c99 on Microsoft Visual Studio

[xuanbai98]

I want to know the current support status of c99 on Microsoft Visual
Studio. Anyone know whether Microsoft has any plan to support c99 on
Visual Studio?

Thanks,

Xuan

[Greg Comeau]

In article <35*************@individual.net>, Eltee <el***@hotmail.com> wrote:

Alan Balmer wrote:
All the work? If this wasn't funny it'd be tragical. This is comp.lang._C_,
isn't it? People in here should know something about C, shouldn't they?
Well, if this whole business wasn't just academic they should know about
compilers, too. Especially those that adhere to the standard.
I never realized that listing some
of those would be such an unthinkable thing to ask.

I'll side step the above and provide this (which I has posted
Aug 26, 2003 at 12:18PM) and believe that it is still accurate:

"To the best of my understanding, at least the following
is the situation as regards C99 implementations:

* Full: Comeau C/C++ (compiler) + Dinkumware (lib)
For: many platforms
* Full: IBM C for AIX version 6
For: AIX
* Intel C (at least compiler)
For: x86
* Full: Compaq/HP/Dec C (this has had various flavors)
For: Unsure
* Full: SGI C
For: SGI platforms (MIPS IRIX I guess)
* Full: Lund Multiprocessor Compiler Company
For: PowerPC LINUX on Apple G4
* Unsure: SAS/C
For: IBM mainframes
* Partial: gcc
For: Many platforms
* Partial: lcc-win32
For: MS-Windows
Only the latter two have "free" possibilities.
As an aside, there is now a small number of books where C99 is
discussed. Search for C99 at http://www.comeaucomputing.com/booklist "
--
Greg Comeau / Comeau C++ 4.3.3, for C++03 core language support
Comeau C/C++ ONLINE ==> http://www.comeaucomputing.com/tryitout
World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?

[Eltee]

Greg Comeau wrote:

In article <35*************@individual.net>, Eltee <el***@hotmail.com> wrote:

Alan Balmer wrote:
All the work? If this wasn't funny it'd be tragical. This is comp.lang._C_,
isn't it? People in here should know something about C, shouldn't they?
Well, if this whole business wasn't just academic they should know about
compilers, too. Especially those that adhere to the standard.
I never realized that listing some
of those would be such an unthinkable thing to ask.

I'll side step the above and provide this (which I has posted
Aug 26, 2003 at 12:18PM) and believe that it is still accurate:

"To the best of my understanding, at least the following
is the situation as regards C99 implementations:

* Full: Comeau C/C++ (compiler) + Dinkumware (lib)
For: many platforms
* Full: IBM C for AIX version 6
For: AIX
* Intel C (at least compiler)
For: x86
* Full: Compaq/HP/Dec C (this has had various flavors)
For: Unsure
* Full: SGI C
For: SGI platforms (MIPS IRIX I guess)
* Full: Lund Multiprocessor Compiler Company
For: PowerPC LINUX on Apple G4
* Unsure: SAS/C
For: IBM mainframes
* Partial: gcc
For: Many platforms
* Partial: lcc-win32
For: MS-Windows

Thank you, thank you, thank you. Now my heart is full. ;-)


Only the latter two have "free" possibilities.
As an aside, there is now a small number of books where C99 is
discussed. Search for C99 at http://www.comeaucomputing.com/booklist "

[Greg Comeau]

In article <sl******************@ccserver.keris.net>,
Chris Croughton <ch***@keristor.net> wrote:

On 26 Jan 2005 20:02:20 -0500, Greg Comeau
<co****@panix.com> wrote:

In article <sl******************@ccserver.keris.net>,
Chris Croughton <ch***@keristor.net> wrote:

On Wed, 26 Jan 2005 04:48:37 GMT, Randy Howard
<ra*********@FOOverizonBAR.net> wrote:
In article <ct**********@panix2.panix.com>, co****@panix.com says...
> Comeau, gcc, and Dinkware are available across the most popular
> platforms, and that covers at least a reasonable subset.

gcc (even with the c99 command line options) is not a c99 compiler.
It's "part" of one perhaps, but that's it.

And doesn't provide the library, which is a large part of the
specification. GLIBC2 is a lot of the way there but still has some
deficiencies.
This suggestion is not intended to change the above fact,
but it's worth pointing out that in addition to use with
Comeau on various platforms, that Dinkumware's C99 library
is available for use with gcc too on various platforms.

Where is it stated which versions of gcc and which platforms? I can
find the price ($135 per user for the "Single User License" -- which I
can't find stated on their site) and a vague statement that the
libraries can be configured for "several compilers on Windows, Unix,
Linux, and Mac OS/X" but not which ones (Debian 'stable'? Gentoo?
Cygwin?). Or there's the source licence for $600 which presumably will
compile on any C89 and later compiler (but presumably also needs to know
about system 'hooks'). Whether the users of applications compiled with
those libraries also need to buy the libraries I can't find (it's
probably in the licence which I can't find on that site).

This seems a good summary http://www.dinkumware.com/libdual.html
Plus a lot of programmers in the Real World(tm) have to work with the
library supplied by the customer, they can't just use another library
which might not be compatible with the customer's other software and
bought-in 3rd-party libraries.

As usual, that can be said of any compiler/lib.
And it also goes as well that if a lot do, then
a lot don't, both in the real world and not in it.
One needs to start somewhere. Your implication is that
say Dinkumware can't be involved in the lib supplied by
the customer. Of course it can. And often is.
There is no escaping issues depending upon the choice,
so are generally neutral although they are different.
If not, I suspect we're talking about different things.
--
Greg Comeau / Comeau C++ 4.3.3, for C++03 core language support
Comeau C/C++ ONLINE ==> http://www.comeaucomputing.com/tryitout
World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?

[Michael Wojcik]

In article <ct**********@panix1.panix.com>, co****@panix.com (Greg Comeau) writes:

I'll side step the above and provide this (which I has posted
Aug 26, 2003 at 12:18PM) and believe that it is still accurate:

"To the best of my understanding, at least the following
is the situation as regards C99 implementations:
Of course it's still important to check the conformance claims for
the particular release you're using (as I'm sure Greg would agree).
Case in point:
* Full: Compaq/HP/Dec C (this has had various flavors)
For: Unsure

I know HP claims that its current OpenVMS C implementation is fully-
conforming, and it may be that its other implementations (which may
be largely the same implementation, for different platforms, under
different names) now conform as well. But the (relatively recent) HP
C compilers we're using on HP-UX and Tru64 do not conform completely
to C99, in a particularly annoying fashion: they have the old, broken
snprintf return semantics.

That might seem unimportant to some, but snprintf is very useful for
avoiding buffer overflows - more useful than, say, strncpy, since
it's not trivial to do the same length checks manually. And the old
return semantics for snprintf (-1 is returned both for overflow and
for formatting error) not only make it more cumbersome to, say,
reallocate the destination buffer until it's sufficiently large;
they make it impossible to distinguish among simply needing a
larger destination, needing a destination that the program would
consider *too* large, and a formatting error which no size of
destination will ever correct.

Consequently, code using snprintf and expecting the C99 semantics
will likely be unpleasantly broken under these nonconforming HP
implementations. (And for that reason I wrote and ran a little
snprintf test program on all of the platforms we're currently
supporting. Microsoft Visual C 6 and the two HP platforms were
the only ones which still had the old semantics, I think.)

So while Greg's list is a useful starting point, and a handy thing to
have for the general discussion of C99 uptake, I wouldn't rely on it
(except for the case of Comeau/Dinkumware) when you're looking at
writing C99 code for some particular implementation.

--
Michael Wojcik mi************@microfocus.com

Pocket #9: A complete "artificial glen" with rocks, and artificial moon,
and forester's station. Excellent for achieving the effect of the
sublime without going out-of-doors. -- Joe Green

[S.Tobias]

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:

Eltee <el***@hotmail.com> wrote:

>I'm particularly interested in ISO C99 standard defects.
>Would you care to elaborate?

[snip]
> You had better
get yourself seriously interested in the Standard in the first place.
Take it as my good-will advice.

This is a good start, isn't it? Getting to know the defects.
Absolutely. As good as building a house and starting with the smoke
from the chimney.

Studying the defects is very educating, but first you have to well
understand what they are for.

[snip]
So I'm only allowed to ask quesions to which I know the answers already (after a
looooooooong study of the standard in this case)?
No. But it's considered a good behaviour to do some *basic* research
yourself. By not doing so you show your disrespect towards those who
answer your questions. It's not that someone has to read your posts,
because nobody has to and it doesn't cost a lot to kill entire thread;
its your _attitude_ that annoys others. When I ask a question I often
prepare myself for the discussion first. Sometimes it takes days;
sometimes minutes, when I discover the answer is obvious.
I think that having the
(alleged) defects (of the ISO C standard) exposed is quite on-topic in this NG.
Yes, and even more so in c.s.c. What have you exposed?
Besides, I wasn't even aware that there were any (defects). I learned that fact
in this NG. In this particular thread, for that matter.
You sound very naive.
Are those defects a c.l.c-taboo or something?
Not in the least, it's just they're a few clicks away from you. If only
you just cared to have a look at the proper place...

[snip]
I beg you to grow up. "Plonk" is usually the last word you hear from
a person here, and it means you lost. If anybody is still speaking to
you, he's a friendly soul, even if he's giving you a reprimand. If you
want to learn C, you're invited, but try to be nice to others and don't
try to fight existing customs or invent your own. If you just want to
annoy others (you have done so recently), please go away.

Google working good, thank you very much. In fact, I don't know if this NG even
deserves to be alive anymore. Google outdoes it in every aspect.
It's your opinion. Some believe giraffes can't exist.
Oh, well, maybe
it can serve as a meeting point for those ol'-c-geeks who want to avoid going
through too many cold turkeys.
Many newbies, who perhaps know C less than you, come here and obtain
very good answers. Wonder where the problem lies...
> Don't you
read c.s.c.? Is WG14 page down or is it too hard for you to find any
DRs there?

64

Well, you've found them youself, good! Is that too few?

--
Stan Tobias
mailx `echo si***@FamOuS.BedBuG.pAlS.INVALID | sed s/[[:upper:]]//g`

[Greg Comeau]

In article <ct********@news4.newsguy.com>,
Michael Wojcik <mw*****@newsguy.com> wrote:

In article <ct**********@panix1.panix.com>, co****@panix.com (Greg Comeau) writes:

I'll side step the above and provide this (which I has posted
Aug 26, 2003 at 12:18PM) and believe that it is still accurate:

"To the best of my understanding, at least the following
is the situation as regards C99 implementations:
Of course it's still important to check the conformance claims for
the particular release you're using (as I'm sure Greg would agree).

I do agree. That said...
Case in point:

* Full: Compaq/HP/Dec C (this has had various flavors)
For: Unsure

I know HP claims that its current OpenVMS C implementation is fully-
conforming, and it may be that its other implementations (which may
be largely the same implementation, for different platforms, under
different names) now conform as well. But the (relatively recent) HP
C compilers we're using on HP-UX and Tru64 do not conform completely
to C99, in a particularly annoying fashion: they have the old, broken
snprintf return semantics.

That might seem unimportant to some, but snprintf is very useful for
avoiding buffer overflows - more useful than, say, strncpy, since
it's not trivial to do the same length checks manually. And the old
return semantics for snprintf (-1 is returned both for overflow and
for formatting error) not only make it more cumbersome to, say,
reallocate the destination buffer until it's sufficiently large;
they make it impossible to distinguish among simply needing a
larger destination, needing a destination that the program would
consider *too* large, and a formatting error which no size of
destination will ever correct.

Consequently, code using snprintf and expecting the C99 semantics
will likely be unpleasantly broken under these nonconforming HP
implementations. (And for that reason I wrote and ran a little
snprintf test program on all of the platforms we're currently
supporting. Microsoft Visual C 6 and the two HP platforms were
the only ones which still had the old semantics, I think.)

So while Greg's list is a useful starting point, and a handy thing to
have for the general discussion of C99 uptake, I wouldn't rely on it
(except for the case of Comeau/Dinkumware) when you're looking at
writing C99 code for some particular implementation.

.....I think at the time I wrote the original post i was
considering the compiler aspect not the libs. I believe
you are correct that the libs are still across the baord,
though I have not done any additional major investigations
since the Aug 2003 post. Probably even the compilers are
across the board. As I think Larry wrote about C89,
vendors were claiming conformance even before the standard
was finished. The same thing happened before C++98,
if I recall years before it, so it wouldm't surprise me if
there are even major issues with some of the claims.
So sure, don't take the list as authoritive, it was mainly
to show progress was occuring, not a compliance survey.
--
Greg Comeau / Comeau C++ 4.3.3, for C++03 core language support
Comeau C/C++ ONLINE ==> http://www.comeaucomputing.com/tryitout
World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?

[Randy Howard]

In article <ct**********@panix2.panix.com>, co****@panix.com says...

In article <MP************************@news.verizon.net>,
Randy Howard <ra*********@FOOverizonBAR.net> wrote:

In article <ct**********@panix2.panix.com>, co****@panix.com says...

Comeau, gcc, and Dinkware are available across the most popular
platforms, and that covers at least a reasonable subset.

gcc (even with the c99 command line options) is not a c99 compiler.
It's "part" of one perhaps, but that's it.

I feel like I've been tricked. Maybe I misunderstood,
but I thought somebody was asking for a subset of C99 that
could be used across multiple compilers and on multiple
platforms. gcc is definitely not a full c99 compiler,
but I think can meet the _substandard_ 5 features or whatever
that was being requested.

Ahh, I guess I misunderstood you, no "trick" intended. Are you
saying that whatever gcc -std=c99 provides is the "common"
subset that the other compilers you mentioned (Intel, IBM,
Compaq and a vague "etc.") support? If true, it seems like
a reasonable way to approach the problem.

--
Randy Howard (2reply remove FOOBAR)
"For some reason most people seem to be born without the part
of the brain that understands pointers." -- Joel Spolsky

[Eltee]

S.Tobias wrote:

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:

Eltee <el***@hotmail.com> wrote:
I'm particularly interested in ISO C99 standard defects.
>>Would you care to elaborate?

[snip]

> You had better

get yourself seriously interested in the Standard in the first place.

Take it as my good-will advice.

This is a good start, isn't it? Getting to know the defects.

Absolutely. As good as building a house and starting with the smoke
from the chimney.

I don't know about you, but I wouldn't even consider building the house if I
learned beforehand that the smoke is going to backfire on me, in the first
place. It's no use blinding yourself with "we're going to cross that bridge when
we get to it" when you can easily find out right at the begining that the gap
you're supposed to cross has interplanetary "dimensions" and would, in fact,
demand a space ship. I'm not saying that this is the case with C99 defects, but
I don't think that learning what they are, before you familiarize yourself with
all the intricacies of the language itself, is a bad thing.
Studying the defects is very educating, but first you have to well
understand what they are for.
Understand what the defects are for?!? Even before you go studying them?!? This
is a joke, right? No, wait, those are _two_ separate jokes.
[snip]

So I'm only allowed to ask quesions to which I know the answers already (after a
looooooooong study of the standard in this case)?

No. But it's considered a good behaviour to do some *basic* research
yourself.

But this _is_ what I'm doing.
By not doing so you show your disrespect towards those who
answer your questions.
That's the point. They aren't answering them. Not at first, that is. I have to
drag every word out of their "mouths". This whole thread would have been
unnecessary if that first someone simply had posted a few links (_one_ even)
instead of saying that this has been answered before someplace.
It's not that someone has to read your posts,
because nobody has to and it doesn't cost a lot to kill entire thread;
its your _attitude_ that annoys others. When I ask a question I often
prepare myself for the discussion first. Sometimes it takes days;
sometimes minutes, when I discover the answer is obvious.

I think that having the
(alleged) defects (of the ISO C standard) exposed is quite on-topic in this NG.

Yes, and even more so in c.s.c. What have you exposed?

Not me. But others have. Still others are trying to ban such discussions from
clc. I can't understand such behaviour.

Besides, I wasn't even aware that there were any (defects). I learned that fact
in this NG. In this particular thread, for that matter.

You sound very naive.

Maybe. But, what would you say of a man who tells you that before you study the
defects you first have to understand them?

Are those defects a c.l.c-taboo or something?

Not in the least, it's just they're a few clicks away from you. If only
you just cared to have a look at the proper place...

There we go again. Proper place. We all know where that's at, don't we. For all
the cases, for all the situations. Why, oh, why do we need clc then, if there's
some other proper place for everything?
[snip]
I beg you to grow up. "Plonk" is usually the last word you hear from
a person here, and it means you lost. If anybody is still speaking to
you, he's a friendly soul, even if he's giving you a reprimand. If you
want to learn C, you're invited, but try to be nice to others and don't
try to fight existing customs or invent your own. If you just want to
annoy others (you have done so recently), please go away.

Google working good, thank you very much. In fact, I don't know if this NG even
deserves to be alive anymore. Google outdoes it in every aspect.

It's your opinion.

But of course!
Some believe giraffes can't exist.
You among them, I take it.

Oh, well, maybe
it can serve as a meeting point for those ol'-c-geeks who want to avoid going
through too many cold turkeys.

Many newbies, who perhaps know C less than you, come here and obtain
very good answers.

In spite of all the proper places out there?
Wonder where the problem lies...

Me too, me too. Have been wondering right from the first non-answer I got in
this NG.

> Don't you

read c.s.c.? Is WG14 page down or is it too hard for you to find any
DRs there?

64

Well, you've found them youself, good! Is that too few?

Samoa.

[Greg Comeau]

In article <MP************************@news.verizon.net>,
Randy Howard <ra*********@FOOverizonBAR.net> wrote:

In article <ct**********@panix2.panix.com>, co****@panix.com says...

In article <MP************************@news.verizon.net>,
Randy Howard <ra*********@FOOverizonBAR.net> wrote:

>In article <ct**********@panix2.panix.com>, co****@panix.com says...
>> Comeau, gcc, and Dinkware are available across the most popular
>> platforms, and that covers at least a reasonable subset.
>
>gcc (even with the c99 command line options) is not a c99 compiler.
>It's "part" of one perhaps, but that's it.
I feel like I've been tricked. Maybe I misunderstood,
but I thought somebody was asking for a subset of C99 that
could be used across multiple compilers and on multiple
platforms. gcc is definitely not a full c99 compiler,
but I think can meet the _substandard_ 5 features or whatever
that was being requested.

Ahh, I guess I misunderstood you, no "trick" intended. Are you
saying that whatever gcc -std=c99 provides is the "common"
subset that the other compilers you mentioned (Intel, IBM,
Compaq and a vague "etc.") support?

Probably.

BTW gcc's features is no secret, or should not be, see here:
http://gcc.gnu.org/c99status.html
If true, it seems like
a reasonable way to approach the problem.

I don't know if it's reasonable, but at least it is a start.
No doubt for every project where that's ok, there's
another where it is completely insufficient.... assuming
compilation across 5 compilers (assuming on 5 different OSs)
is/was the goal.
--
Greg Comeau / Comeau C++ 4.3.3, for C++03 core language support
Comeau C/C++ ONLINE ==> http://www.comeaucomputing.com/tryitout
World Class Compilers: Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?

[S.Tobias]

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:
[cut the crap]

Studying the defects is very educating, but first you have to well
understand what they are for. Understand what the defects are for?!? Even before you go studying them?!? This
is a joke, right? No, wait, those are _two_ separate jokes.
In case I didn't express myself clearly, I meant there's no point in
studying the Standard defects if you don't know the Standard in the first
place; it was just a piece of advice. But this is not my business,
now you know where to look for them, go figure yourself. It's not
what my argument is about anyway.

So I'm only allowed to ask quesions to which I know the answers already (after a
looooooooong study of the standard in this case)?

No. But it's considered a good behaviour to do some *basic* research
yourself. But this _is_ what I'm doing.
No, this is exactly what you are expected to do, but not doing it at all.
I have pointed out to you, in this particular case you had plenty of
sources to answer your particular question, you just had to reach out
for them. In the clc FAQ you have "11. ANSI/ISO Standard C" where you
can find more than you asked for, and links for specific information.
(Of course, you have read the FAQ, haven't you?) If you at least new
what the latest ISO C standard version was, you'd realize how idiotic
your question was.

(I hope you don't start asking now what the current version is.)
> By not doing so you show your disrespect towards those who
answer your questions. That's the point. They aren't answering them. Not at first, that is. I have to
drag every word out of their "mouths". This whole thread would have been
unnecessary if that first someone simply had posted a few links (_one_ even)
instead of saying that this has been answered before someplace.
Because Usenet was not created specially for you. If someone gives you
a *friendly response* that something has been already discussed before,
it roughly means: "I can't answer your question, but I remember this topic
appeared here before, you can easily find it there and get comprehensive
answer". What do you think, that anybody remembers by heart all you
ask about? No, those people often would have to get to their keyboards
and first find and check the answers themselves. If you can't make
any effort, why can you expect others to do it for you? Nobody is your
servant.

[snip]

> Some believe giraffes can't exist.

You among them, I take it.

If you look for friends, you'll find many in c.[ls].c. If you look
for enemies, don't bother, you'll find none; those people are really
serious here.

--
Stan Tobias
mailx `echo si***@FamOuS.BedBuG.pAlS.INVALID | sed s/[[:upper:]]//g`

[Eltee]

S.Tobias wrote:

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:

[cut the crap]

Studying the defects is very educating, but first you have to well
understand what they are for.

Understand what the defects are for?!? Even before you go studying them?!? This
is a joke, right? No, wait, those are _two_ separate jokes.

In case I didn't express myself clearly, I meant there's no point in
studying the Standard defects if you don't know the Standard in the first
place;

Well, I see it this way. Why study the standard if I can get to know the defects
from someone else who already studied it in great detail and knows it so much
better than me? If that someone is prepared to give me that information or to
show me where I can find it myself (thanks, Greg!) then it saves me a lot of
time. Idealy everybody should study everything, get to know everything,
(re)verify all the proofs of all the theorems (I think that R. Feynman said
something like that once). Unfortunately we don't live in an ideal world. In our
(_real_) world time plays an important role, you see. Like most people, I'm
frequently delegating my time and - most importantly - trust to people more
knowledgeable (in some particular field of knowledge) than me. Having never
written a serious compiler (let alone C99 compliant one) and not having any
aspirations to do so in the future, I clearly don't have other choice but to
trust other people, like Greg, who did. Call it naive if you want, but I don't
think that in this particular case I'm in much danger of being doublecrossed.
it was just a piece of advice. But this is not my business,
now you know where to look for them, go figure yourself. It's not
what my argument is about anyway.
So I'm only allowed to ask quesions to which I know the answers already (after a
looooooooong study of the standard in this case)?
No. But it's considered a good behaviour to do some *basic* research
yourself.

But this _is_ what I'm doing.

No, this is exactly what you are expected to do, but not doing it at all.
I have pointed out to you, in this particular case you had plenty of
sources to answer your particular question, you just had to reach out
for them. In the clc FAQ you have "11. ANSI/ISO Standard C" where you
can find more than you asked for, and links for specific information.
(Of course, you have read the FAQ, haven't you?) If you at least new
what the latest ISO C standard version was, you'd realize how idiotic
your question was.

(I hope you don't start asking now what the current version is.)

> By not doing so you show your disrespect towards those who

answer your questions.

That's the point. They aren't answering them. Not at first, that is. I have to
drag every word out of their "mouths". This whole thread would have been
unnecessary if that first someone simply had posted a few links (_one_ even)
instead of saying that this has been answered before someplace.

Because Usenet was not created specially for you. If someone gives you
a *friendly response* that something has been already discussed before,
it roughly means: "I can't answer your question, but I remember this topic
appeared here before, you can easily find it there and get comprehensive
answer". What do you think, that anybody remembers by heart all you
ask about?

No, not at all. But the question of compiler compliancy is, in my view at least,
of utmost importance, if someone is seriously involved in C, as I would
absolutely expect the people (well, the experts at least) of c.l.c to be. Ergo,
I would expect that people would have no trouble listing the major compliant
compilers if someone asked. Don't misunderstand me: I don't believe my
expectations were misplaced. Not then, not now. I'm certain that people do know,
yet they are somewhat reluctant to tell. I'm still figuring out the reason
behind that reluctancy.
No, those people often would have to get to their keyboards
and first find and check the answers themselves. If you can't make
any effort, why can you expect others to do it for you? Nobody is your
servant.

[snip]
> Some believe giraffes can't exist.

You among them, I take it.

If you look for friends, you'll find many in c.[ls].c. If you look
for enemies, don't bother, you'll find none; those people are really
serious here.

OK. We friends now? ;-)

[S.Tobias]

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:
[snip]

Because Usenet was not created specially for you. If someone gives you
a *friendly response* that something has been already discussed before,
it roughly means: "I can't answer your question, but I remember this topic
appeared here before, you can easily find it there and get comprehensive
answer". What do you think, that anybody remembers by heart all you
ask about? No, not at all. But the question of compiler compliancy is, in my view at least,
of utmost importance, if someone is seriously involved in C, as I would
absolutely expect the people (well, the experts at least) of c.l.c to be.
Yes, it is important, but there's no need to exaggerate here. We all
work with broken compilers of broken languages and broken tools on
broken OS-s with broken protocols on broken hardware. And we
still live.
Ergo,
I would expect that people would have no trouble listing the major compliant
compilers if someone asked. Don't misunderstand me: I don't believe my
expectations were misplaced. Not then, not now.
Sometimes it's impossible to give a single and unambiguous answer
to a simple question. For any Yes-es there may be as many No-s.
If you trust people more knowledgeable than you, as you say
(snipped away), why haven't you trusted Michael Wojcik? He gave you
a long and over-broad explanation why such an answer in not possible.
He has been extremely kind to you. You really _were_ to look first
into the past discussions, and then come back again if you _needed_
more answers. If you hadn't done that, you weren't even aware what
you were asking about.
I'm certain that people do know,
yet they are somewhat reluctant to tell.
People here are not automatic machines for answering questions.
They are here to *help* other people. It means that you have to
treat them as humans, and you have to be human too. They have their
feelings, too. If someone doesn't want to answer you question,
you have to respect him. Don't treat clc as a private chat room.
The answers that you get here are serious. Patience and humility
are good companions here.
I'm still figuring out the reason
behind that reluctancy.
There're many reasons why certain questions are not and should not
be answered. Stay here a longer time, you'll learn why.

If you look for friends, you'll find many in c.[ls].c. If you look
for enemies, don't bother, you'll find none; those people are really
serious here.

OK. We friends now? ;-)

Yes, with pleasure. After you learn how to behave here.

--
Stan Tobias
mailx `echo si***@FamOuS.BedBuG.pAlS.INVALID | sed s/[[:upper:]]//g`

[Keith Thompson]

Eltee <el***@hotmail.com> writes:

S.Tobias wrote:

[snip]

In case I didn't express myself clearly, I meant there's no point in
studying the Standard defects if you don't know the Standard in the first
place;

Well, I see it this way. Why study the standard if I can get to know
the defects from someone else who already studied it in great detail
and knows it so much better than me?

Knowing about the defects in the standard is a very small part of
understanding the standard, and it's probably not the first thing you
should worry about.

It's up to you, of course, but I don't see the point of studying, for
example, a defect in the wording that describes the conversion of an
imaginary type to _Bool (DR 285) without first understanding
conversions, imaginary types, and _Bool.

Why do you want to focus on the defects rather than on understanding
the standard as a whole?

--
Keith Thompson (The_Other_Keith) ks***@mib.org <http://www.ghoti.net/~kst>
San Diego Supercomputer Center <*> <http://users.sdsc.edu/~kst>
We must do something. This is something. Therefore, we must do this.

[Eltee]

S.Tobias wrote:

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:

[snip]

Because Usenet was not created specially for you. If someone gives you
a *friendly response* that something has been already discussed before,
it roughly means: "I can't answer your question, but I remember this topic
appeared here before, you can easily find it there and get comprehensive
answer". What do you think, that anybody remembers by heart all you
ask about?

No, not at all. But the question of compiler compliancy is, in my view at least,
of utmost importance, if someone is seriously involved in C, as I would
absolutely expect the people (well, the experts at least) of c.l.c to be.

Yes, it is important, but there's no need to exaggerate here. We all
work with broken compilers of broken languages and broken tools on
broken OS-s with broken protocols on broken hardware. And we
still live.

Right. Me, I use gcc in spite of it not being compliant. But, from what I've
learned in this discussion, I might have to change that.

Ergo,
I would expect that people would have no trouble listing the major compliant
compilers if someone asked. Don't misunderstand me: I don't believe my
expectations were misplaced. Not then, not now.

Sometimes it's impossible to give a single and unambiguous answer
to a simple question. For any Yes-es there may be as many No-s.
If you trust people more knowledgeable than you, as you say
(snipped away), why haven't you trusted Michael Wojcik?

Hm. Let me see. I asked for a list of compliant compilers. I would have been
happy with a very short list: one entry would've been enough. He gave me nothing
of the sort. There was actually nothing to put my trust into. Greg, on the other
hand, gave me what I wanted.
He gave you
a long and over-broad explanation why such an answer in not possible.
He has been extremely kind to you. You really _were_ to look first
into the past discussions, and then come back again if you _needed_
more answers. If you hadn't done that, you weren't even aware what
you were asking about.

I'm certain that people do know,
yet they are somewhat reluctant to tell.

People here are not automatic machines for answering questions.
They are here to *help* other people. It means that you have to
treat them as humans, and you have to be human too. They have their
feelings, too. If someone doesn't want to answer you question,
you have to respect him.

Right. I could live with no answers. What I don't understand is the answers
saying "I don't know the answer to your question, but here's something else if
you like". Well, sorry folks, no offence meant but ... me no like.
Don't treat clc as a private chat room.
The answers that you get here are serious. Patience and humility
are good companions here.

I'm still figuring out the reason
behind that reluctancy.

There're many reasons why certain questions are not and should not
be answered. Stay here a longer time, you'll learn why.
If you look for friends, you'll find many in c.[ls].c. If you look
for enemies, don't bother, you'll find none; those people are really
serious here.

OK. We friends now? ;-)

Yes, with pleasure. After you learn how to behave here.

Doing my best to learn. Just give me some time, OK?

[Eltee]

Keith Thompson wrote:

Eltee <el***@hotmail.com> writes:

S.Tobias wrote:

[snip]

In case I didn't express myself clearly, I meant there's no point in
studying the Standard defects if you don't know the Standard in the first
place;

Well, I see it this way. Why study the standard if I can get to know
the defects from someone else who already studied it in great detail
and knows it so much better than me?

Knowing about the defects in the standard is a very small part of
understanding the standard, and it's probably not the first thing you
should worry about.

It's up to you, of course, but I don't see the point of studying, for
example, a defect in the wording that describes the conversion of an
imaginary type to _Bool (DR 285) without first understanding
conversions, imaginary types, and _Bool.

Why do you want to focus on the defects rather than on understanding
the standard as a whole?

I just recently became aware of those defects. Feeling more than a little
surprised by their existence I wanted to learn what they were. Not knowing the
standard I thought my "learned colleagues" could tell me more about them.

Secondly, I don't have any real aspirations in knowing the standard itself. My
interest lies more within the field of the language the standard "describes". So ...

....thirdly, I'm worried that the language inherited those defects from the standard.

[S.Tobias]

Eltee <el***@hotmail.com> wrote:

S.Tobias wrote:

Sometimes it's impossible to give a single and unambiguous answer
to a simple question. For any Yes-es there may be as many No-s.
If you trust people more knowledgeable than you, as you say
(snipped away), why haven't you trusted Michael Wojcik?

Hm. Let me see. I asked for a list of compliant compilers. I would have been
happy with a very short list: one entry would've been enough.
You can't impose limits on your interlocutor. This is not a quiz.
He gave me nothing
of the sort.
He gave you a fishing rod.
There was actually nothing to put my trust into.
Stay here longer, you'll get the Tao.
Greg, on the other
hand, gave me what I wanted.
Yes. The world is not all black-and-white only.
[snip]
Right. I could live with no answers. What I don't understand is the answers
saying "I don't know the answer to your question, but here's something else if
you like". Well, sorry folks, no offence meant but ... me no like.

Then look into "something else" and tell why you're not satisfied with it.
I'm sure everybody will be happy to answer you, and you'll be happy
with the answers you obtain.
Look, I don't want to convince you any more, I think I have said enough.
Life is not easy at clc, but staying here can be very rewarding. Just
go with what people tell you, and do it _for your own sake_. Otherwise
people will just ignore you.

Bye, and good luck getting your answers!

--
Stan Tobias
mailx `echo si***@FamOuS.BedBuG.pAlS.INVALID | sed s/[[:upper:]]//g`