Fortran expected lifetime

[markmerritt]

Can anyone give me information regarding Intel's committment to supporting Fortran in the future? Is this language going to be around in 60 years? I work for a major aerospace company and we need to make a decision whether to migrate one of our simulations (currently written in Fortran) to a different language (Matlab/simulink), even though it works perfectly well in Fortran. If we keep it in Fortran we need to have strong confidence that Fortran will be supported for the next 60 years.

Perhaps someone can tell me how to get hold of the Fortrans sales dept to get more info about this issue.

Thanks all.

[reneeculver]

It's a none issue. Do you think any company can give you a commitment to support any language for the next 60 years?

How does Intel insure that they beare business 60 years from now?

Renee

[psantos]

Hello markmerritt,

I thing you should make the question the other way around: Will Mathworks support Matlab/simulink in the next 60 years?
I suspect no one can answer this question. Perhaps your company should update the FORTRAN code to reflect the changes of the past few years, not start rewriting the entire code to another language. I thing this is a big waste of time and resources. On the other hand, there is one major drawback with this transition. Since Matlab is an interpreted language there will be a huge loss of performance.
I can also take to your attention that very recently, a revision of FORTRAN standard was made (F2008). I think that indicates that FORTRAN is very active and isn't an obsolete programming language that is near expiration period.

Pedro

[Paul_Curtis]

If you try to predict the future based on the past, note that there has been more (and better) code written in Fortran, for far longer, than any other language, and other languages have undergone many faddish evolutions of questionable utility and which have all been comparatively short-lived.

Will you compare the customer base and market capitalizations of Matlab/Simulink vs. Intel, and then seriously wonder which company is likely to be around for longer? (I cannot speak for Intel and their continuing committment to Fortran support, but this is a really silly question.)

[Steven_L_Intel1]

If any company tells you that they will support X for 60 years, they're being dishonest. Fortran has been around for more than 50 years and shows no signs of going away. Same for COBOL. I would not even try to guess whether we'd be "writing programs" 60 years from now. (I very much doubt I'll be around to see.)

As for Intel, we have not even entertained the suggestion of dropping support for Fortran. But even if we did at some point, there are quite a few other Fortran compilers available and Fortran is extremely portable and has a better history of upward compatibility than any other programming language in common use.

My personal opinion is that your application is likely to have a longer lifetime if left in Fortran than if converted to any other language.

[dannycat]

Will your aerospace company be around in sixty years time? I'd say not ifit is run by people who expect proper answers to such questions.

I'm afraid the aerospace industry is being taken over by people who expect everyone to predict the future and then are surprised and disappointed when your prediction was not correct.

So I wouldn't worry about FORTRAN's longevity and stick with it.

[Jacob_Williams]

I can guarantee that any code you write in Matlab today will not work in 60 years... I have Matlab code from 10 years ago that doesn't work the same on the latest version of Matlab. Meanwhile, I routinely use Fortran code from 20-30 years ago that compiles and runs just fine on any of the latest fortran compilers (Intel/gfortran/g95/etc.)

[WSinc]

You should be much more concerned that YOUR company will be around for 60 years.

Anyway, FORTRAN is such a superior choice for engineering and science applications,
I seriously doubt it will disappear. I am talking about:
(1) different types and precisions of variables
(2) superior I/O handling capabililty
(3) superiorARRAY handling
(4) Efficiency of processing.

I really feel sorry for any engineer trying to write a heavy-duty application (i.e. like simulations of supernovas) in C++

Rather, I think it will continue to evolve to meet programmer needs, just like it has for the last 50 years.
If you doubt me, take a look at what it was like back in the late 50s and 60s. I actually had to
work withFORTRAN 2back then, when we were carrying around punched cards.

If Intel doesn't want to support it, they would probably sell the rights to some other company, rather
than just let it die.

Yours; Bill S.

[WSinc]

Hi Mark -

Maybe you could tellus what the simulation is (assuming it's not classified).

If it's already in Fortran, I think it would be best to upgrade the code to
take advantage of the more modern language features.

I recently woked on a package that used code from the 70s, and
it ran somewhat faster.

Yours; Bill

[martin_upsdellagrese]

Then there is the quote by Seymour Cray, John Backus and others.
"I don't know what language I will beusing in 20 years time, but I know it will be called Fortran"

[Steven_L_Intel1]

That quote was originally from C Anthony "Tony" Hoare, "I don't know what the programming language of the year 2000 will look like, but I know it will be called FORTRAN."