vAbazarov@dAnai.com ("Victor Bazarov") wrote in message news:<uni9a4agiq92e5@corp.supernews.com>...
>
> Both times where you say it's "1" (digit one) and should be "i"
> (the ninth letter of English alphabet), it already is.  If your
> i's are really not as sharp as they used to be, you could simply
> increase the zoom factor in your Acrobat reader.
>
> Although, I am sure that the community at large does appreciate
> your effort.
>
> Victor
> --
> Please remove capital A's from my address when replying by mail
>

Mr. Bazarov you are completely correct.  Damn these i's anyway!  :-)

Randy.

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

On Fri, 6 Sep 2002 22:03:34 +0000 (UTC), vAbazarov@dAnai.com ("Victor
Bazarov") wrote:

>"Randy Maddox" <rmaddox@isicns.com> wrote...
>> Replacement text for subclause 24.1 paragraph 5.  The sentence
>> beginning on the third line of this paragraph says:  "Values of an
>> iterator 1 for which the expression *1 is defined are called
>> dereferenceable."  I believe that both of the '1' characters should be
>> 'i' instead.

>Both times where you say it's "1" (digit one) and should be "i"
>(the ninth letter of English alphabet), it already is.  If your
>i's are really not as sharp as they used to be, you could simply
>increase the zoom factor in your Acrobat reader.

A very humble suggestion to make reading more easy: use "iter" instead
of "i" when writing about iterators.

Regards,
Natale Fietta

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

Balog> If the new text is to expect, when is it due?

See http://www.research.att.com/~ark/c++std/2002/pdf/revisions.pdf,
where you will find a complete (unofficial) list of the changes to the
C++ standard.  It is 294 pages long, but the length is misleading
because it contains the complete old and new text of every paragraph
that contains any change, however small.

This list is generated by a program (slightly more than 700 lines
of C++) from the source text of the original and revised standards,
so it is as accurate as the program is correct.

So there are probably errors in the list, which is one of the reasons
that it is unofficial.  Nevertheless, I am not presently aware of any
inaccuracies, and I would appreciate it if you tell me about any that
you find.

--
Andrew Koenig, ark@research.att.com, http://www.research.att.com/info/ark

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

On Thu, 5 Sep 2002 18:52:55 +0000 (UTC), ark@research.att.com (Andrew
Koenig) wrote:

>Balog> If the new text is to expect, when is it due?
>
>See http://www.research.att.com/~ark/c++std/2002/pdf/revisions.pdf,
>where you will find a complete (unofficial) list of the changes to the
>C++ standard.  It is 294 pages long, but the length is misleading
>because it contains the complete old and new text of every paragraph
>that contains any change, however small.

This is the document that will lead to the TC2, right? If so, what
about adding a mark beside paragraphs that have undergone non-trivial
changes in consequence of an accepted DR? The fact that there was a DR
means often that the wording was not clear or that it addresses a
particularly delicate question. A special symbol would be a signal to
the reader that an important issue is hidden behind that wording and
help compiler implementers to identify points where language semantics
have been changed or better specified.

While we are at it, and considered that proceedings for the new
standard will begin soon, I would also like to encourage/ask for a new
policy as it concerns the clarity of the standard. Consider e.g. the
proposed resolution for DR 201: it relies on a phrase in 1.9/12 to
which no reference (not even with a note) is made from 12.2/4. I think
it would greatly help clarity if, first of all, a link/note to 1.9/12
was added and then if that paragraph was tagged with a special mark.
The standard is often written as if they had to economize the number
of words, and examples are rare, even in situations where I'm sure a
*lot* of examples where considered in the discussion phase. This seems
to me like writing a program and then stripping all the comments from
it. Regulars of this group frequently hears that this is so because
the standard is not a "tutorial"; but the purpose of notes and
examples would not be to make it a tutorial but to make it clearer and
maintainable. Just for the same reason we add comments to programs.
Why we have a standard if not even expert and competent compiler
writers can understand it? Of course we know there are vendors who
don't want compliance, but there are vendors who simply make mistakes
because a paragraph was ambiguous or even, say, obfuscated. You know
that we could do a lot of examples in this regard. A recent one is the
thread "Cast forms and reference initializations" started by Scott
Meyers. We have top-level experts like Francis Glassborow, John
Potter, Gabriel Dos Reis and Scott Meyers himself discussing on the
validity of a simple cast. Doesn't this mean that there's a problem
(in the standard, not in the experts, of course)?

The terminology, also, seems to be chosen exactly to mislead:
structures (often called structs) are classes defined with the keyword
struct, but POD-structs are not necessarily structs because they may
be defined with the class-key class. This also means that "POD struct"
(without dash) has a different meaning from "POD-struct". As it
concerns dashes, we have the following:

  (15/1) "[Note: within this clause =93try block=94 is taken
          to mean both try-block and function-try-block. ]"


A further example concerns "unions". See for instance:
http://groups.google.com/groups?selm=3Dhotaeuct0gtnpkpqlqqsiri74jcb54dqih=
%404ax.com


I want to emphasize that I'm a great fun of C++ and I have a deep
respect for the work of the committee. But I think something has to be
made in this regard. I'm quite confident in your competence and your
conscientiousness to believe that you will take this post as a honest
and non-polemic remark from an amateur.


>[...] there are probably errors in the list, which is one of the reasons
>that it is unofficial.  Nevertheless, I am not presently aware of any
>inaccuracies, and I would appreciate it if you tell me about any that
>you find.

I have just taken a look, and scrolling through the document I noticed
an error in the section "Replace subclause 17.4.1.2, paragraph 2":
with the addition of <strstream> in the table the headers are, of
course, 33 and not 32 as the text states.


Genny.

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

ark@research.att.com (Andrew Koenig) wrote in message news:<yu99d6rse0h2.fsf@europa.research.att.com>...
> Balog> If the new text is to expect, when is it due?
>
> See http://www.research.att.com/~ark/c++std/2002/pdf/revisions.pdf,
> where you will find a complete (unofficial) list of the changes to the
> C++ standard.  It is 294 pages long, but the length is misleading
> because it contains the complete old and new text of every paragraph
> that contains any change, however small.
>
> This list is generated by a program (slightly more than 700 lines
> of C++) from the source text of the original and revised standards,
> so it is as accurate as the program is correct.
>
> So there are probably errors in the list, which is one of the reasons
> that it is unofficial.  Nevertheless, I am not presently aware of any
> inaccuracies, and I would appreciate it if you tell me about any that
> you find.

Mr. Koenig, many thanks for posting this revision list.  It is most
interesting reading.  As per your request, here are a couple of things
that I have noted thus far.

Replacement text for subclause 6.4 paragraph 1 reads:  "Thus after the
if statement 1 is no longer in scope."  I believe the '1' should be
'i' to agree with the example.

Several of the replacement texts are not discernably different from
the original.  For example, the replacement text for subclause 24.1
paragraph 2 appears to be identical to the original text.  Perhaps
there are whitespace differences?  Or perhaps these old eyes are just
not as sharp as they used to be...

Replacement text for subclause 24.1 paragraph 5.  The sentence
beginning on the third line of this paragraph says:  "Values of an
iterator 1 for which the expression *1 is defined are called
dereferenceable."  I believe that both of the '1' characters should be
'i' instead.

Finally, as noted by Alan Stokes in the posting "Library defect report
208 is wrong; can it be fixed?" the final sentence in the replacement
text for subclause 24.1 paragraph 5 seems to leave open the
possibility that a past-the-end iterator may be singular.  It seems
that this would make the value returned by end() less than useful
since it could no longer serve for comparisons.

Thanks again.

Randy.

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

"Randy Maddox" <rmaddox@isicns.com> wrote...
> ark@research.att.com (Andrew Koenig) wrote in message
news:<yu99d6rse0h2.fsf@europa.research.att.com>...
> > Balog> If the new text is to expect, when is it due?
> >
> > See http://www.research.att.com/~ark/c++std/2002/pdf/revisions.pdf,
> > where you will find a complete (unofficial) list of the changes to the
> > C++ standard.  It is 294 pages long, but the length is misleading
> > because it contains the complete old and new text of every paragraph
> > that contains any change, however small.
> >
> > This list is generated by a program (slightly more than 700 lines
> > of C++) from the source text of the original and revised standards,
> > so it is as accurate as the program is correct.
> >
> > So there are probably errors in the list, which is one of the reasons
> > that it is unofficial.  Nevertheless, I am not presently aware of any
> > inaccuracies, and I would appreciate it if you tell me about any that
> > you find.
>
> Mr. Koenig, many thanks for posting this revision list.  It is most
> interesting reading.  As per your request, here are a couple of things
> that I have noted thus far.
>
> Replacement text for subclause 6.4 paragraph 1 reads:  "Thus after the
> if statement 1 is no longer in scope."  I believe the '1' should be
> 'i' to agree with the example.
>
> Several of the replacement texts are not discernably different from
> the original.  For example, the replacement text for subclause 24.1
> paragraph 2 appears to be identical to the original text.  Perhaps
> there are whitespace differences?  Or perhaps these old eyes are just
> not as sharp as they used to be...

"template function" is replaced by "function template".  And it
has really nothing to do with your eyes, I am afraid.

> Replacement text for subclause 24.1 paragraph 5.  The sentence
> beginning on the third line of this paragraph says:  "Values of an
> iterator 1 for which the expression *1 is defined are called
> dereferenceable."  I believe that both of the '1' characters should be
> 'i' instead.

Both times where you say it's "1" (digit one) and should be "i"
(the ninth letter of English alphabet), it already is.  If your
i's are really not as sharp as they used to be, you could simply
increase the zoom factor in your Acrobat reader.

Although, I am sure that the community at large does appreciate
your effort.

Victor
--
Please remove capital A's from my address when replying by mail


---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

After the defect reports (or some subset of them) are considered and closed,
will we have a new standard text, with all the relevant sections consolidated?
Or everyone has to do the merge for himself, and we'll see a new text only as the C++0x waiting the decade?

If the new text is to expect, when is it due?

Paul

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

Balog> After the defect reports (or some subset of them) are
Balog> considered and closed, will we have a new standard text, with
Balog> all the relevant sections consolidated?

We will have a new standard text.

Balog> If the new text is to expect, when is it due?

I delivered what I hope is the final copy this past Wednesday.

ISO will publish it when they publish it, either late this year or
early next year.

Meanwhile, I will put a complete (unofficial) list of changes on my
website, possibly as early as next week.

--
Andrew Koenig, ark@research.att.com, http://www.research.att.com/info/ark

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]