Tweet
As a midrange 'dinosaur', I sympathize entirely with the basic belief behind this article. We are out of control; it's no longer a matter of getting the job done that counts the most but has become a matter of how you go about getting the job done. While not denying that our discipline needs help, I really do think we are running down the acronym highway to hell (without brakes). Some things that are old are good - that's how they got old. On the other hand, some things that are new are also good - they're likely to get old. I think tools such as WDSc/Eclipse are very helpful as are things like service programs, free format RPG, Java, OOP (if not given some holy grail-like status) but some of the things that are come up with make me shudder. They have a grain of sense, which is probably how they get accepted in the first place, but the emphasis is on the 'how' not on actually getting something done. There has to be a balance between theory and practice. As programmers, we use the best tools we have available and maybe the fact that some of these acronymic 'tools' don't really help will cause them to fall on their face. Unfortunately, I feel that what is likely to happen is that they will simply be replaced by some new thingy that everyone will view as a savior (until it too soon gets replaced). Life is interesting. Larry Manter
-
-
Service-Oriented Architecture (SOA)
Thanks, Larry, you got it in one. The fact is that the tools are now quickly being replaced by new tools in an ever-accelerating cycle. I've watched this for years in the Open Source world. Because backwards compatibility is usually compromised in the name of new features, that means that you must constantly rewrite your system or use old versions of the packages (it's even more of an issue when a new release actually breaks something in an older release). However, if you use old versions, then you run the risk that the next great Open Source project you want to incorporate uses a different version, and now you're stuck. Now entire technologies are flitting in and out of fashion. Has anyone tried to sell you EJB lately? I doubt it. Except for a few diehard consulting firms that basically bet the farm on EJB, nobody recommends it except for the very largest distributed pure Java projects. How about Struts? That's now being replaced by JavaServer Faces, which in mu opinion still isn't mature and is likely to be replaced or substantially rewritten yet again. This middleware issue is another example of the acronym soup. SOA is an umbrella for SOAP, UDDI and WSDL, all of which are actually being challenged as standards. So what you call SOA today may have NOTHING in common with what SOA is tomorrow. Funny, but I'm just not comfortable betting my business on soemthing like that. Especially when, as I said in the article, we've been doing the exact same thing for 20 years, WITHOUT all the acronyms. Joe -
Service-Oriented Architecture (SOA)
Joe Pluta wrote: Because backwards compatibility is usually compromised in the name of new features, that means that you must constantly rewrite your system or use old versions of the packages This is an interesting statement and leads to the topic of "Software Investment". AFAIK, very little is written for execs to absorb about the consequences of ignoring the current investment in their software. The end result can be tons of money being wasted year after year in the name of competitive progress. And after all of that the company ends up with no more additional functionality than they had previously. I don't have enough toes or fingers to count the number of times users have expressed the desire to return to their "old system". DaveComment
-
Service-Oriented Architecture (SOA)
Joe, your historical perspective articles are the most interesting reading of all, although trials and travails articles on getting future history to work run a close second. Reminding people of the jow cost effective our AS/400 and prior systems are and the people who made it that way is a good thing. rdComment
-
Service-Oriented Architecture (SOA)
Hi, I used to work for a trucking firm as well. And I remember writting code like you described to make PC's talk to our System/38. I'm in a big shop now, and don't do 'everything' anymore. Ahh, the good life. But I find that I've been out of it too long. Now I need to find out about new technologies so that my customers can work on the web and access our AS/400 data. I'm starting to look into JAVA and would appreciate any new articles in that area. I don't want to end up like the previous mainframe programmers around here. I noticed that some of my customers are starting to call the AS/400 a mainframe. Great...Comment
-
Service-Oriented Architecture (SOA)
And after all of that the company ends up with no more additional functionality than they had previously. This is the killer. Companies are being pushed into new technologies without a good idea of the ROI. Not only that, theyre being guided down buzzword alleys that are less and less stable, and ending up with not only a major investment that asnt really gained them anything, but in fact with a technological dead end that is already obsolete by the time theyve implemented it. We really need to step back a minute from this rush to deployment and return to the idea of basic business functions. Im going to try this month to get back to the IAAI website and build on the application issues I started. People need to realize that none of these new tools are silver bullets, and that you must make reasoned decisions on which technologies to implement. JoeComment
-
Service-Oriented Architecture (SOA)
Thanks for the kind words Ralph. More and more these days I find that people are madly careening down technological tunnels with their lights off -- heck, in some cases nobody's even at the wheel! I am gonig to steal one of our pop culture phrases and call this the "rush to deployment". They're deploying things simply because some techie pundit or self-hyping columnisttells them to, without making any sort of real business analysis as to how (or even whether) this new gizmo is going to help the company. WE MUST HAVE WEB SERVICES! WE NEED GRID COMPUTING! PLATFORM INDEPENDENCE IS KEY! Uh... why? Why do you need these things? How much money will it make your company, and how much will it cost? Is the former larger than the latter? It may turn out that some of these technological goodies are good for your company. Great! Find out how they work, and then incorporate them into an overall architectural direction. Don't just grab the latest wizard and plunk its output into your production systems. The less you know how your software works, the more likely it will break in ways you cannot fix, which ultimately is going to cost you far more than you could ever make from any new gadget. Joe Comment
-
Service-Oriented Architecture (SOA)
What version of OS/400 are you running? If you're at a relatively recent release, I am going to unabashedly recommend my two books to get you started. You will need to learn WDSC (WebSphere Development Studio Client for iSeries) in order to get into the world of web application design, especially if you choose servlets and JavaServer Pages (which is in my opinion the correct path). WDSC is the PC-based tool that takes the place of SEU for developing iSeries code, but it also adds ALL of the features you need to develop web applications. However, it's a pretty big learning curve, and it's not easy to get started. Which is where my books come in. WDSC is based on a product called Eclipse, which is basically a really powerful Java development tool (also know as an IDE, for Integrated Development Environment). Eclipse: Step by Step takes you through the basics of creating a simple web application using Eclipse. The next book, WDSC: Step by Step, then takes you through the process of using WDSC to create a web application that uses an RPG back end to present data - pretty much exactly what you're trying to do. While the example in the book is VERY simplistic, it will walk you through all the steps you need. Jon Paris reviewed the book and had some great input to make sure that the book is indeed readable by an RPG person with no prior Java experience. Somewhere along the line you may also need a book on beginning Java, such as Java for RPG Programmers. And all of that is available right here at your friendly MC Press Bookstore. I hope you don't mind the self-promotion. Joe Comment
-
Service-Oriented Architecture (SOA)
Joe Pluta wrote: People need to realize that none of these new tools are silver bullets, and that you must make reasoned decisions on which technologies to implement. You're a funny guy, Joe. DaveComment
-
Service-Oriented Architecture (SOA)
JThomas said: "I don't want to end up like the previous mainframe programmers around here." Why ? Whatever happened to them ? Hit by a meteorite called technology paralysis ... Scary yes ... JThomas said: "I noticed that some of my customers are starting to call the AS/400 a mainframe. Great..." Is AS/400(I5) another dinosaur in the making ? ... Not yet for sure...Comment
-
Service-Oriented Architecture (SOA)
Is AS/400(I5) another dinosaur in the making ? Well, before you go pronouncing the demise of the platform, let's get a couple of things straight. First, it's been around longer than just about any other platform you might be using today. Unix is older, but how many people are actually using Unix variants from the 70's? Instead, they're using a relatively new Unix, like FreeBSD, or even Linux, both of which are younger than OS/400. Heck, one of few Unix variants older than OS/400 is AIX... from you know who. So, the beast has been around a looooong time, demonstrating staying power. Since then, shall we see a few things that haven't quite panned out? How about J++? Or maybe EJB? Or maybe Struts? Let's see how long UDDI lasts. No, I think OS/400 or i5/OS or whatever they want to call it will be around a lot longer than a lot of the recent buzzwords. Or even some current ones which are losing favor (when's the last time you heard somebody rave about Extreme Programming?). Joe P.S. Anyone got a ZIP drive?Comment
-
Service-Oriented Architecture (SOA)
Stability used to count for something. Particularly when the CEO or CFO came up from accounting rather than sales. I'd be curious to know if anyone sees stability (of the OS) being valued in their shop. DaveComment
-
Service-Oriented Architecture (SOA)
It seems the value of a stable OS went away when GUI interfaces started becoming a "business requirement." You just have to put up with the rest of the mess that comes along with it (viruses, worms, continual patching, etc.). For that, you get the "added productivity" of a GUI interface. Now, don't you feel better that you are so much more productive?Comment
-
Service-Oriented Architecture (SOA)
Was that a rhetorical Question? It matters to me..... but obviously not to everybody. The quote below is from a Comp Sci professor at Berkeley. Maybe that explains something or maybe we're just doomed. At any rate, I found it to be both amusing and just a little scary. http://www.iseriesnetwork.com/resour...ageView=SearchComment
-
Service-Oriented Architecture (SOA)
The quote below is from a Comp Sci professor at Berkeley. Yeah just some professor ! Recovery Oriented Computing seems to be an extension into internet services of the RAID disks where Prof Patterson was a leader. (presume a disk will fail but implement a mechanism to recovery quickly and transparently.) Prof Paterson currently serves on the Presidential Information Technology Advisory Committee, Microsoft's Trusted Computing Academic Advisory Board, and IBM's Autonomic Computing Advisory Board. Check out http://www.almaden.ibm.com/institute...patterson.html Of course you could have submitted all those blue screen of death dumps to his research... http://winerror.cs.berkeley.edu/crashcollection/ Great StuffComment
Comment