potent mixture of interface speed, power, and richness. rd

                      Comment


                      • #86
                        IMHO: What Will It Take to Turn the System i Around?

                        Ralph Daugherty wrote: I do think and advocate the 5250 interface be extensible to a richer standard desktop interface, IMO There is much unfinished work left to do in 5250 land. It is hard for me to accept IBM's line that all the new i5 OS development will be web based. After all they did recently enhance CL after many years of inactivity. Some 5250 basic improvements on the wish list (nothing fancy here) :
                          [*] Ability to select and mix typestyles on a single screen.[*] Ability to select and mix font sizes on a single screen.[*] More Colors.[*] Ability to imbed graphic files (JPG, BMP, GIF, TIF, etc) at specific character locations, The files could reside in the IFS, or be a BLOB field in a DB2 table.[/list]None of this is OOP, or revolutionary. It is merely an evolutionary and natural step in the right direction. Such enhancements will also give shops with heavy investment in HLL programming, the ability to enhance existing programs without having to sell the farm. Dave

                        Comment


                        • #87
                          IMHO: What Will It Take to Turn the System i Around?

                          Obviously the target demographics you discuss are, in my opinion at least, firmly still in midrange / mainframe market. Obviously, the vendors of these replacement solutions don't look at it that way though. But to be fair a lot of software implementation issues relate to the quality of their developers and how they target the implementation layer. I wonder how many people have dropped the AS/400/ISeries/I5/SeriesI over the years as the software applications offered on them were shown up by alternative products in performance or function. Does this mean that all Green Screen implementations are bad? Of course not. Anyway - work is over now. It's Xmas eve. Let's hope for a bright future for the Series I and a great new year to all. Ron.

                          Comment


                          • #88
                            IMHO: What Will It Take to Turn the System i Around?

                            Well, my work starts when working for someone else stops. I'm wrapping up some research for my next project. My previous post was on yet another J2EE failure in replacing green screen systems that are still chugging along, still doing the job that J2EE and web pages can't for any amount of money! No one knows how much could be spent trying because the plug is pulled after several hundred million dollars of futilty. All those hundreds of millions and billions of dollars are going to the major vendors, IBM, EDS, SAIC, BearingPoint?, and others who keep changing their names to protect the guilty. If you read the original justification for all these failed systems, they were like they were written by the same few people. Probably were. Having looked at Java web pages trying to replace green screens, let's look at Windows web pages trying to replace them: From Wikipedia: CalWIN, the CalWORKs Information Network, is an automated information system to automate eligibility determination and case maintenance functions for specific county-administered social services programs in the state of California, including CalWORKs, Food Stamps, Medi-Cal, CAPI, General Assistance, and Foster Care. CalWIN was developed by Electronic Data Systems (EDS), which also built, owns, and operates other major health and benefits information systems in the state. Under the CalWIN contract, state and county consortium pay EDS more than $800 million for the system. EDS is promoting the same technology in several states for proportionally equivalent fees; the Colorado Benefits Management System, now in operation, is one such variation. CalWIN is a Windows-based software package radically evolved from its predecessor, the mainframe, 'green screen' style Welfare Case Data System (WCDS), also developed and maintained by EDS. end quote So EDS develops a Windows based package to replace green screens, including the Colorado Benefits Management System. And what do the people in Colorado who have to use it think of it? from AP: "We were told by EDS they had a computer program that would address the needs of our constituents. It's a mess, it's a disaster." end quote from Glen Emerson Morris: "Automated functionality testing would have made EDS aware that the Colorado Benefits Management System ran so slowly that data entry timed out when moving between data entry screens. According to some reports 17 different screens had to billed out and it took up to 24 minutes for each new screen to load." end quote from Rocky Mountain News: "To make the system work, county officials have developed a six-volume set of "manual work-arounds" for resolving problems. Counties also have had to increase their staffing levels by 30 percent to 40 percent." end quote from The Gazette: "The CBMS System Cost 200 million Dollars. It was supposed to be more efficient, faster, require less staff, and provide better services. It will cost more to maintain, it is hardly working, much slower, and requires more staff to deliver the same benefits." end quote These Colorado and the previously posted Texas systems, by the way, being compared in speed to green screen systems developed in the 1980's and running on old IBM mainframes. And still they are faster and unmatched by hundreds of millions of dollars of state of the art web page technology by EDS or IBM. Ahhh, but new technology is maintainable and "legacy" technology isn't, yada, yada, yada is the IT mantra to justify all this. Let's take a look at the "maintainability" of the brand new Windows system. From Wikipedia: "Major operational problems, delays, and systems failures are occurring on a frequent and regular basis. Service requests to correct such errors are backlogged to the point that a two-month delay between a county filing and initial EDS response -- usually a rejection taking the form "Functions As Designed" -- is the norm. A period of debate and negotiation follows the rejection, consuming additional weeks. If EDS ultimately agrees to make the change -- or if the county pays for it -- actual implementation takes another six months or more. CalWIN programmers do not keep pace with regulatory changes: A relatively simple adjustment in income limits for the Food Stamps program, published early 2006 and taking effect November 1 of that year, was not programmed in time for the change. Instead, EDS recommended that counties employ a manual budget system, and has not set an estimated target for releasing accurate code in this area." end quote The people who have to use these "new technology" web pages want to go back to decades old green screens. They can actually get their work done that way. Merry Christmas to them and the lucky ones of us on an iseries. rd

                            Comment


                            • #89
                              IMHO: What Will It Take to Turn the System i Around?

                              Again, please don't get me wrong - I LOVE the AS400/iSeries/WhateverIBMcallsTheIbox. In referring to learning Webshpere, we get into the semantics of what IS Websphere. IBM has branded so much under the name, but I digress. I have used Code400 and RSE to program RPG, and it IS relatively easy to use. I don't see the productivity gains that are touted by using it, so prefer to use SEU for RPG coding. The part that I would LOVE to learn is creating the web interfaces, and so forth. But, that's just not needed at my current job, so if I want to learn, it's got to be at home on my own time. With no AS400 to attach to, AND no version of the software that I can own legally, that's impossible. I'LL SAY IT AGAIN: I as a person cannot own Websphere, unless I buy an AS400. If I could, there's AS400 sharing sites online that I could connect to with my own account. Then I would be able to learn the newer technology,and maybe show the benefit to the employer, as opposed to half-**s'ed thrown together stuff in my "Spare time" at work, that would not be seriously looked at. In summary, using RSE/Code400 and saying that I'm using Websphere is not what I'm after. IBM needs to get off it's corporate butt, and aggressively market the AS400 and find a way to really promote it's development tool and system to newer generations, or the machine will die from attrition. I don't necesarily LIKE MS Visual Studio programming, but there ARE jobs out there for it, and not too many for AS400 programmers EVEN IF they can do web programming with Websphere. It makes me sad.

                              Comment


                              • #90
                                IMHO: What Will It Take to Turn the System i Around?

                                I agree with all that, TAGrove. Well said. And certainly, I also came very close to going off to another platform because of no RPG jobs. I had been working on a personal Java project to come up to speed for Java certification for a few months before getting very lucky in finding an excellent employer running on iseries after I finally got my house sold and was able to move elsewhere to look. There was never an RPG centric replacement to our 5250 architecture because IBM staked their future on their Websphere J2EE server. At that point to them it became here's a web page, call something to do something with it. Thus a brave new world of stateless web pages calling programs rather state 5250 sessions calling screens was born. I'm convinced it's idiotic but IBM has bet their future on it. I just looked last weekend at a Websphere project that Texas pulled the plug on, CRIS, a Crash Records web page system. Maybe IBM knows what they're doing, but I think the architecture is backwards. Web people have taken something intended for random viewing of content and tried to replace the successful terminal session screen architecture with a Rube Goldberg meets chaos hodgepodge. Try for example even finding a web hosting company that will host Websphere if you were to develop something for the internet. I don't think it's possible without paying a professional services company, with matching fees. At best you can get some Tomcat JSP hosting (which JSP is the recommended solution by some on the iseries), but it is a mess. People stick with PHP because you basically have to have your own server and nurse it for Websphere Java serving, so hosts won't touch it. So there's a basic disconnect from an individual running this stuff when an enterprise team with their own servers is required. So to tap into it to train? I'd say IBM would have to have a training server set up to promote their technology, but all in all I'd say they probably have their own problems with it vis a vis Texas CRIS and the like and I wouldn't be looking anytime soon for IBM showcasing this stuff for iseries people. Just my opinion. Whatever is going on out there is all over the map. But I'm now focusing personally on RPG solutions communicating in a number of ways, not web pages calling CGI programs in a number of ways. Thanks for the clarification, TAGrove. And I hope something good happens for you and everyone else in the new year. rd

                                Comment

                                Working...
                                X