Hi, guys
For years people, mostly academics, have been talking about COBOL's death. Finally, I came accross an article more realistic and from someone inside the business world which say what we all should be aware. COBOL is not dying?
Do you have news like this to share?
Regards,

Power Programming Point
-----------------------
www.powerprogrammingpoint.blogspot.com

Well, since I'm still writing new programs in COBOL after 39 years, I know it's not dying.  COmmon Business Oriented Language is still the best tool for coding most Business rules.

It's interesting how the majority of professors say the contrary. Many of them are out of the business field and have the wrong ideas.

I've been working with COBOL for ten years and many times, when I say to young IT professionals what I do, they seem surprised, to know that COBOL still exists, or worried, for fear that I won't have work for a long time.

Now, I discover that even the current fear of workforce decline for COBOL developers might be untrue. That is something.

http://powerprogrammingpoint.blogspot.com/2007/12/how-is-cobol-today.html

Power Programming Point
-----------------------
www.powerprogrammingpoint.blogspot.com

Quote:

It's interesting how the majority of professors say the contrary. Many of them are out of the business field and have the wrong ideas.

I've written on this much in the past (maybe here, but definitely other boards).  Usually they have a "we'll tell you what you should use" mentality, more than a "we'll teach what our students actually need".  A lot of pride, arrogance, and hubris is there, which goes along with the position.

Of course, if you go by job ads that I've seen, COBOL is pretty dead in the US (maybe not in India?).  If I go by message board posts, it seems a lot of people are converting COBOL systems to something else.  Dead?  No.  Dying?  Seems to be, and probably will die when the folks that claim double-digit experience retire.

Of course, if I look at job ads in general,  IT itself seems to be dying where I live (7 jobs in a major met. newspaper?).  There's always the pride in saying you were employed in something that went the way of the buggy whip.  Of course, there's always the retraining for the new career - assuming you can find the money to do it with.

COBOL is fast 100-200 times faster than Chrystal Reports to generate a report.  Monday "morning" report takes CR 8-1/2 hours, that means the president gets it Tuesday morning.  The COBOL version took 20-30 minutes.

100,000 times faster than VB.  VB File maintenence progran took 3-6 hours.  COBOL replacement completed in the bink of an eye, less than one second.

COBOL is self documenting, if written properly.  I have had useres and auditors vet COBOL code without knowing anything about programming.

COBOL supports business functions.  The Univerity of Nebrask at one time wrote a payroll system in FORTRAN!  What a disaster!  LAUSD payroll system is unable to pay any teacher or employee properly.  It is the current scandal.  It sure wasn't written in COBOL.

The article / author you cite... works for MicroFocus, so many readers might expect him to be a little biased toward COBOL.

Here's a very recent article that might be considered a bit more balanced / un-biased:

http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9062478&intsrc=hm_list

In the early 60s I used to write in Fortran, assembler and Cobol (IBM, burroughs, univac etc) but I love cobol very much because of it's simplicity, speed and a document by itself. Then came RPG and Basic which was followed by an explosion of object programming languages which turned the programming world around.

Today you can ask those big ERP software developers such as Peoplesoft, Lawson etc and you'll find that the core of their programming engine is written in what else but cobol!

Cobol nowadays are even more beautiful and powerful for they have now sub routines to call APIs written in C, java or any other scripting languages.

Truly the news of its demise is highly exaggerated (Twain)..

Yeah, the situation is worse: cobol programmers are dying, not cobol itself.
For me, I'm 49 next month. Under dutch law I have to work until I'm 65 at least.
As I forsee my future now: the last 5 years of my working life I will be a teacher earning more money that I can imagine now.

Wait and see.

Ha ha ha ha.... fun.... It would be great for you, Truus... but... if you see that a lot of work is exported to countries with low salaries.... then what is there to expect? Listened to BNR the ceo of LOGICA???? but.... I hope you are right....

Joepiemeloenie, in 10 to 15 years from now we search for topic again and the winner posts: "hate to say told you so".

And about the exported work: I am a strong believer in the 80 - 20 ratio. Even if 80% of all the work is exported, the remaining 20% is enough to fullfill my prophecy.

It will be as it always has been in the last 10-20 years.  There will be people that know how to do things, but the companies will do everything they can (including migrate systems away from COBOL) to not hire them and pay what they deserve to be paid given supply and demand.

Trans will be teaching some African tribe who only knows the $100 laptop that they've been distributing down there.

As an unemployed COBOL programmer, maybe the day will come that they might have to come ask me to do the teaching.   That'll be the day!  Ha!

Hi Glenn, how is it possible to be unemployed at this moment? There is a lot of work here....(NL) Tell us about your situation.

Yes. Please do tell us (me) of your situation.

Private email if you wish. frederico_fonseca at the domain below

Regards

Frederico Fonseca
SysSoft Integrated Ltd
www.syssoft-int.com

FAQ219-2884: How Do I Get Great Answers To my Tek-Tips Questions?
FAQ181-2886: How can I maximize my chances of getting an answer?

Sorry for the delay in a response.  Anyhow it's not anything I would want to get into in a public forum.  As for where I'm at (the US), most places are either rewriting their systems into something else or offshoring it (or a combination of both).  For the precious few advertisements I do see, there is incredible competition for them, or they are with terms that would end up to be foolish for me (or anyone else in my position) to take (i.e. move halfway across the country at my expense for a contract work that would only last 6 months, among other things)

Glenn,

Out of curiosity, do you by any chance have knowledge of PAXUS ?

Regards

Frederico Fonseca
SysSoft Integrated Ltd
www.syssoft-int.com

FAQ219-2884: How Do I Get Great Answers To my Tek-Tips Questions?
FAQ181-2886: How can I maximize my chances of getting an answer?

Glenn, sorry the hear that. So the U.S. is no longer the "promised land" for cobol programmers... used to be....

And how about the average age of the working programmers compared to your age? Are you "young & beatifull" and the're all grandpa's?

Quote (fredericofonesca):

Out of curiosity, do you by any chance have knowledge of PAXUS ?

No.

Quote (Truusvlugindewind):

And how about the average age of the working programmers compared to your age? Are you "young & beatifull" and the're all grandpa's?

I'm younger in age (much).  Part of the problem, I think is that I came out of a school that taught predominantly mainframe things just as the world (here in the US evidently) was switching to a PC/C++/Java/.NET focus.  Doesn't bode very well for one whose main programming classes in the bachelor's degree involved COBOL, DB/2 and CMS.  

Not to mention that no company ever is willing to train and wants work experience alone (and not stuff like what I do over in the Delphi forum).  At least the plus side is that I'm getting a few positive comments over my little hobbyist projects.

And not coincidentally the college I got my degree from switched focus to the PC not long after I graduated.  Says something, no doubt.

I just retired from programming after 33 years of using COBOL.  For the past 20 years I taught new hires to our organization how to write COBOL code.

Just like with most places, our management believed all of the hype about COBOL dying.  In the mid 80's when PCs were getting big, we went with a company who had developed a new language that could access our mainframe database.  Our VP declared that we were going to convert all of our COBOL to the new language.  We taught all new programmers this new language and did not teach COBOL.  Soon we found out as others have noted earlier, the PC language could not process the massive amounts of data we had like the COBOL code could.

We finally came to the realization that you don't throw out your tools just because there is something new.  You use the best tool for the job.  The PC languages were great for painting screens and collecting input, but COBOL was best for processing all the data we collected.  We went back to teaching COBOL along with VB and Java and are using all of these together to produce our systems.

I agree with many of your comments regarding professors in colleges.  We almost exclusively hired new college grads and trained them.  Only one or two of the colleges we recruited at continued to teach COBOL.  Most of the new hires could not believe we still used COBOL.  But also, most of the new hires had no idea how to write a computer program.  They could use project management software, e-mail, and search engines on the web very well, but couldn't put together a simple program to read a data file and write a report.

Interesting thread.

<soapbox on>
I come from a Computer Science educational background.  Colleges offering Computer Science have NO business teaching COBOL and never did (except perhaps smatterings of COBOL in a Programming Languages or compiler design class).  Their focus is on broad, theoretical topics related to the underlying science of computing, not on purely practical topics.  A CS major ought to learn about operating systems, programming languages, algorithms, numerical methods, data structures, computer architecture, data base theory, and computer networking, among other things.  None of these is particulary practical in terms of getting a job as an applications programmer.  However, a CS degree is excellent preparation for any number of practical careers, including COBOL programming if that's your thing.

To carry that a step farther, any baccalaureate program needs to teach its students in broad brushstrokes and not in narrow niches.  It's OK to teach COBOL, but not as an end in itself.  It ought to be done in the framework of a degree in MIS (or whatever its called these days) and ought to include training in other related skills (e.g. accounting, data base design, systems analysis), languages (e.g. java, VB), and tools.  The key point is to produce graduates who know how to think and how to learn and who have a solid, well-rounded foundation to build on.  That way they can learn the language du jour and specific business practices on the job as they need them.

My present employer is a bit of an anomally.  We have been training 30-45 or so new college graduates in mainframe skills every year for the past 10 years.  Most of the 400 or so graduates of that program are still employed here.  We also employee a huge number of mainframe contractors for multi-year contracts.  In my group, the FTE/Contractor ration is 55/45.  Yes, many of us are old and gray, but we do have a good bit of new blood.  AND we're also pushing the frontiers of mainframe programming as we delve into SOA, Linux on the mainframe, etc.

We're ALWAYS looking for good mainframe talent and we have a hard time finding it.  I'd guess 80% of the mainframe COBOL programmers whose resumes pass initial screening can't satisfactorily pass even a cursory technical interview on COBOL, CICS, JCL, and DB2.  The point I'm trying to make is that many of us have an outsized view of our own knowledge and abilities.  Remember the old saw about the guy with 30 years of experience who really had 1 year of experience 30 times.  If you're not paying attention, its easy to let your skills atrophy or become completely invested in your particular company's standards, tools, etc.  I think that in the long run it pays to broaden your skills and invest in keeping your technical skills sharp even if its just within the framework of COBOL applications development on the mainframe.

<soapbox off>

Regards,

Glenn

Quote (coboldeke):

But also, most of the new hires had no idea how to write a computer program.

Quote (3gm):

I'd guess 80% of the mainframe COBOL programmers whose resumes pass initial screening can't satisfactorily pass even a cursory technical interview on COBOL, CICS, JCL, and DB2.  

None of this really surprises me for what I've seen in my college experience.  Usually, most of them would get help to the point of it being a "do my program for me" affair or just muddle through with sub-par passing grades (C or D).  And quite easy to figure out that they were learning nothing out of the venture, too.  

A great example (out of many!) was a large group of people in a senior-level Oracle class (with pre-requisite class involving SQL - the senior-level class was about database administration) who could not even perform the most basic of SQL queries.

Rant time
It's become quite obvious to me in my numerous interviews that the last thing all the companies I have encountered are hiring for is ability to do the job at hand.  With the first being the ability to be the life of the party (i.e. exemplary social communication skills - not professional communication skills), it should be no surprise that they are getting what they bargained for.

And I know I wasted all my time studying/working/preparing/training to be as good as I could be at the job I was looking into, given how I fared compared to these folks in my examples.  Perhaps I should have been like those in the examples and just merely skate by in the course work, and concentrate my time on plotting the next party to attend to sharpen those prized social communication skills while getting drunk.

But that's the nature of the beast...
Ending rant now

Glenn9999, did you receive my email?

Tom Morrison
www.liant.com