The iSeries: The Once and Future King

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Joe, I have always shared your enthusiasm for the iSeries (i5) and native development. I have yet to see an application in the Java world that could not be replicated natively on the iSeries. The problem is IBM needs to turn around and preach to the congregation not the choir. True they are spending mega bucks in advertisements, but I have yet to see one marketing rep in our company that is promoting the iSeries as a complete solution. Why? A move to Java, Websphere e-Commerce, Portal would require new hardware sales. Now isn't the real profits there for IBM. I share yours and everyone's enthusiasm for the native iSeries development, I will simmer down my optimism for IBM's announcement until they make them a reality.

[charles400]

** This thread discusses the article: The iSeries: The Once and Future King **
Two weeks ago my boss and I had lunch with our new IBM representative. We were discussing Websphere Portal, which IBM is pushing very hard. She asked what we were running in our shop. I was astonished that any IBM rep would not have done this homework before a meeting. We just upgraded to a nice big i5 through a business partner last year. In any case, we explained the ERP configuration on our i5. In response she said, "i5... is that like i/2? I've think I've heard of them..." Good Lord. From anybody but IBM I would have expected this. The sure sign of a platform in trouble is when the vendor selling it doesn't know what it is. I hope IBM's own people are learning from their own literature.

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Apologies in advance for what is a lengthy and bitter post. I think it's too late. I think it'll work, and I think it'll be great. For about 30 people. I think that there is no way IBM is going to win over a mass of new converts, particularly when it comes to RPG, and AS/400 technology in general. True, IBM has enhanced the ILE RPG language significantly, and actually brought RPG forward. There's even something of a migration path (although third-party tools seem to do a more thorough job than IBM's). True, IBM has done some remarkable things with DB2/400 in V5R3. True, you can code in C. True, you can try to code in C++. The problem is that while they were building a better mousetrap, people started figuring out how to either live with the mice, or have moved on to different locales where the mice don't exist. In my opinion, IBM has done absolutely nothing for the RPG community at large in terms of providing people with ***reasons to move to ILE RPG***, other than "it's better." I've been on the iSeries as long as you have, Joe. However, what I've seen several times is this: IBM's pitch: "It's better. See? Really! It's better! Look at it! Do you code RPG? Look what we have for you-- and it's ILE, it's free and it makes RPG coding easier! And look-- you can integrate Java, and CL, and COBOL and C. And look-- the database is slicker. You can do some superb stuff with journalling. And it's FASTER. How about it, huh?" And the RPG coders and the DBAs and the techie application designers scratch their heads and say, "Wow! It IS better." And these poor saps with big grins on their faces, go back to their bosses. These people are excited at the prospect of learning or implementing new technology on their favorite platform that will make their job easier. However, their bosses aren't interested. They're happy with their twenty-year old OPM RPG apps. The code works, it costs a lot less to maintain than developing new code, especially in what is essentially a new language. These bosses say things like: "How much is it going to cost me to use ILE?" "Why would I want to do this-- just because my staff thinks it's 'neat' or 'cool'?" "How long will this take me to realize any benefit?" "Oh, HELL, no." "You guys are nuts, and IBM is nuts for thinking I'd fall for this." "There's no benefit in employing new technology on the AS/400. The 400 will be dead in a few years anyway." I'm not just making this up-- I've heard these sentences from several "managers", many times. I suspect I'm not alone, either. Worse, I've heard people out-and-out lie regarding the stability or usability of the ILE environment on the 400. They'll make up the most egregious nonsense to keep people from learning or using "new stuff". My current client has a pack of team leaders who "search and destroy" any attempts by their staff to use or even learn ILE. I've only managed to get ILE in as production code because it meets project requirements that OPM RPG can't, and... well, I'm a consultant. I don't answer to those team leaders. There's a disconnect-- people can't see the value in investing in new techology, or get over the fear of new technology. As a result, the AS/400 evolves, but the shops don't. And IBM has absolutely nothing, repeat nothing, that will show a high-level manager in pictures (because after a certain level in Corporate America, nobody has time to read) that the overall cost of ownership for an AS/400 is competitive. In fact, I've seen eyes glaze over as soon as the words "Cost of Ownership" were used. And vendors are exploiting this disconnect. If your company has a project requirement which includes technology that you don't associate with the AS/400, like serving web pages, boom! There's a vendor out there who's ready to say: "The 400? Well, I know we've got someone who's done some work with it..." "The 400? Isn't that a dead box?" "The 400? Well, if I knew you were made of money, I would have raised my price 50%! Ha-ha-ha!" The vendor wants to make a sale. And his comments aren't unethical, he's just doing the best he can with what he knows. As an aside, most techie guys think you need x86 box with Apache to run a web server; never mind Apache runs on the AS/400, most of them don't know or care. (The amount of head-scratching when non-AS/400 people do when they learn all of the 400's capabilities would make you think there's a lice festival somewhere.) And the sales pitch sticks and suddenly people are saying, "There's a guy who wants to sell me a Linux box, with mySQL, and he says it'll be GREAT. He says I can retrieve four million rows in 2 seconds! And it'll cost me 10 grand to put the whole thing together? Wow. Neat. Cut a PO!" And then reality dawns: "Who's going to do this work? Hmmm... well, it could be the guy who's been on the AS/400 and written about half the apps in our system and maintained the other half... trouble is, he spits up hairballs and threatens to quit when someone says 'UNIX' or 'PC' within thirty feet of him. We could hire a consultant, but that's knowledge drain... maybe we can just get some kid out of college, pay him peanuts, and work him like a dog. Yeah, that's what we'll do." So, they hire the guy, and because everyone's busy and nobody "trusts" anyone else's technology, nobody gets cross-trained. The result? Silos. Everyone works in their little camps. Their attitudes are summed up as follows: "Stupid UNIX punk," says the AS/400 geezer, "_my_ stuff works." "Stupid AS/400 fossil," says the UNIX guy, "_my_ stuff works." The problem isn't the AS/400. The AS/400 is a superb box. Always has been, always will be. And also, the problem isn't the UNIX box. I've used 'em, and I like 'em. So, if the problem isn't the platforms, what is it? The problem is diversity. Many shops are multiplatform these days. And IBM has done a superb job of social engineering on the village scale, with their loyal bands of AS/400 folks, fighting for their machines. However, IBM has completely dropped the ball with regard to scaling the social engineering-- the AS/400 is an island, just like the Mac. Too little. Too late. Sorry, I don't believe it.

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Doc, You sound like a man scorned who thinks all women are out to get your money and treat you bad. While I believe your experiences are real I'm not sure I believe they are universal. There are many shops, including the one that I run, that are very progressively using the iSeries as the company's information backbone. While it's true that we don't exploit the deep crevaces of ILE (we only use free format, not modules, procedures, etc.) we use it to serve up web pages. A lot. In fact, any new user facing development is done via web pages. We do NOT use Websphere and, it seems, that IBM now understands that Websphere isn't the answer to everything. Also, there are many large companies doing similar things. Nearby companies to my own such as Countrywide Funding, Amgen, Disney, Columbia Studios, etc. (to name only a few) all rely upon the iSeries to run their business and serve up web pages. IBM is investing BILLIONS on the iSeries and one thing IBM doen't do is waste billions of dollars. My bet is on the iSeries. It may never become the Windows XP of the computer world but as a niche it'll do just fine. chuck Opinions expressed are not necessarily those of my employer.

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
I'm sorry that my previous post was so long. My points were actually pretty small, buried under a ton of rubble. To sum up, here they are: 1. IBM has treated the legacy AS/400 codebases as a waste of time and money for too long. 2. Other vendors are supplying solutions that meet immediate needs far better than IBM. Vendors have learned to exploit the gulf that's come up between the AS/400's users and IT management. 3. This isn't a case of EVOLUTIONARY v. REVOLUTIONARY. It's a simple case abandonment by IBM. While my criticisim of IBM's new ideas may be too broad with regard to the industry, I think the counterarguements are two narrow. For example, Joe's sold his (most excellent) tool to a shop in Wisconsin. However, just because one shop in Wisconsin is using your product, does not mean that you're getting full industry exposure either. You have a tool that you're marketing to AS/400 shops. In other words, the AS/400 is your market. Your customers are interested in keeping and modernizing their 400's and their 400 applications. Chuck, I work in integrated shops that use the 400 in a variety of ways. Without exception, all of these shops have had their 400's for at least 10 years, and most of their systems were migrated to the 400 from the S/38. Other systems arrived on the scene after the AS/400 became marginalized by IBM. I have yet to see a non-400 shop embrace the AS/400. By "embrace," I mean "buy a new AS/400 for core business applications". Even in buyout situations, I've yet to see the 400 become the main system. Also, where is the next generation of AS/400 jockies going to come from? Just curious. IBM doesn't seem to be too interested in that, either. Again, I think it's too little, too late. I hope I'm wrong.

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Doc, I agree with you on #2. Our programming efforts are based upon products supplied by BCD, a third party vendor. However, this is normal in the computing industry. Intel doesn't supply compilers, third party vendors do. Don't take this as a knock, but I think that sometimes a consultant's view of the iSeries world is a bit warped. Often consultants are called into companies that don't have the commitment to the platform to hire employees to do the work. They hire consultants to "get them over the hump" until they can move onto the projects that they have employees available to do the work. I don't use consultants in the iSeries world and hire only permanent programmer employees. We're not a big shop at only 10 iSeries programmers, but we're typical of a stable iSeries environment where we know what we expect of the iSeries and plan our staff accordingly. It is shops like our that consultants would have no experience with. And, believe me, there are TONS of iSeries shops that are staffed entirely by employee-only programmers. chuck Opinions expressed are not necessarily those of my employer. "E Doc" wrote in message news:6b229b3b.14@WebX.WawyahGHajS... > I'm sorry that my previous post was so long. > > My points were actually pretty small, buried under a ton of rubble. To sum > up, here they are: > > 1. IBM has treated the legacy AS/400 codebases as a waste of time and > money for too long. > 2. Other vendors are supplying solutions that meet immediate needs far > better than IBM. Vendors have learned to exploit the gulf that's come up > between the AS/400's users and IT management. > 3. This isn't a case of EVOLUTIONARY v. REVOLUTIONARY. It's a simple case > abandonment by IBM. > > While my criticisim of IBM's new ideas may be too broad with regard to the > industry, I think the counterarguements are two narrow. For example, Joe's > sold his (most excellent) tool to a shop in Wisconsin. However, just > because one shop in Wisconsin is using your product, does not mean that > you're getting full industry exposure either. You have a tool that you're > marketing to AS/400 shops. In other words, the AS/400 is your market. Your > customers are interested in keeping and modernizing their 400's and their > 400 applications. > > Chuck, I work in integrated shops that use the 400 in a variety of ways. > Without exception, all of these shops have had their 400's for at least 10 > years, and most of their systems were migrated to the 400 from the S/38. > Other systems arrived on the scene after the AS/400 became marginalized by > IBM. > > I have yet to see a non-400 shop embrace the AS/400. By "embrace," I mean > "buy a new AS/400 for core business applications". Even in buyout > situations, I've yet to see the 400 become the main system. > > Also, where is the next generation of AS/400 jockies going to come from? > Just curious. IBM doesn't seem to be too interested in that, either. > > Again, I think it's too little, too late. I hope I'm wrong.

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Joe, Joe said: JSP Model II has many advantages that RPG-CGI simply can't compete with (ranging from security to scalability to flexibility to cost). Can you expand on this? I'd love to know the details. One obvious advantage is being able to extend or implement a java class. RPG can't do that. The April 2005 issue of iSeriesNetwork has 2 articles on RPG-CGI. Chris

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Joe, I'd like to see an expansion of this also. Please place the explanation in the context of how it's REALLY used, not theoretical usages. For example, one would argue that a Plymouth Voyager simply can't compete with a Ferrari. But, in the real world we all drive the same freeways with the same speed limits so even though a Ferrari is a better car it may not be the better choice given constraints and requirements. We have hundreds of CGI programs in production that seem to work just fine. I'd like to know what, in the real world, what I'm not doing that I could have done in Java. chuck Opinions expressed are not necessarily those of my employer. "Chris Ringer" wrote in message news:6b229b3b.18@WebX.WawyahGHajS... > Joe, > > Joe said: JSP Model II has many advantages that RPG-CGI simply can't > compete with (ranging from security to scalability to flexibility to > cost). > > Can you expand on this? I'd love to know the details. One obvious > advantage is being able to extend or implement a java class. RPG can't > do that. The April 2005 issue of iSeriesNetwork has 2 articles on RPG-CGI. > > Chris

[buck.calabro@commsoft.net]

** This thread discusses the article: The iSeries: The Once and Future King **
Joe Pluta said the following on 3/29/2005 4:22 PM: > But all things being equal, the JSP/servlet architecture > is far more technologically advanced than RPG-CGI. Hi Joe! First, I agree with you. Second, having fought this battle and lost, I can tell you that pushing technology alone can be a frustrating experience. Without a real-world example of what JSP II brings to the table, I couldn't sell Java/JSP. Heck, I couldn't sell CGI either and for the same reason. The biggest problem for a green screener like me is that I haven't got instant command of a suite of JSP-enabled front end business solutions ready to pull out of my hat. On the other hand, I have a TON of RPG procs and snippets that I dazzle people (OK, maybe not dazzle) with routinely. I don't think there's a good answer for that dilemma. Without examples, there isn't a demonstrable reason to have a management push in favour of new tech. Without new tech, we can't really deliver web-enabled apps. And that's where the tools people make their living. --buck

[Guest.Visitor]

** This thread discusses the article: The iSeries: The Once and Future King **
Joe, Thanks for the reply. You asked: "You use a tool. So right off the bat, we have the licensing costs for the tool. How much are you paying for that?" It cost about $12k. Considerably less than the cost to train 10 programmers in Java. Joe said: "Second, you can't offload your CGI programs. This affects security, scalability and cost." I'm not sure what you mean by offloading CGI programs. Can you explain further? Joe said: "Changing a logo on a JSP is as quick as editing a text file." Same with us. I don't see any difference here. I can change a style sheet and/or an "include" and the logo is changed. Joe said: "Third, you cannot write and call Java methods. This simple capability allows you to do all kinds of custom formatting routines, write them once, and use them throughout your system." The "tool" we use allows us to create custom functions that are stored as procedures on the iSeries and can be called as often as we like. I don't know what Java methods are so maybe I'm not missing anything? Joe said: "There are just a few. I'm not going to argue the merits of your decision. In the end, it's a business decision. But all things being equal, the JSP/servlet architecture is far more technologically advanced than RPG-CGI." We agree. I'm all for getting solutions in place. With existing staff. Java doesn't offer that, the tool I chose does. With our purchased tool a programmers are productive with 4 hours of training and can be doing very complex tasks with less than 1 month of using the tool. Everything I've seen about Java tells me that programmers that have never programmed in anything but RPG their entire career will not be so productive so soon. As for scaling, I won't have a problem with that since we only deploy AS/400 related web pages to our WAN, they never see the Internet. All of our Internet development is done in C# using IIS and that seems to scale enough for our purposes. chuck Opinions expressed are not necessarily those of my employer.