petition to retire Fortran
E. Robert Tisdale
Brian Salter-Duke
How do I vote against the petition?
--
Brian Salter-Duke Humpty Doo, Nr Darwin, Australia
My real address is b_duke(AT)octa4(DOT)net(DOT)au
Use this for reply or followup
Rich Townsend
You're sleeping on the job, E. Bob; this troll was posted to c.l.f a few
days ago. See the thread entitled "Re: A petition to J3 apropos
FORTRAN's future".
Gary Strand
> PETITION TO RETIRE FORTRAN:
>
> http://www.fortranstatement.com/
Puhleeze.
Go away.
--
/\ Gary Strand (303) 497-1336 NCAR ML220
\_][ www.cgd.ucar.edu/ccr/strandwg 1850 Table Mesa Dr
\___stra...@ucar.edu Boulder, Colorado, USA 80305-3000
Stewart Gordon
<snip>
> If you do not like Fortran 77/90/95, don't use it.
<snip>
Some people, myself included, are made to use it as that's what the
department uses.
I think the idea is that the sooner an official retirement comes about,
the sooner these departments are eventually going to think "this is an
ancient language, it's time to move on" and stop inflicting this
archaism on their professors/students/whomever.
Stewart.
--
My e-mail is valid but not my primary mailbox, aside from its being the
unfortunate victim of intensive mail-bombing at the moment. Please keep
replies on the 'group where everyone may benefit.
Paul van Delst
>
> Madhusudan Singh wrote:
>
> <snip>
> > If you do not like Fortran 77/90/95, don't use it.
> <snip>
>
> Some people, myself included, are made to use it as that's what the
> department uses.
You're *made* to use it? That doesn't sound like a very enlightened (or productive)
attitude for a university department...even if you thought Fortran was the best thing
since sliced bread.
> I think the idea is that the sooner an official retirement comes about,
> the sooner these departments are eventually going to think "this is an
> ancient language, it's time to move on" and stop inflicting this
> archaism on their professors/students/whomever.
It sure sounds like you have a weird setup there. Aren't the professors the ones that
decide what gets inflicted on the students?
If you're talking about Fortran-77 only, I would mostly agree. If the department is
forcing[*] students (and professors/whomever) to use Fortran-77, and only Fortran-77, then
yes I also think they're doing a disservice to their staff and students. At the same time,
depending on your field, there is a *lot* of Fortran-77 code "out there" that is still
used heavily, so the instruction is not a total write-off.
If, however, the department is advocating use of Fortran-90/95, I don't see the problem at
all.
Referring to a Fortran -- which is a computing language under active development (whether
or not one approves of the development path) -- as an "ancient language" or an "archaism"
exposes your ignorance... or stridency. But fear not... I feel exactly the same way about
C/C++! ;o)
cheers,
paulv
[*] If you tell the Dept you want to use <insert non-Fortran programming language of
choice> because you will be more productive, what would they do? Toss you out?
--
Paul van Delst
CIMSS @ NOAA/NCEP/EMC
Dan Tex1
>Date: 2/27/04 9:20 AM Pacific Standard Time
>Message-id: <c1nu84$991$1...@sun-cc204.lut.ac.uk>
>
>Madhusudan Singh wrote:
>
><snip>
>> If you do not like Fortran 77/90/95, don't use it.
><snip>
>
>Some people, myself included, are made to use it as that's what the
>department uses.
>
>I think the idea is that the sooner an official retirement comes about,
>the sooner these departments are eventually going to think "this is an
>ancient language, it's time to move on" and stop inflicting this
>archaism on their professors/students/whomever.
The problem with such logic is that they NEED something to replace the Fortran
with. Thus far... no one appears to have come up with a very suitable ( or
rather, acceptable ) replacement language. Heck, a lot of people who wish to
see Fortran go away think that Java would be a good language to use instead.
There's nothing wrong with Java. But, Fortran kicks butt in some important
areas where Java doesn't even have a butt.
Why replace a language that serves it's purpose with a language which isn't as
capable of serving the same purpose??
>Stewart.
Rob
news:c1nu84$991$1...@sun-cc204.lut.ac.uk...
> Madhusudan Singh wrote:
>
> <snip>
> > If you do not like Fortran 77/90/95, don't use it.
> <snip>
>
> Some people, myself included, are made to use it as that's what the
> department uses.
That's a problem with your department, or with you for not being
able to muster a rational argument to use something else.
There are some application areas where Fortran is better than
other languages, and some where it's not. If your department works
with one of those, they may be justified in their position. If your work
suits another language better, then *you* need to justify an
exception.
>
> I think the idea is that the sooner an official retirement comes about,
> the sooner these departments are eventually going to think "this is an
> ancient language, it's time to move on" and stop inflicting this
> archaism on their professors/students/whomever.
>
That argument won't wash as age of a language does not change the
fact it is still better for some applications.
And Fortran is still active: there have been extensive modifications
to Fortran in recent times, as reflected in F90 and F95 standards
and the effort to evolve the language continues.
While I agree a case may be made for moving away from F77,
I wouldn't describe F95 as "archaic". None of the languages are
perfect (no language is!) but there still exists compelling reasons
for using older languages. The most common one is an existing
code base in (say) F77 and a high cost of modifiying it for another
language and then verifying it.
Rob
"Madhusudan Singh" <spammers...@yahoo.com> wrote in message
news:c1p8q8$1le47o$1...@ID-159130.news.uni-berlin.de...
> [Snip]
I agree with your other comments (in fact I've said something similar in
a parallel post). But
> Further, if fortran 95 or the forthcoming f2003 is "archaic", I wonder
what
> does that make ANSI C which was invented in early 1970's.
This statement is wildly inaccurate. C was standardised by ANSI (and ISO)
in 1989, and the standard has evolved: there is another standard dated
1999.
Precursors of ANSI/ISO C (eg K&R C) date back to the 70s, but
their relationship to standard C is more or less equivalent to the
relationship between FORTRAN IV and modern versions of
Fortran.
Greg Lindahl
Rob <nos...@nonexistant.com> wrote:
>Precursors of ANSI/ISO C (eg K&R C) date back to the 70s, but
>their relationship to standard C is more or less equivalent to the
>relationship between FORTRAN IV and modern versions of
>Fortran.
Congratulations, that was the point that the posting you replied to
was making.
-- greg
Dan Tex1
>news:c1nu84$991$1...@sun-cc204.lut.ac.uk...
>> Madhusudan Singh wrote:
>>
>> <snip>
>> > If you do not like Fortran 77/90/95, don't use it.
>> <snip>
>>
>> Some people, myself included, are made to use it as that's what the
>> department uses.
That's a lot like quite a few engineers I know of who "have" to use C++ where
I currently work.
>That's a problem with your department, or with you for not being
>able to muster a rational argument to use something else.
>
>There are some application areas where Fortran is better than
>other languages, and some where it's not. If your department works
>with one of those, they may be justified in their position. If your work
>suits another language better, then *you* need to justify an
>exception.
>
>>
>> I think the idea is that the sooner an official retirement comes about,
>> the sooner these departments are eventually going to think "this is an
>> ancient language, it's time to move on" and stop inflicting this
>> archaism on their professors/students/whomever
By that EXACT same logic, it can easily be argued that C and C++ should be
retired. You know... so that some companies and schools can move on and stop
inflicting their archaicsm on others.
The only problem is... that logic just isn't very logical. It's not a good
argument for getting rid of C and C++ and it's not a good argument for getting
rid of Fortran either.
Dan :-)
Stewart Gordon
<snip>
> It sure sounds like you have a weird setup there. Aren't the
> professors the ones that decide what gets inflicted on the students?
They probably inflict it on each other.
> If you're talking about Fortran-77 only, I would mostly agree. If the
> department is forcing[*] students (and professors/whomever) to use
> Fortran-77, and only Fortran-77, then yes I also think they're doing
> a disservice to their staff and students. At the same time, depending
> on your field, there is a *lot* of Fortran-77 code "out there" that
> is still used heavily, so the instruction is not a total write-off.
>
> If, however, the department is advocating use of Fortran-90/95, I
> don't see the problem at all.
My dept is mostly on F90, so it isn't too bad. Though there is still
quite a bit of legacy F77 code lurking around. Notice that I wasn't
complaining, only trying to infer the intention of the petitioner.
<snip>
> [*] If you tell the Dept you want to use <insert non-Fortran
> programming language of choice> because you will be more productive,
> what would they do? Toss you out?
I already was using one, namely C++. But my bosses have been suggesting
to me for a long time that I should write a F90 version. After all, it
does interface with external F90 code.
I think they're saying using F90 is in itself a way of being more
productive, as:
- my C++ code had got rather over-complex with its use of OOP and that
- it's expected to go into a departmental code repository, for which F90
is the standard
- because it's the department standard, other people'll easily be able
to read/understand/help debug the code.
The main strength I can see in F90 is built-in vector arithmetic. I'm
not sure how many other languages have it. One that has it specified is
D <http://www.digitalmars.com/d/>, but:
- that's still in alpha stage
- its vector arithmetic hadn't yet been implemented last time I looked
- it doesn't yet have a Mac implementation
so my department's unlikely to switch to it in the near future. But I
wonder how well D would serve as a scientific programming language....
Helge Avlesen
| so my department's unlikely to switch to it in the near future. But I
| wonder how well D would serve as a scientific programming language....
yet another family member with the brace disease. it is now at least
the fourth attempt to make multidimensional arrays in this language
family, and the users still have to say a[i][j][k] instead of
a[i,j,k]. better luck next time I guess...
--
Helge
Paul van Delst
>
> Paul van Delst wrote:
>
> <snip>
> > It sure sounds like you have a weird setup there. Aren't the
> > professors the ones that decide what gets inflicted on the students?
>
> They probably inflict it on each other.
:oD
<snip>
> My dept is mostly on F90, so it isn't too bad. Though there is still
> quite a bit of legacy F77 code lurking around. Notice that I wasn't
> complaining, only trying to infer the intention of the petitioner.
Yeah I know, but I was feeling indignant on your behalf because of how your dept was
treating you!
> <snip>
> > [*] If you tell the Dept you want to use <insert non-Fortran
> > programming language of choice> because you will be more productive,
> > what would they do? Toss you out?
>
> I already was using one, namely C++. But my bosses have been suggesting
> to me for a long time that I should write a F90 version. After all, it
> does interface with external F90 code.
>
> I think they're saying using F90 is in itself a way of being more
> productive, as:
>
> - my C++ code had got rather over-complex with its use of OOP and that
> - it's expected to go into a departmental code repository, for which F90
> is the standard
> - because it's the department standard, other people'll easily be able
> to read/understand/help debug the code.
I think they're all facetious arguments (on your depts part). As for the over-complex
part, I reckon that's a subjective statement. The stuff I do in Fortran90 is brain-dead
simple, but when people look at my code they think it's complex. If there is a design
problem with your code, then switching to another language won't fix that. Although if you
treat your current version as a prototype and decide to rewrite from scratch with an
updated design, then maybe a f90 only version is a more useful thing. It can be a tough
call.
My original reply was responding to my perception that someone in your dept has a bee in
their bonnet about certain things being "just so". Your department should be happy to have
a competent C++ programmer.....diversity is a good thing, right? :o) It might be more
productive (on both your and your dept's part) for you to spend some time putting together
an interface (API? Dunno what the correct term is) to your code so that fortran-only
programmers can use it. ala the f90 interface to the netCDF library. The netCDF stuff is
all in C, but I use the f90 interface exclusively.
> The main strength I can see in F90 is built-in vector arithmetic.
I like Fortran because it's simple to read! :o) When I see C/C++ source code, my eyes
glaze over. But, that doesn't stop me from using useful C/C++ libraries that have f90
interfaces (when they exist).
Anyway.... apologies for the blathering.
cheers,
paulv
Greg Lindahl
Stewart Gordon <smjg...@yahoo.com> wrote:
> Please keep replies on the 'group where everyone may benefit.
I know that Usenet etiquette is dead and gone, but I would like to
encourage everyone to send private mail when it makes sense, and post
when it makes sense.
-- greg
Les
>
> <snip>
> > If you do not like Fortran 77/90/95, don't use it.
> <snip>
>
> Some people, myself included, are made to use it as that's what the
> department uses.
>
> I think the idea is that the sooner an official retirement comes about,
> the sooner these departments are eventually going to think "this is an
> ancient language, it's time to move on" and stop inflicting this
> archaism on their professors/students/whomever.
>
There could be a positive side to this.
It's an additional skill for your cv.
I got my present job (almost a year ago) because of having experience in
Fortran. And now I have the opportunity, with the same company, to extend my
C/C++ skills (into MFC for example). An attitude of adaptability, a
willingness to take on new skills, looks good to many exployers I'm sure.
BTW my next big project will involve working with an ancient, archaic
language, which (to me) has a very difficult syntax and structure - chinese!
I wonder who would be willing to suggest that it be retired :-)
Les
Stewart Gordon
> In article <c1nu84$991$1...@sun-cc204.lut.ac.uk>, Stewart Gordon
> <smjg...@yahoo.com> wrote:
>
>> Please keep replies on the 'group where everyone may benefit.
>
> I know that Usenet etiquette is dead and gone,
Where has it gone? The same place as English spelling, grammar and
punctuation?
> but I would like to encourage everyone to send private mail when it
> makes sense, and post when it makes sense.
Not sure how that's a "but", but I generally post stuff to which it's
most unlikely that sending private mail'll make sense.
Stewart.
--
My e-mail is valid but not my primary mailbox, aside from its being the
unfortunate victim of intensive mail-bombing at the moment. Please keep
Stewart Gordon
<snip>
> There could be a positive side to this.
> It's an additional skill for your cv.
<snip>
Let's just wait for a replacement language to come out. Then you'll
have one more skill to add.
Greg Chien
> BTW my next big project will involve working with an ancient,
> archaic language, which (to me) has a very difficult syntax and
> structure - chinese! I wonder who would be willing to suggest
> that it be retired :-)
How can one retire that? It is the most "popular" language :-)
--
Best Regards,
Greg Chien
e-mail: remove n.o.S.p.a.m.
http://protodesign-inc.com
Rob
> On Saturday 28 February 2004 21:45, Rob (nos...@nonexistant.com) held
forth
> in comp.lang.fortran (<40415...@news.iprimus.com.au>):
>
>
> > This statement is wildly inaccurate. C was standardised by ANSI (and
> > ISO)
> > in 1989, and the standard has evolved: there is another standard dated
> > 1999.
> >
> > Precursors of ANSI/ISO C (eg K&R C) date back to the 70s, but
> > their relationship to standard C is more or less equivalent to the
> > relationship between FORTRAN IV and modern versions of
> > Fortran.
>
>
> 1989 still predates Fortran 95 etc. So the original logic of questioning
my
> interlocutor's comments about Fortran being "archaic" is still valid.
I understood (and agree) with you querying Fortran being
described as archaic. My concern was that your example potentially
conveyed an incorrect impression that C is archaic.