December Article by Ted Holt: The Death of OPNQRYF

[David Abramowitz]

December Article by Ted Holt: The Death of OPNQRYF

I don't know if there is a performance hit, but I'm sure IBM would love to supply additional storage and memory to make up for any. The real question here concerns maintenance costs. As Ted points out in the article, the AS/400 is, has been, and will most likely forever be, the only box, where a RDBMS may be directly accessed by a programming language. It is one, of among many reasons AS/400 advocates are so fervent in their support of the OS/400 operating system. IBM is going to bequeath new features, yet, not make them natively accessible. The whole purpose behind the zealous support of OS/400 will be gone! In order to access these features, you will be doing so from another OS, most likely Windows. DB2/400 is now called UDB, the "U" standing for universal. IBM's strategy here is to offer the same data access from I-series, P-series, X-series, etc. If I'm a corporate guy, I will either hold on to my existing hardware/software set (No sale here, IBM), or, I will examine my options. Since everything is now universal and equal, and I have to get new skill sets anyway, my next system, will be a less expensive server, where personnel with the skill set is literally around the corner, or coming off the boat. Once again, IBM, no sale. Ted's reference to YAS/400 is most apt. Once the 400 becomes the same as everything else, the purpose of existence will be eliminated. . . . . . which may have been IBM's game plan all along. Here's the potential good news: What IBM does or does not do is customer driven. A unilateral decision to do one thing or another can often be met by a brick wall of customers demanding someone's head on a stick. IMO, IBM will be put in touch with its customer base very quickly, once V5 is released. I suspect we will see native features enhanced. All it takes is a few big customers with a lot of noise. History does tend to repeat itself. Dave

[R.Daugherty]

December Article by Ted Holt: The Death of OPNQRYF

It's ironic that while IBM pushes Java and SQL which would require us to buy 20 AS/400's to do the same work as one AS/400 running RPG, those pesky competitors out there are writing Java and SQL that won't work on IBM computers (ok, maybe Oracle on AIX, if IBM really wants to stoop that low). Given the nature of the AS/400, many of us have experience with manufacturing and interfacing to shop floor control systems. Given my constant quest to find an actual working web page business app, a recent ComputerWorld article, "The Assembly Line Gets A Web Interface" in the September 25, 2000 issue, about a Manufacturing execution system called DataSweep piqued my interest. Written in Java and "web enabled", the hype is stomach turning: Implementation only takes 3 months versus 2 years for all that horrible stuff not written in Java; ability to feed real time information to a web site so that customers can track the status of their order (obviously requires Java to feed data to a web site); and reducing the manufacturing defect rate from 10% to 1% due to the miracle of viewing data through a browser, according to one of their customers. Yes, this Java web enabled software is the future, is it not? Oh by the way, those web enabled "clients" have to be Windows NT Workstation or Windows 2000 Professional, while the server has to be Windows NT or Windows 2000 Server running Microsoft SQL Server or an Oracle 8i database. IBM's vision of Java and SQL serving to browsers is startling in its clarity and foresight, a view so startling that it could only be obtained from the edge of a precipice... Ralph ralph@ee.net

[David Abramowitz]

December Article by Ted Holt: The Death of OPNQRYF

I understand what you are saying Ralph, and I do not disagree, but it should be remembered that IBM has backtracked, and made mistakes before. Some things that IBM has told us:

[*]The S/36 would no longer be manufactured after 6/17/1988[*]PL-I is the computer language of the future.[*]The keyboard of the future (no keys, just touch plastic) is being used on the PC Jr.[*]The PC Jr.[*]All PCs would be using OS/2[*]Taligent (The big pink project) would be the development platform of the future.[*]The MCI bus is way to go for PC add-on cards, and will be used on all other platforms in the future.[*]You must code to SAA standards on the AS/400 (1988-1990)[*]Office Vision will be dead by 1998.[*]You must change all your programs to modules in order to be competitive.[*]If you don't code for web development, you'll be flipping burgers[/list]In each and every case, IBM has failed to consider a company's existing investment, and the cost of shifting gears. The smart talk on wall street right now (read Barron's or Wall Street Journal) is that web development (for retail purposes), is simply not worth the investment. ROI is negligible at best, and market share growth is pure fiction. Yet every IBM demo I have ever attended uses retailing as the justification for web development. I have worked on a variety of platforms, and IMO, the native development methods of the AS/400 has (far and away) the fastest implementation times bar none. I find it most interesting that rather to bolster the "best" with additional features, IBM chooses to ignore that it exists. Even venerable RPG compiler development is trying to turn RPG into something else, forgetting the reasons that the language has been, and remains successful. Eventually, the market, and IBM's own customer base, will take IBM to the woodshed. Don't worry, , , , be cynical. Dave

[T.Holt]

December Article by Ted Holt: The Death of OPNQRYF

Frank said: DDS usage won't be eliminated completely -- you'll still need DDS for creating display and printer files because no SQL equivalent exists for those file types. Not so fast, Frank. You don't need DDS for display files, because you don't need any stinkin' display files. Green-screen is dead. There are only 3 places on the planet still using text-based computer interfaces, & yours is one of them. The rest of us write for browsers. As for printer files, none of us still use them, do we? Everybody I know went paperless years ago. Don't you do all your invoices, ship notices, checks, etc. thru EDI? SRC="http://home.earthlink.net/~tmeinen/faces/smilewink.gif" ALT="wink" WIDTH="15" HEIGHT="15">

[Guest.Visitor]

December Article by Ted Holt: The Death of OPNQRYF

As far as I Know - there are no Column Headings in SQL Design of Files. Moving From One File to Another - the code looks Awkward. it is hard to debug and to see which variable goes to which. in DDS it neat and RPG defines exactly which column goes to which. In any case if th S36 with all its limitations (4GIG of Disk, 64k programs etc.) was for so long, so the RPG and DDS much more so. Avrohom Notik Blue Square Co. Israel

[David Abramowitz]

December Article by Ted Holt: The Death of OPNQRYF

It took me a little while before I noticed the happy face. Whew. Dave

[R.Daugherty]

December Article by Ted Holt: The Death of OPNQRYF

Avrohom wrote: "In any case if th S36 with all its limitations (4GIG of Disk, 64k programs etc.) was for so long, so the RPG and DDS much more so." You and Dave and Ted are getting us off to a good start this week... Ralph ralph@ee.net

[B.Myrick]

December Article by Ted Holt: The Death of OPNQRYF

Avrohom, You can use the LABEL ON statement to do this. The following is the SQL statement to do such. Note the spacing. That gives me the three columns for description of the field names. > label on bret/brettest ( bmcust is 'Customer ID', bmname is 'Customer Name', bmstreet1 is 'Street Address 1', bmstreet2 is 'Street Address 2', bmcity is 'City Name', bmstate is 'State Name', bmzip is 'Zip + 4') -bret

[R.Daugherty]

December Article by Ted Holt: The Death of OPNQRYF

An IBM announcement on Dec. 15 that the Apache HTTP server is now available for the AS/400 in addition to the existing HTTP server had a couple of statements that caught my eye. One was that IBM is basing this server on the Apache open source web server, duh, and that IBM is providing this for free on the AS/400, like, duh dude. I don't know how IBM is skirting around the requirement to keep the code open source but I'm sure that the Apache Foundation would sue them if they tried to charge for it. The great statement in there is that the AS/400 now has a new visual interface. Huh? I looked hard at the Apache announcement trying to figure out what was the new visual interface and found that IBM created some admin web pages for the Apache server. This is now going beyond incompetence into realms of stupidity never trod by a major computer manufacturer before. The Windows admin program is a visual interface, and admin web pages for the Apache server is a new visual interface. The Sun's and HP's and Oracles's and Microsoft's of the world have got to be getting a chortle out of this one. Create some web pages and you have a new visual interface. While I was looking at Linux and the possibility of retaining some semblance of a career, I saw that HP-UX and Sun Solaris as well as Linux have implemented KDE. They know what a visual interface is. Neither AIX or OS/400 have an implemention of KDE, but for some reason HP and Sun saw fit to provide this Windows 95 like visual interface to their computers. We have the Apache admin web pages as a new visual interface. The Somers people need to be sent immediately on a forced march to Rochester for some rehabilitation training, otherwise known as getting a clue before we all go down with these chumps, that is, if they make it to Rochester... Ralph ralph@ee.net

[frankgw@adelphia.net]

December Article by Ted Holt: The Death of OPNQRYF

O.K. O.K. Ted, I won't ask you a green screen or printer question. We're phasing them out by year end anyway! But to Ted or others - My newest question is this. Should I plan to use SQL or Operations Navigator as my main tool of choice for my future external file descriptions? To Ted - Perhaps this could be the subject of a future article. Hopefully, very soon. We have a new programmer trainee. In what direction should we aim this guy? For that matter, in what direction should we aim the one and only current programmer at my site? Help! We are a small site and can only go in a limited number of directions simultaneously. TIA

[David Abramowitz]

December Article by Ted Holt: The Death of OPNQRYF

Frank Whittemore asked: In what direction should we aim this guy? IMO, the bullseye is directly determined by the corporate goals and objectives. What are your users' needs? What does your budget look like? And, most importantly, as Director, where do you see your company headed? Take all of this into account, and make the decision your own. Dave

[T.Holt]

December Article by Ted Holt: The Death of OPNQRYF

In what direction should we aim this guy? For that matter, in what direction should we aim the one and only current programmer at my site? IBM wants you to abandon native interfaces to the database & just use SQL. However, I know of shops that still use RPG III or even RPG II for development. IBM is warning us that they aren't going to make new enhancments available to native interfaces. I know people who aren't using constraints, triggers, journaling, or commitment control. Obviously they aren't worried about not having access to new database enhancements. IBM is telling us to move away from green screens. The new iSeries machines carry stiff fees for interactive processing. You have to decide what you think will get the job done for you. But if you remain with native interfaces & RPG, you are limiting your choices. How many hardware and/or software vendors support your decision? If you want to run RPG, you have 2 basic choices: an AS/400 or Unibol. (You might get Baby/4xx California Software to work in a pinch, but I wouldn't count on it.) Do you want to lock yourself into these options, especially knowing that IBM is abandoning the very stuff you're choosing to run your business? So, to answer your question, if you want to remain with the iSeries, then you should follow IBM, even if you think they're wrong, & use what they actively support -- RPG IV (not RPG III), SQL (not native interfaces) & some GUI strategy (either Windows or CGI) for the user interface. I'm talking only about new development. Some or all of the old stuff (at least the batch stuff) should continue to run as it is.

[J.Pluta]

December Article by Ted Holt: The Death of OPNQRYF

So, to answer your question, if you want to remain with the iSeries, then you should follow IBM, even if you think they're wrong, & use what they actively support -- RPG IV (not RPG III), SQL (not native interfaces) & some GUI strategy (either Windows or CGI) for the user interface. I'm talking only about new development. Some or all of the old stuff (at least the batch stuff) should continue to run as it is. So what do you think of my strategy, Ted, which is to develop your programs as you normally do, but deploy them using something like my revitalization architecture, which uses a browser or very thin graphical client? This way they can be developed and tested with a green screen, but in production they run in batch (allowing a much cheaper iSeries) and connect to a cheap PC running Java or a browser. Joe

[R.Daugherty]

December Article by Ted Holt: The Death of OPNQRYF

Ted wrote: "SQL (not native interfaces)" Ted, while I understand the practicality of what you're saying, could we drill down a little on this SQL situation? It's one thing for IBM to make no further enhancements to record level access and put all efforts into making SQL perform in competition with all the other SQL databases, but it's another altogether to read into that that record level access is to be discouraged and new development to be done in SQL. DDM and Java Toolbox implement record level access, and record level access is the very core of the AS/400. IBM's pricing on batch systems is not to punish green screens but to compete on price with other web servers without giving away the store. Again, I understand this as necessary to compete. Even if this were to somehow be construed as a hex against green screens as if to make them go away, what comparable action have they taken to inhibit use of record level access? When did they indicate DDM and Java Toolbox record level access classes would no longer work? Why does putting all their efforts into competing with SQL translate into something that makes no sense? It would take a strong statement of direction, way beyond we're putting all our future efforts into SQL, to make me think that it's not better to develop with RPG I/O statements. And I haven't seen it. Being competitive with the other SQL based servers does not require IBM to make any such statement discouraging use of native access, and if that statement I saw yesterday of one SQL statement being "easier" than two CL statements to add a databas e RI check is an example of IBM making such a statement, then I disagree... Ralph ralph@ee.net

[David Abramowitz]

December Article by Ted Holt: The Death of OPNQRYF

Ted Holt wrote: IBM wants. . . . IBM is warning us. . . . IBM is telling us. . . . I have no disagreement with anything you have said, Ted. My observations are that IBM is tiptoeing around this whole issue. This is because IBM is scared of alienating customers, , , ,and with this attitude they should be. If all that is left (of choices) is an IBM server to serve windows applications, or an NT server at 1/5th the price to do the same, guess where the customer will go. Once you become like everbody else, you lose the special qualities that made you attractive in the first place. ergo If IBM does not turn around, and start enhancing the features that made OS/400 the OS that people love to defend, these same people will go elsewhere. OTOH, if IBM makes native enhancements, then it will have a competitive OS that can be touted as an alternative to anything else. My prediction - Look for 128-bit, or even 256-bit hardware/software in a year or two. Dave