Re: Challenge 4 (matrix diagonal)

From: Richard Maine (nospam_at_see.signature)
Date: 12/29/03 

Date: Mon, 29 Dec 2003 08:56:19 -0800

"Gary L. Scott" <garyscottNOSPAM@ev1.net> writes:
> Dan Nagle wrote:
>>
>> On Sun, 28 Dec 2003 21:58:35 GMT, "David Frank"
>> <dave_frank@hotmail.com> wrote:

>> >Below story says it [F2003] likely will be approved in Mar. 2004
>> >
>> >http://sdtimes.com/news/092/story9.htm
>>
>> The story is wrong. What happens in March?
>
> Certainly not a very well written or thorough article.
> Hard to imagine what the purpose of it was.

Sure isn't. Since it has a pointer to the NAG compiler at the bottom,
my wild guess is that it is a press release/product announcement
disguised as an article. Probably created by some collusion between
NAG's marketing department and the hacks of sdtimes; hard to guess
where most of the "blame" lies. Has the typical flavor of such
poorly disguised ads.

Odd collection of features to emphasize. VOLATILE gets the largest
number of words of any feature; that wouldn't have made the list of
the features I'd have even named in an article that short. C interop,
which I consider the most important thing, isn't even mentioned.

The descriptions of IMPORT and ISO_FORTRAN_ENV have been "dumbed down"
so much as to be completely wrong. ISO_FORTRAN_ENV is not a keyword
and does have anything even vaguely to do with "data stored elsewhere
on the host computer". And while I pushed to have the command line
argument stuff and get_environment_variable put into ISO_FORTRAN_ENV,
I lost, so it doesn't have anything to do with them either. That
makes all 3 things said about ISO_FORTRAN_ENV completely wrong - not
just incomplete and imprecise as you'd expect from such a short
article, but completely wrong and not even hinting at the actual
functionality. If you wonder why I pick on such a relatively
small thing as ISO_FORTRAN_ENV, it isn't because all the more
significant features were well described; it is because ISO_FORTRAN_ENV
came in second behind VOLATILE on the list of things getting
the most words in the article.

Glad to see that the IEEE and allocatable TRs are "long-used".
My editorial ear winces at the poor writing (or was it poor
understanding?) that manages to completely invert the meaning
of one important feature. No, the IEEE TR is not about
exceptions *TO* the IEEE standard. It is about the exceptions
*IN* the IEEE standard. I'm afraid that one two-letter
word changes the sentence to mean pretty much its opposite.

And, back to what it was cited for...I'd say that whoever wrote
the part about the schedule for the standard fell into the category
of knowing just enough to be dangerous - as in managed to sound
knowledgable without actually getting a single thing correct.

The final draft has not yet been released. Yes, I know that FCD
stands for "final committee draft", which includes the word "final".
Confusing ISO terminology, but it is still a committee draft, which
isn't the last kind of draft. The final draft is a DIS (Draft
International Standard), which isn't even scheduled for release
until...sometime soon after the May meeting, I forget the exact date
but probably June or so.

J3 doesn't have anything to do with ratifying the standard; that
would be WG5. While possibly a small detail at some level, it is
quite relevant here in that WG5 doesn't meet in March. The March
J3 meeting is sort of unfortunately timed and won't be able to
do much of anything official on F2003. The US position is due
before then, and official actions on comments can't be done
without WG5. Likely some work on proposed actions can be done,
but nothing can be "ratified". Looks to me like the writer
knew that the next J3 meeting would be in March and assumed that
therefore that's when the standard would be ratified. He sure
can't have gotten that from any published material or knowledgable
source. I'd surmise that he just assumed it. He's wrong, and not
even particularly close. 

-- 
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

Relevant Pages

  • Re: Ancient history
    ... proposed deprecated features were just too controversial. ... I'd advise against basing much on a draft that early; ... Equivalence has never been "scheduled" for removal in any draft of any ... warning that it *MIGHT* be removed in the next standard. ...
    (comp.lang.fortran)
  • Re: struct tmx
    ... I guess the list of new features Im looking at was based on a ... draft and not the final standard then? ... The things marked C99 are the standard. ...
    (comp.lang.c)
  • Re: book on C99
    ... > Is there a book on the C language that covers the features added in ... If you want to buy a book, you can buy the standard or use the draft. ... which also covers C99. ...
    (comp.lang.c)
  • Re: WANTED: Volunteer to Scan Old Programs
    ... BASIC to the American National Standard for Minimal BASIC, ... Implementation-defined features ... Error and exception reporting ... of interpretation, and Section 5 for information peculiar to ...
    (comp.os.cpm)
  • Re: Fortran Features
    ... without a specific list of extensions. ... In order to make the features optional like this, ... will need to be included in the standard. ... means that you get into a substantial debate on each ...
    (comp.lang.fortran)