Some one was describing FORTRAN and COBOL programming languages as archaic and obsolete; but i later on discovered on the internet that people still make use of them, and they have been undergoing reviews to meet up with 5th generation programming languages.

I discovered on the internet that there is an object-oriented FORTRAN version that allows you program just like visual basic. Please I need to hear from you guys about this object-oriented FORTRAN version and the latest COBOL version.

You haven't stated what your interest is, so I assuming that you are considering a learning/career path.

In that case, I would like to share my experience; usually I have come across FORTRAN/COBOL programmes when undertaking a migration from a legacy application to a more distributed one. Legacy apps are usually mainframe ones and are considered too rigid for today's dynamic use.

It may be a good associated skill but, it will be plain suicide to hedge your bets with such a small arena. Another thought is the demand will be more in demand in self build mainframe markets like eastern Europe and defence related markets, even they are gradually moving to more maintainable architectures based either on UNIX or even Linux.

The only guys I know who have experience in FORTRAN/COBOL are 55yrs plus and only get work on highly security-classified system they worked on when in their 30yrs !

On the other hand there might be a Renaissance!

Enjoy!

It's easier to get free assistance from others on popular programming languages like PHP, Java, C++, C# and Visual Basic .net because there are so many people who use them. If you get stuck with COBOL or FORTRAN who is going to help you out? If you write your company's software with COBOL or FORTRAN where will you get good programmers to maintain the software for you? They'll be more expensive!

pls i need to know if FORTRAN and COBOL languages are archiac and obsolete languages. I discovered recently on the internet that there is an object-oriented version of FORTRAN now and a reviewed COBOL version. Pls i need to hear from u guys if these languages are still useful to us. Do people really use them to develop applications presently, compared to the more recent languages available?

Sorry i did not state my area of interest. My area of interest is developing scientific, engineering and mathematical applications. Do you think FORTRAN will be suitable?

If you use any imperative procedural language, you are already a fortran user!
This applies to people who program in vb (pre dotnet) or c. Or even to pseudo-object oriented languages like c++.
this is because fortran is simply a subset of the lexicon of modern imperative languages- it was the first imperative language created.
thus it is useless in today's world, cimply because it exists in so many advanced forms. Choose one of the advanced forms for practical use, and study fortran as an academic excercise. Fortran is an excellent introduction to procedural programming because, unlike the others, it does not railroad you into a particular way of thinking.


COBOL should, similarly, be the father of declarative languages, but today's declarative languages are so far removed from it that it cannot be approached in the same way. In many ways cobol is obsolete because the things it does are done so much better by today's rdbms packages and pseudo-imperative query languages. However because it is so far removed from modern languages, a lot of legacy code exists in cobol that cannot be trivially rewritten (unlike fortran), so ironically the skill still has some value, in maintaining or updating legacy code.

In particular the Y2k crisis required the work of several cobol programmers before it was smoothly averted.

My area of interest is developing scientific, engineering and mathematical applications. Do you think FORTRAN will be suitable?

I think so. I'm just concerned that your area of interest might not be lucrative since Matlab does all. It's always good to pick one's area of interest based on what is in demand, so one can make money.

Fortran can be suitable for scientific programming, indeed it was designed for it (Formula Translator). As for matlab, it depends on what you are trying to achieve, if you want nice graphs etc, to present your result, then go for it, but it will cost you - a lot besides the syntax is very similar to C. There are so many other good options than may prove more useful after your scientific programming days, principally C/C++, Java ( except for the massive overhead) etc, You an achieve the same results with these languages. C/C++ share an extensive math library for many of the basic functions you will need. I think many of the engineering software, at least the underlying algorithm, is still written in C because of its speed advantage.

The idea that fortran is suitable for scientific programming is what is obsolete.

Remember, at the time fortran was invented, imperative languages did not yet exist.

The formula translator was created for scientific use at that time. It just so happened that its style turned out to be so flexible, it practically overshadowed all the other major programming styles. In fact most 'programmers' today dont know anything more than imperative and object oriented programming- and most oo languages are in fact, imperative (even java). Nowadays any imperative language will serve the same purpose much better.

In short, take the advice of people on this forum and grab a modern language.

the main people using fortran these days are people supporting legacy code.
New developemnt projects are done in the newer languagues

please i need to know if FORTRAN and COBOL languages are archiac and obsolete languages. I discovered recently on the internet that there is an object-oriented version of FORTRAN now and a reviewed COBOL version. Please i need to hear from u guys if these languages are still useful to us. Do people really use them to develop applications presently, compared to the more recent languages available?

Yes people/company still use em. You will be surprise to hear that Walmart, Target and most of the big retail stores still use Cobol today.  Its one of the requirement for their in-house programmers. If you have the chance/time to learn fortran or cobol, do it. Thing is, be the best (or at least strive to be) in whatever comp language you are using or decide to use.

Walmart uses COBOL?

what for?

I can't imagine pos systems being written in cobol? what do they use it for? do they still use dumb terminals and mainframes? how do they set them up in diff. stores?

Thanx guys. Its just dat some are saying its ok to learn it while others are saying its not ok. Which ever sha i think its still ok 4 one to ve a little knowledge on fortran and COBOL too. I,m already into visual Basic programming. Still need your advice.

Ynot:

Yes people/company still use em. You will be surprise to hear that Walmart, Target and most of the big retail stores still use Cobol today. Its one of the requirement for their in-house programmers. If you have the chance/time to learn fortran or cobol, do it. Thing is, be the best (or at least strive to be) in whatever comp language you are using or decide to use.

The reason Walmart and the likes like various banks still use it is becasue of the need to support legacy code that was writtin in the 70's and 80's . I have been involved in a project like that We had a legacy Credit derivatives code written in fortran for an investment bank. They wanted to be able to provide new services for some high profile coustomers via web services. This was not possible to do in Fortran so what we did was that we developed a bridging code in fortran to a new application which was developed in Java . Clients still access the legacy code but through a web service interface developed in Java

Why not just rewrite the legacy code?

I imagine this bridging solution will incure a serious performance hit.

Fortran should not be too difficult to translate to a new language? Its not like COBOL where entire programming concepts will have to be rethought- a fortran program should be quite easy to rewrite in any modern imperative language?

My Final year project depended almost entirely on programming concepts illustrated in fortran, which I implemented in VB. Wasn't hard at all- in fact the code looks so similar.

dakmanzero:

Why not just rewrite the legacy code?

I imagine this bridging solution will incure a serious performance hit.

Fortran should not be too difficult to translate to a new language? Its not like COBOL where entire programming concepts will have to be rethought- a fortran program should be quite easy to rewrite in any modern imperative language?

My Final year project depended almost entirely on programming concepts illustrated in fortran, which I implemented in VB. Wasn't hard at all- in fact the code looks so similar.

Writing an application with about 1 million lines of code just because u want to provide new services for some new customers? oh pls, Thats not only lame its just not practical. The Fortran code works perfectly ok and there is no reason whatso ever to attempt a rewrite. The bridge code were just mere API'S to some underlying logic much where called via JNI from the java interface and considering that this code was called by JNI, it performed just as well as a Fortran code and much more faster than a similar java code.

Lame? impractical?

I suppose it would be impractical and/or lame from a consultants' point of view.

I imagine what you probably meant to say was 'rewriting legacy code would have been too much work- and why do ajayi work, afterall na my faddah own di bank?'

In summary: code lying around in a nonportable language is a bad thing. To rewrite it is neither lame nor impractical, but neccesary. Just say you would rather *you* didn't have to do it.

dakmanzero:

Lame? impractical?

I suppose it would be impractical and/or lame from a consultants' point of view.

I imagine what you probably meant to say was 'rewriting legacy code would have been too much work- and why do ajayi work, afterall na my faddah own di bank?'

In summary: code lying around in a nonportable language is a bad thing. To rewrite it is neither lame nor impractical, but neccesary. Just say you would rather *you* didn't have to do it.

Majority of furtune 500 companies have legacy code writtin in your so called nonportable langugues driving there business. Even the US defence department and NASA have a lot of code writtin in Fortran, Lisp and smalltalk langugues

Nobody in his right senses would justify rewriting a perfectly working application just because it is writen in a Fourth generation language.

<<snip>> don't even know what a fourth generation language is.

Anyway, it's not just because it was written in a THIRD gen language or whatever.

It is written in a NONPORTABLE language.

Because big companies use crap doesn't make it right.

Afterall, Yahoo uses PHP, ne? And I agreed with you on another forum that php is a rather ugly language due to its weak typing.

dakmanzero:

<<snip>> don't even know what a fourth generation language is.

Anyway, it's not just because it was written in a THIRD gen language or whatever.

It is written in a NONPORTABLE language.

Because big companies use crap doesn't make it right.

Ignorance is bliss, educate yourself here http://foldoc.org/?legacy+code

Big companies dont deliberately write code in these languages these days but they still do because they sometimes have to provide backward compatiblity.

A computer system or application program which continues to be used because of the cost of replacing or redesigning it and often despite its poor competitiveness and compatibility with modern equivalents. The implication is that the system is large, monolithic and difficult to modify.

If legacy software only runs on antiquated hardware the cost of maintaining this may eventually outweigh the cost of replacing both the software and hardware unless some form of emulation or backward compatibility allows the software to run on new hardware.

Thanks for proving my point!

Emulation is a performance-sucking kludge, and backwards compatibility is a sometimes neccesary evil, but evil nonetheless. Legacy free is good. That is the beauty of PORTABILITY.

My point: just because something is difficult to do doesn't mean that managing a second-rate solution is right.

dakmanzero:


Emulation is a performance-sucking kludge, and backwards compatibility is a sometimes neccesary evil, but evil nonetheless. Legacy free is good. That is the beauty of PORTABILITY.

My point: just because something is difficult to do doesn't mean that managing a second-rate solution is right.

Nobody is talking about rightousness here. The augument is that these langugues are still relevant to todays busnessis because there are lots of code laying around in this langugues. And i can tell you with a fact that companies pay huge for people that can help them maintain these systems if required.

Yeah, you are absolutely right

We have a consultant at the company I work for who makes obscene amounts, so huge I cannot mention them here writing ridiculously simple translation layers, wrappers, etc for some of our legacy systems, whenever we need to connect them to modern 'channels'.

He has succesfully reverse engineered a large, poorly supported system that we and a number of other companies in our line of business depend upon. Since he is the one and only person around who can do this, he can demand for whatever he wants and he is paid! I swear you will pass out if you hear what he makes.

We are currently in the process of putting an end to that and guess what? He is in a state of panic. All of a sudden it costs more to pay him, and less to simply replace the legacy system, hardware, infrastructure, everything! The new system will support open interfaces and will integrate better.

So, in summary, that the legacy languages (in this case it is RPG, the report generator, similar in construction to COBOL and ugly as hell) are relevant is absolutely true. However, the need to replace them with newer, more portable, more open systems is neither LAME nor IMPRACTICAL. And to demonstrate a preference for the legacy-free approach does not demonstrate IGNORANCE.

dakmanzero:


So, in summary, that the legacy languages (in this case it is RPG, the report generator, similar in construction to COBOL and ugly as hell) are relevant is absolutely true. However, the need to replace them with newer, more portable, more open systems is neither LAME nor IMPRACTICAL. And to demonstrate a preference for the legacy-free approach does not demonstrate IGNORANCE.

It was lame and impractical in our own case because the legacy system still does work perfectly even though it was no longer supported . Therefore the cost of a total redevelopment far outweighed simple maintaince that were carried out.

The world would definately be a better place if it was free of legacy code but in reality it definately would never be,  take it or leave it.

key word: No longer supported.

They will soon run into issues of media incompatibility, parts scarcit,

Jesus! This was a discussion about fortran and cobol being obsolete! See how the thing done metamorphosize.

Funny enough, name-calling aside, it does indeed seem hat we are in agreement about quite a number of things.

So, lets round this up:

I guess it didn't make sense for your firm to rewrite the code for them, since you weren't being paid a sum commensurate to such a task. Indeed it would be impractical for you to do such. And, I reluctantly agree, lame, (tho I'm not so sure that what amounts to a huge act of charity qualifies as lame!). But it would not be lame and impractical for THEM to do so! In fact it would benefit them in the long run, ne?

Glad you agree that the world would be a better place, legacy free. And rest assured it will be! Although it is not the most wholesome of methods, the Trusted Computing Initiative will soon make sure of that. I assume the only holdout against this onslaught of DRM and 'managed' programming languages will be FOSS software, and they are focused on portability, which is also an opponent of obsolete legacy systems!

As Richard Stallman said, code should be allowed to die. Standards should be allowed to live. Ignoring these basic rules is what leads to the mess of obsolete, cuddy code mucking up our systems and giving us all a bad name.

dakmanzero:




Glad you agree that the world would be a better place, legacy free. And rest assured it will be! Although it is not the most wholesome of methods, the Trusted Computing Initiative will soon make sure of that. I assume the only holdout against this onslaught of DRM and 'managed' programming languages will be FOSS software, and they are focused on portability, which is also an opponent of obsolete legacy systems!

As Richard Stallman said, code should be allowed to die. Standards should be allowed to live. Ignoring these basic rules is what leads to the mess of obsolete, cuddy code mucking up our systems and giving us all a bad name.

Wake up from your dream man ,  u seem to read too much junk and theory,  The real world is not as utopia as u wish it to be.

@ topic

Yes, they are obsolete. It seems they still teach it in schools but I think lecturers should look more into C++, Java.

I wouldn't say Fortran is OO oriented although its got functions (which are like methods in java) that act seemingly so.

Fortran might be good for engineering, its got lots of maths functions (they say, I couldnt find them) .

Theory is where everything you are living today, was in the past.

I've been into computers for the better part of my life, and I've seen mere ideas become reality and *then* accepted fact! Its a lovely experience and you can try it!

That's one of the best things about computing- in our world, the world of code, ideas can become reality! And they do, by far more often than they do in the dreary, mundane non-digital world, where ideas remain ideas forever.

If you can abstract it, it CAN happen in the digital world! And the total obliteration of crappy legacy nonsense is much closer than you would think.

If you really are world-weary and cynical enough to turn up your nose at such a possibility, I urge you to read up on the TCPA, or trusted computing initiative. It is, far from being an idealistic fantasy, a terrible, sinister reality that is already taking root- witness windows vista, the intel core duo, the EFI in intel imacs, we recently took order of IBM thinkcentre servers and guess what? TCPA signatures in the root of the drive.

Anyone basing anything of value on legacy-encumbered, unportable systems is soon going to have to face a harsh reality! could YOU imagine a world without floppy drives? a world where operating systems in excess of ten gigabytes are freely available and distributable over a worldwide network of billions of diverse computing devices? These were mere dreams when I started programming on my 32 kilobytes strong BBC computer back in the late eighties, when I amused myself burying my nose in one-year-out-of-date copies of PC world, where men like John Dvorak predicted things that have come to pass almost a decade ago!

My friend, in the world of code and coding, it pays to think ahead, it pays to dream, to imagine. Because here where we are, dreams are made real. And the day that ends is the day I give up my lifelong fascination with the timeless poetry of code.

Have u guys not seen this before?

News
Forpedo Includes Run-Time Polymorphism
Forpedo has been upgraded to include support for run-time polymorphism, as described here. Run-time polymorphism is an important part of Object-Oriented Programming (OOP), and is supported in Fortran 2003, but not earlier standards. With Forpedo you can mimic run-time polymorphism in Fortran 90/95 programs, making the transition to Fortran 2003 easier when compilers of that standard become readily available.

Don't you guys think that the latest version of FORTRAN is fully objected oriented now?

It must offer what no other language offers before it can be reckoned with. It wont be much good if the coding format is not similar to the popular programming language. Its difficult to get help on FORTRAN.

very informative Post (thanks to darkmanzero and clocky!!! ) I think why ishmael is interested (dont know, really) is that nigerian univs (in their glorious semi-archaic[i]ness[/i] usually focus on these two and not on C, vb etc.

Darkmanzero, I feel you, looking in a pragmatic and profitability view, but there will always be (a lot of) ppl for legacy code, just like there will always be ppl that still buy customized 1980s caddilacs, etc. But, ofcourse, the key-word is customized .

What I'll say is that (I dont know much oh!) Ishmael shud focus on c, vb, java demdem with peripheral view on cobol/fortran.

You gotta agree with me, thats a nice bargain

Majority of the very large companies such as banks, insurance companies, etc.  have a lot of COBOL that you will not believe it! These are companies that process a lot of data and there is no programming language that does it as well as COBOL. These applications have been out there functioning for years, so there is no reason to rewrite it. (If it ain't broke, don't fix it!).

What these companies are doing is adding a web front to their applications. But their mission-critical data still resides on the mainframe. The last three clients that have worked, all have web fronts accessing data on the mid range, and this data is propagated to the mainframe. I have worked at a couple of major insurance companies. These are companies that have thousands of developers, of which 90% are mainframe developers.

As I am typing this, I am at a clients in downtown Chicago, another very big Health Insurance company, a good majority of the consultants in this room are doing mainframe. As anyone in this field  will tell you, having both mainframe and Object-Oriented Programmong (OOP) skills is a plus. Check out monster.com for OOP related jobs, on a lot of them they always specify that mainframe skills is a plus. The bottom line is, there is a lot of mainframe code out there, and it's not going anywhere.