W
WW
Ron said:
That was in Herbs first post. And I did not want to overload his brain.
That was in Herbs first post. And I did not want to overload his brain.
T
The_Sage
Reply to article by: "Attila Feher said:
No you do. The spec clearly stated that "...its type is implementation-defined"
and you failed to explain why IBM, MS, and Borland all implemented-defined void
main() and yet still claim to be ISO compliant. Or maybe it wasn't that you
failed, but that you dodged and evaded the issue because it exposes a gap in
your claim bigger than the gap between your ears?
The Sage
=============================================================
My Home Page : http://members.cox.net/the.sage
"The men that American people admire most extravagantly are
most daring liars; the men they detest the most violently are
those who try to tell them the truth" -- H. L. Mencken
=============================================================
W
WW
The_Sage said:
I did. So I *know*. You only whish.
The spec clearly stated that it *must* return int, "otherwise its type is
implementation-defined".
Herb Sutter explained it for you.
Excatly. You failed.
I have answered your stupid claims every time. You delete them from the
answer every time. You know you are wrong. Stop trolling.
T
The_Sage
Reply to article by: "WW said:
Well?
The Sage
=============================================================
My Home Page : http://members.cox.net/the.sage
"The men that American people admire most extravagantly are
most daring liars; the men they detest the most violently are
those who try to tell them the truth" -- H. L. Mencken
=============================================================
W
WW
The_Sage said:
Please look at Chapter 28 and let me know what do you see there about void
main.
A
Attila Feher
WW said:
I mean The Sage's brain. (You have to admit I am nice to assume he has one)
D
David B. Held
The_Sage said:
I have challenged you several times to show statements by any
of these organizations claiming that their compilers are 100%
conforming. Assuming that they are and then assuming that
any code they accept is conforming C++ is very much like
assuming that every car built by the Big Three is crash-proof,
and that any road you drive such a car on is also crash-proof,
because you drove a few cars on a few roads and didn't crash.
You start out with an incorrect assumption, and then throw in
a non-sequitur. Combine that with an inability to correctly
parse entire English sentences (perhaps they overload your
stack), and it's a wonder you can even tie your shoes in the
morning (though I wouldn't be suprised if you stick with sandals
or velcro straps).
Dave
A
Attila Feher
David B. Held wrote:
[SNIP]
Hey! I have velcro straps! And sandals.
(Shoelaces somehow always break within a short time. Might be the hight of
my feet...)
#include <iostreams>
typedef int not_void;
not_void main() {
std::cout <<
"Having velcro straps and sandals"
" does not make you a moron." << //
std::endl;
}
[SNIP]
Hey! I have velcro straps! And sandals.
(Shoelaces somehow always break within a short time. Might be the hight of
my feet...)
#include <iostreams>
typedef int not_void;
not_void main() {
std::cout <<
"Having velcro straps and sandals"
" does not make you a moron." << //
std::endl;
}
N
Noah Roberts
Attila said:
I have velcro straps on my sandals
I also only tie my shoes when theundo themselves as I am too lazy to do it in the morning
NR
A
Andrew Koenig
Your interpretation of the standard is incorrect.
No, I don't mean that. If I meant that, I would have said it.
Correct.
Please read that last sentence again:
A conforming implementation may provide more versions
of main(), but they must all have return type int.
If you write "void main() { }", then main does not have return
type int. Therefore, a conforming implementation is not permitted
to accept it.
That depends on whether the books purport to describe Standard C++, or
whether they are describing the dialects accepted by particular
implementationa.
Strictly speaking, programs are neither conforming nor nonconforming.
The C++ standard does not impose requirements directly on programs; it
imposes requirements only on implementations. So if you want to be
strictly correct about it, you should not talk about conforming or
non-conforming programs, only about conforming or non-conforming
implementations.
It is also correct to say that conforming implementations are not
required to accept "void main()" -- in fact, they are required to
produce a diagostic message for such programs. This requirement is
often ignored in practice, but it's still there.
What it proves is that arguments of the form "My compiler accepts it,
so it's right" are specious.
Yes it does, in subclause 3.6.1, paragraph 2.
No they're not -- subclause 3.6.1 paragraph 2 makes that clear.
No, I don't mean that. If I meant that, I would have said it.
Correct.
Please read that last sentence again:
A conforming implementation may provide more versions
of main(), but they must all have return type int.
If you write "void main() { }", then main does not have return
type int. Therefore, a conforming implementation is not permitted
to accept it.
That depends on whether the books purport to describe Standard C++, or
whether they are describing the dialects accepted by particular
implementationa.
Strictly speaking, programs are neither conforming nor nonconforming.
The C++ standard does not impose requirements directly on programs; it
imposes requirements only on implementations. So if you want to be
strictly correct about it, you should not talk about conforming or
non-conforming programs, only about conforming or non-conforming
implementations.
It is also correct to say that conforming implementations are not
required to accept "void main()" -- in fact, they are required to
produce a diagostic message for such programs. This requirement is
often ignored in practice, but it's still there.
What it proves is that arguments of the form "My compiler accepts it,
so it's right" are specious.
Yes it does, in subclause 3.6.1, paragraph 2.
No they're not -- subclause 3.6.1 paragraph 2 makes that clear.
A
Andrew Koenig
Andrey> Strictly speaking, implementations are allowed to do anything they
Andrey> please as long as any compliant C++ code is interpreted correctly.
No, they're not. Implementations are also required to produce
diagnostic messages when presented with ill-formed programs, unless
the ill-formedness is described with "no diagnostic is required" in
the standard.
Andrey> please as long as any compliant C++ code is interpreted correctly.
No, they're not. Implementations are also required to produce
diagnostic messages when presented with ill-formed programs, unless
the ill-formedness is described with "no diagnostic is required" in
the standard.
A
Andrew Koenig
Strictly speaking, implementations are allowed to do anything they
Sorry, but you're both mistaken.
See subclause 1.4 of the standard, which says, among other things:
The set of "diagnosible rules" consists of all syntactic and
semantic rules in this International Standard except for those
rules containing an explicit notation that "no diagnostic is
required" or which are described as resulting in "undefined
behavior."
...
If a program contains a violation of any diagnosible rule,
a conforming implementation shall issue at least one
diagnostic message...
So conforming C++ implementations are not allowed to accept anything
at all unless they issue a diagnostic message for nonconforming programs.
Sorry, but you're both mistaken.
See subclause 1.4 of the standard, which says, among other things:
The set of "diagnosible rules" consists of all syntactic and
semantic rules in this International Standard except for those
rules containing an explicit notation that "no diagnostic is
required" or which are described as resulting in "undefined
behavior."
...
If a program contains a violation of any diagnosible rule,
a conforming implementation shall issue at least one
diagnostic message...
So conforming C++ implementations are not allowed to accept anything
at all unless they issue a diagnostic message for nonconforming programs.
A
Andrew Koenig
Then explain why the standard says it is okay to use return type IN ADDITION to
--> BUT<--
No, that's not what the sentence says. The sentence is
It [the type of main] shall have a return type of type "int",
but otherwise its type is implementation-defined.
The return type is only part of the type of a function.
So the standard is saying that the return-type part of the type of
main is required to be "int" and the rest of the type of main is
implementation-defined.
--> BUT<--
No, that's not what the sentence says. The sentence is
It [the type of main] shall have a return type of type "int",
but otherwise its type is implementation-defined.
The return type is only part of the type of a function.
So the standard is saying that the return-type part of the type of
main is required to be "int" and the rest of the type of main is
implementation-defined.
A
Andrew Koenig
Try another guess. The "but" allows for different returns types, as
and who is also the convener of the ISO C++ standards committee.
and who is also the convener of the ISO C++ standards committee.
W
WW
Andrew said:
Yep. It is in his signature. But I doubt that a Seriously Abusive
Gibberish Emitter knows what a convener is. (To be honest I don't know
really either. I even misspell it all the time as convenor. :-( )
A
Andrew Koenig
I'm not saying that implicit return 0 is the greatest thing since
Especially if you know the history.
Here's the order in which things happened:
1) Someone noticed that the following program:
main() /* yes, I meant "main", not "int main" */
{ }
was a valid C program but not a valid C++ program.
I forget the exact reason, but it was something having to do
with subtle differences in behavior when a program fails to
return a value and its caller doesn't look at the value.
2) Accordingly, there was a proposal in the C++ committee
to put an implicit "return 0;" at the end of main, as a
compatibility hack. This proposal was accepted.
3) The C++ committee then learned that the C committee had
prohibited implicit return types entirely in C9x. This was
a surprise, as we had not expected the C committee to break
any valid C89 programs.
4) As a result, the C++ committee also proposed to prohibit
implicit return types. This proposal was also accepted.
5) Because
main() { }
was now broken (because of the lack of an explicit return type),
there was no further need for the implicit "return 0;"
I think there was even a proposal to remove it. However, the
committee decided to leave it alone because there might already
be programs out there relying on it.
Isn't compatibility wonderful?
Especially if you know the history.
Here's the order in which things happened:
1) Someone noticed that the following program:
main() /* yes, I meant "main", not "int main" */
{ }
was a valid C program but not a valid C++ program.
I forget the exact reason, but it was something having to do
with subtle differences in behavior when a program fails to
return a value and its caller doesn't look at the value.
2) Accordingly, there was a proposal in the C++ committee
to put an implicit "return 0;" at the end of main, as a
compatibility hack. This proposal was accepted.
3) The C++ committee then learned that the C committee had
prohibited implicit return types entirely in C9x. This was
a surprise, as we had not expected the C committee to break
any valid C89 programs.
4) As a result, the C++ committee also proposed to prohibit
implicit return types. This proposal was also accepted.
5) Because
main() { }
was now broken (because of the lack of an explicit return type),
there was no further need for the implicit "return 0;"
I think there was even a proposal to remove it. However, the
committee decided to leave it alone because there might already
be programs out there relying on it.
Isn't compatibility wonderful?
W
WW
Andrew Koenig wrote:
[SNIP]
That reminds me of something I have read in one of those a graffiti books.
It was something like this: "Hey, Lord! Shouldn't we erase this whole thing
and start over with some other kind of animal?"
[SNIP]
That reminds me of something I have read in one of those a graffiti books.
It was something like this: "Hey, Lord! Shouldn't we erase this whole thing
and start over with some other kind of animal?"
T
The_Sage
Reply to article by: "WW said:
You didn't answer the question. I take that as an admission you have no clue.
The Sage
=============================================================
My Home Page : http://members.cox.net/the.sage
"The men that American people admire most extravagantly are
most daring liars; the men they detest the most violently are
those who try to tell them the truth" -- H. L. Mencken
=============================================================
T
The_Sage
Reply to article by: "David B. Held said:
I have challenged you several times to show statements by any of those
organizations claiming that their compilers are less than 100% conforming. They
say they conform, therefore they are conforming. If you want to make up new
definitions for words so that "conforming" now means "almost conforming", go
right ahead, but don't be surprized if no one ever take anything seriously that
you ever say from then on.
The Sage
=============================================================
My Home Page : http://members.cox.net/the.sage
"The men that American people admire most extravagantly are
most daring liars; the men they detest the most violently are
those who try to tell them the truth" -- H. L. Mencken
=============================================================
T
The_Sage
Reply to article by: Andrew Koenig said:
That proves nothiong expect there are alot of people who resort to the logical
fallacy of appeal to authority. The standard allows room for other types,
whether deliberately or not, and many major compiler makers use additional
types. Whether they like it or not, the standard needs to conform to what the
user needs, not what Herb or you want.
The Sage
=============================================================
My Home Page : http://members.cox.net/the.sage
"The men that American people admire most extravagantly are
most daring liars; the men they detest the most violently are
those who try to tell them the truth" -- H. L. Mencken
=============================================================