Home > Archive > Fortran > June 2005 > Fortran 90/95 standard?
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Fortran 90/95 standard?
|
|
| Bart Vandewoestyne 2005-06-09, 3:58 pm |
| When submitting bug-reports for g95, I have already often felt the need
to look up certain things in the Fortran standard, but unfortunately I
don't have a copy of that standard and the standard + all it's
extensions are quite expensive to buy for an invidivual...
Now I would like to know what documents I can get that are 1) legally
downloadable/usable and 2) as close to the official standard as it can
get (latest versions).
At http://webstore.ansi.org I already found
* draft 'ISO/IEC 1539-1' from 23 september 2002 which seems to be freely
available for download.
But apparently these documents are not freely available from there:
ISO/IEC 1539-1:2004
ISO/IEC 1539-2:2000
ISO/IEC 1539-3:1999
ISO/IEC TR 15580:2001
ISO/IEC TR 15581:2001
ISO/IEC TR 19767:2005
Are there any other places where i can get the 'most recent' drafts
legally and for free so i can at least look up certain things before I
send bug-reports or ask questions here in the group? ;-)
Regards,
Bart
--
"Share what you know. Learn what you don't."
| |
|
|
| beliavsky@aol.com 2005-06-09, 3:58 pm |
| Bart Vandewoestyne wrote:
> When submitting bug-reports for g95, I have already often felt the need
> to look up certain things in the Fortran standard, but unfortunately I
> don't have a copy of that standard and the standard + all it's
> extensions are quite expensive to buy for an invidivual...
The Fortran committee members, who are unpaid, do almost all the work
in creating a standard. I have wondered what ISO does to merit the
large fees for copies of the standard.
My main Fortran reference is the "Fortran 95 Handbook", written by
several committee members. An F2003 edition is in the works.
| |
| Richard E Maine 2005-06-09, 3:58 pm |
| In article <1118318479.285139.279910@z14g2000cwz.googlegroups.com>,
"Joost" <jv244@cam.ac.uk> wrote:
> I have been looking often at this draft standard:
>
> http://j3-fortran.org/doc/standing/...df/97-007r2.pdf
That is the version that the committee itself uses, so you aren't likely
to go far wrong, technically, except...
> However, I'm not quite sure I understand, for example, in which
> document I could find e.g. the edits to the standard that are mentioned
> in the Fortran interprations (e.g.
> http://j3-fortran.org/doc/standing/2002/02-006ar1.txt).
There isn't a version of the standard, even in draft form, which has the
edits from the interps actually applied. What is actually published is
not a revised version of the whole standard, but instead a corrigendum
(standard-speak for errata), which is nothing but a collected list of
the edits. No explanations or anything (explanations are in the 006
document that you cite above).
The final drafts (identical in content to what was published) of f95
corrigendum 1 and 2 are WG5 documents N1421 and N1472, which can be
found on the WG5 web site <http://www.nag.co.uk/sc22wg5/>. I was
thinking there was a third f95 corrigendum, but a quick glance at where
I normally keep my copies doesn't show it. Maybe I'm confusing it with
f90, which did have 3. Or maybe I forgot to stash a copy there; I didn't
go back and check the various meeting papers.
Anyway, the number of WG5 papers isn't excessive. If you paw around in
the documents directory of the wg5 site (look at the Index.txt summary
files), you should be able to find the latest drafts of all of the
corrigenda and TRs. On the TRs, my notes say that N1379 is the latest
thing I have for the 2nd edition of the allocatable TR, and N1378 is the
latest I have for the IEEE one, though N1376 implies one small extra
change. N1602 is the last draft of the modules TR (to f2003).
Oh yes, and the version of f2003 that the committee uses for its own
work is 04-007 from the J3 site. (Sorry that I didn't bother to turn
these into handy clickable urls, but you should be able to find them on
the site and I'm feeling lazy... or maybe it is that there was no coffee
in the pot when I checked this morning. Got to go check again. :-))
As far as legality... Well, I'll give the data... make your own
judgment. The 97-007r2 document cited above is what the committee uses
itself when doing such things as processing interpretations of f95.
Likewise with 04-007 and f2003. They are freely available for anyone
helping the committee work on the standard. You don't have to be a
formal member or anything. As far as I am concerned (but I have no
official say in it), you are involved in helping the committee work on
the standard if you are using the standard in your programming work.
Maybe you'd find a problem in the standard (that perfectly well includes
such problems as features that you think appropriate for a future
version, for example). Or maybe you wouldn't find anything, but the
possibility is enough.
I doubt you will offend anyone on the committee (ANSI/ISO bureaucrats I
can't speak for). As noted, committee members work for "free", which is
actually inaccurate. Committee members have to pay ANSI to be on the
committee. :-(
I haven't checked on f2003, but an unquestionably legal electronic copy
of f95 used to be available from the ansi store for only about $35. They
wanted something ludicrous for a paper copy, but if you looked further,
the electronic one was more reasonable. It wasn't the best job (ISO
never got electrons for f95; instead they scanned the paper. I wish they
had asked me so that I could send them "better" electrons, but
coordination between ISO and editors+committees has long been poor; once
they get the final draft for publication, they don't seem to want to
even answer emails any more.)
--
Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain | experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain
| |
| Bart Vandewoestyne 2005-06-09, 3:58 pm |
| In article <1118322041.804074.242460@o13g2000cwo.googlegroups.com>, beliavsky@aol.com wrote:
>
> The Fortran committee members, who are unpaid, do almost all the work
> in creating a standard. I have wondered what ISO does to merit the
> large fees for copies of the standard.
True, but since it might have its reasons and I don't want to go into a
kind of political discussion, i'll skip your first remark ;-)
> My main Fortran reference is the "Fortran 95 Handbook", written by
> several committee members. An F2003 edition is in the works.
My main references are the 'Fortran X explained' books and I find them
very useful, but sometimes i would like to read what the *real* standard
says...
Regards,
Bart
--
"Share what you know. Learn what you don't."
| |
| wclodius@lanl.gov 2005-06-09, 3:58 pm |
|
Richard E Maine wrote:
><snip>
> The final drafts (identical in content to what was published) of f95
> corrigendum 1 and 2 are WG5 documents N1421 and N1472, which can be
> found on the WG5 web site <http://www.nag.co.uk/sc22wg5/>. I was
> thinking there was a third f95 corrigendum, but a quick glance at where
> I normally keep my copies doesn't show it. Maybe I'm confusing it with
> f90, which did have 3. Or maybe I forgot to stash a copy there; I didn't
> go back and check the various meeting papers.
> <snip>
The last sets of votes on corrigenda have been confusing, because some
of decisions about F95 interpretations got delayed so long that F03 was
the official standard. I sometime follow the email, but in what I read
to my feeble brain it wasn't clear if a separate F95 corrigenda could
be published, or that an F03 corrigenda could include
1. items that were fixed in that standard
2. discussion of differences between F03 and F95 where the F95
interpretation is part of anupublished corrigendum
3. Edits for the F95 standard.
| |
| Richard E Maine 2005-06-09, 8:58 pm |
| In article <1118338121.261090.49130@g14g2000cwa.googlegroups.com>,
wclodius@lanl.gov wrote:
> The last sets of votes on corrigenda have been confusing,
Yes, I found them confusing also. In fact, I voted "no" on a whole bunch
of them because I thought they weren't sufficiently explicit about which
standard they applied to. But you were probably talking about the "last"
sets of votes before the Delft meeting of a few w s ago. (Anyway,
that's what I found confusing). That was not actually votes on a
corrigenda on any standard. It was about stuff planned to go in a
corrigendum, but there was not actually a draft corrigendum being voted
on.
From reading the minutes of the Deft meeting, I see that a corrigendum
draft was prepared (N1641). It is quite clearly and explicitly a
corrigendum to f2003. I don't think it is quite "official" yet until ISO
says so, but it is probably at least close.
I see no mention about an f95 corrigendum in the resolutions of the
meeting, so I don't know the formal status of those interp items that
"passed" but were directed at f95. I'm at least a bit suspicious that
they don't have and won't get the imprimatur of "official" status from
ISO, but just as something that passed both committees.
--
Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain | experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain
| |
| robert.corbett@sun.com 2005-06-10, 8:57 am |
|
Bart Vandewoestyne wrote:
> My main references are the 'Fortran X explained' books and I find them
> very useful, but sometimes i would like to read what the *real* standard
I have yet to see a book on Fortran other than the standard that
explains
the tricky parts of the language.
Bob Corbett
|
|
|
|
|