Re: 4GL Certification Test Writing Contest
Georges A. Tomazi wrote: > > Tim Schaefer wrote: > > > > > Write questions for the INFORMIX-4GL Developer Certification Exam > > > and receive $50 for each one chosen for the final pool of questions! > > > > Forgive me for being a devils' advocate, but I just have to ask: > > > > Why on earth would anyone care about this now, with 4GL on a > > respirator? It's a dead product. > > > > I mean, this looks like something that should have come out when > > 4GL was born, not now, when its' life is ending. Maybe you could > > sponsor this for Universal Server or for something that's worth > > spending the time and money on. > > I'm afraid Tim that you are wrong... > > From what I've heard, today the I-4GL market still represent for > Informix a tremendous income. Thousands of companies all over the world > are still using the I-4GL language and can't afford to switch to > something newer. These companies are still buying development licenses > and runtimes... > Companies are amazing when it comes to what they can afford. I think you are correct in your assertion that INFORMIX is making a lot of money off 4GL--and well they should. I too have benefited from it. However, in planning ahead, I do not see 4GL surviving much longer for several intuitive reasons: 1. Program bloat. 4GL compiles to an ungodly amount of executable now. The average size of an executable has increased steadily over the years to 4 times what it once was. Let's not optimize the compile, let's buy a faster computer, and more disk. Compare 4GL to COBOL, and we see a strange cross-over. Look at the size of an executable COBOL program sometime, compared to the source code. It's tiny. Look at 4GL, small source code, monster compile. Does it really matter? Enterprise-wide it should, as you begin to look at multiples of programs. And the network--oh yeah... uh... And then you see these add-on products that REALLY make a mess of it. And then add new-to-4GL programmers to it, who don't know how to make it optimal--it gets ugly... I don't see INFORMIX denying bloat, I see them offering Universal Server, as a solution to not just 4GL, but the whole data base. They appear to be on track by catering to the current wave for the internet/intranet. I hope the best for it. It will have its' bugs but so has everything else. 2. No new real changes to the language in howmany years. The language has not evolved much into a better product. It didn't really need to be object orientated, for what it is. Many of the improvements sought by leaders in 4GL were never realized. Don't get me wrong, it is a great product, and I'm a big fan of 4GL, but I have to see it for what it is. 3. 4GL is proprietary unlike COBOL, and thus will never become truly mainstream. Two little concepts: ubiquity and portability. In this current market the most portable and most present will survive. Unless contributed to the public domain it will die it's appointed death as the product reaches the end of its' life. This is at the mercy of the market of course. Sure, there are new customers for 4GL every day. But who will code 4GL? Old men and women? There will continue to be a talent pool for 4GL, but the lure of GUI will take many of the new programmers away from 4GL, and thus the product will die of natural causes. It will be the COBOL for the 21st century. As I walk across that bridge I'll have waved good bye to it long ago... :-) 4. New technologies are emerging and gaining ground to take 4GLs' place. It's true, if you skipped the client/server 101 class, your company saved a hell of a lot of money, now that intranets are the buzz. You can go directly from mainframe to intranet, skip UNIX altogether, and still have money left over for a company picnic, and not have to downsize. :-) Seriously UNIX will continue, especially with Windows NT 4.0 and the latest revelations of the scam Microsoft is pulling. ( Server and Workstation kernels are one and the same, see: http://www.zdnet.com/wsources/update/setup.html and http://www.winntmag.com/Editorial/Nov96.htm ) Yes, I'll be using NT, and UNIX. But the idea of being connected to a machine constantly via a terminal is beginning to wane. You can see the writing on the wall. Tcl/tk has saved Xwindows for a bit, but the concepts behind web-based applications are changing everything. There's no reason why we won't see completely browser-based accounting applications in the near future. I can already access my bank account with my web browser, and have you seen the Netscape App Foundry? Even INFORMIX has apps in there... > NewEra is not compatible with I-4GL and the Compass project is very > expensive and takes a lot of time. A lot of people who even tried > NewEra for new developments gave up after a while. > A grand experiment. A noble effort. And we had to learn it to know that what will be in the next few months could not have happened without this learning. I commend INFORMIX for taking the effort they did for New Era, but my cheers really go out to their forward thinking for what Universal Server is and what it will do. It's going to change everything. > Now, Informix doesn't have another development environment than I-4GL, > and it's not a bad language, even if it's 10 years old. In you read > daily "comp.databases.informix", don't you see heaps of job offers > for I-4GL programmers ? Don't you think that's a market sign ? Also > look how many companies are selling developpement tools around that > language (compilers, CASE Tools, etc...). > It's a market sign, not unlike the TV signals that went out into space 20 years ago, and are now returning. They return and we are amused. The 4GL activity will continue, just like TV, and just like COBOL. The current action, TODAY, is all about the web, inside and outside the corporate castle. The sad truth is that many companies are not ready for the web, they still haven't caught on. But this will soon be changing, just like the recent flurry of DBA work that was almost non-existent a few years ago. When we can do apps with a browser or applet that do what 4GL does, what the hell do we need 4GL for? 4GL was good--damn good, but it isn't the smartest way to do things anymore. I'm gonna miss it. The new era is here, but it ain't what we thought it would be...and I really believe the best is yet to come... :-) > Georges > -- > Georges A. Tomazi > Four J's Development Tools > > tomazi@4js.com | 75020 Paris (France) | URL: > http://www.4js.com PS Georges, take a look at the cover of the latest Computerworld, the November 11 1996 issue. My intuition never fails me... -- Tim Schaefer \\\\|// tschaefe@shadow.net (6 6) ------------------------oOOo---( )---o00o--------- http://www.shadow.net/~tschaefe Liberty: http://www.lp.org