C++ compiler problems with FC2

[Steve Friedman]

I've had a strange problem when trying to compile a C++ program.

On FC2, when using the standard C++ compiler (gcc 3.3.3), the first line 
of the program (#include <iostream>) caused a raft of errors regarding the 
file cmath.  I was then pointed to try gcc 2.96, but this didn't work for 
other reasons.

However, on FC1 (with gcc3.3.1), the program compiled cleanly (and ran on 
FC2).

What gives?

Steve Friedman

[Paul F. Johnson]

Hi,

>  I've had a strange problem when trying to compile a C++ program.
>  
>  On FC2, when using the standard C++ compiler (gcc 3.3.3), the first line 
>  of the program (#include <iostream>) caused a raft of errors regarding the 
>  file cmath.  I was then pointed to try gcc 2.96, but this didn't work for 
>  other reasons. 
Can you post an example? I've written and compiled well over 3000 source
files using gcc 3.3.3 and 3.4 without a hitch on the headers.

You should never use 2.96 for C++ - it's awful! Leave that to the debian
crew.

>  However, on FC1 (with gcc3.3.1), the program compiled cleanly (and
> ran on 
>  FC2).
>  
>  What gives? 
Without seeing code, no idea.

Does this compile for you

#include <iostream>
#include <cmath>

using namespace std;

int main()
{
  cout << "sine 63 = " << sin(63) << endl;
  cout << "log to the base 10 of 54 = " << log10(54) << endl;
  double quick = (sin(53) * sqrt(14.3443) + tanh(22)) / log(61.444);
  cout << "quick = " << quick << endl;
}

This should give you...

sine 63 = 0.167356
log to the base 10 of 54 = 1.80618
quick = 0.606956

Very silly test, but should work without a hitch.

TTFN

Paul
-- 
"Our enemies are innovative and resourceful - and so are we,"
"They never stop thinking about new ways to harm our country and our
people - and neither do we." - George W. Bush, Aug 2004

[Robert L Cochran]

>  Does this compile for you
>  
>  #include <iostream>
>  #include <cmath>
>  
>  using namespace std;
>  
>  int main()
>  {
>    cout << "sine 63 = " << sin(63) << endl;
>    cout << "log to the base 10 of 54 = " << log10(54) <<
> endl;
>    double quick = (sin(53) * sqrt(14.3443) + tanh(22)) / log(61.444);
>    cout << "quick = " << quick << endl;
>  }
>  
>  This should give you...
>  
>  sine 63 = 0.167356
>  log to the base 10 of 54 = 1.80618
>  quick = 0.606956
>  
>  Very silly test, but should work without a hitch.
>  
>  TTFN
>  
>  Paul 
[rlc@bobcp4 cpp]$ g++ -Wall -pedantic -o tsine tsine.cpp
tsine.cpp: In function `int main()':
tsine.cpp:8: error: call of overloaded `sin(int)' is ambiguous
/usr/include/bits/mathcalls.h:66: error: candidates are: double
sin(double)
/usr/include/c++/3.3.3/cmath:522: error:                 long double
   std::sin(long double)
/usr/include/c++/3.3.3/cmath:518: error:                 float
std::sin(float)
tsine.cpp:9: error: call of overloaded `log10(int)' is ambiguous
/usr/include/bits/mathcalls.h:113: error: candidates are: double
log10(double)
/usr/include/c++/3.3.3/cmath:437: error:                 long double
   std::log10(long double)
/usr/include/c++/3.3.3/cmath:429: error:                 float
   std::log10(float)
tsine.cpp:10: error: call of overloaded `sin(int)' is ambiguous
/usr/include/bits/mathcalls.h:66: error: candidates are: double
sin(double)
/usr/include/c++/3.3.3/cmath:522: error:                 long double
   std::sin(long double)
/usr/include/c++/3.3.3/cmath:518: error:                 float
std::sin(float)
tsine.cpp:10: error: call of overloaded `tanh(int)' is ambiguous
/usr/include/bits/mathcalls.h:77: error: candidates are: double
tanh(double)
/usr/include/c++/3.3.3/cmath:582: error:                 long double
   std::tanh(long double)
/usr/include/c++/3.3.3/cmath:574: error:                 float
std::tanh(float)
[rlc@bobcp4 cpp]$ rpm -q gcc
gcc-3.3.3-7
[rlc@bobcp4 cpp]$

[Ed Hill]

On Mon, 2004-08-30 at 21:04, Robert L Cochran wrote:
>  > Does this compile for you
>  > 
>  > #include <iostream>
>  > #include <cmath>
>  > 
>  > using namespace std;
>  > 
>  > int main()
>  > {
>  >   cout << "sine 63 = " << sin(63) << endl;
>  >   cout << "log to the base 10 of 54 = " << log10(54) <<
> endl;
>  >   double quick = (sin(53) * sqrt(14.3443) + tanh(22)) / log(61.444);
>  >   cout << "quick = " << quick << endl;
>  > } 

Yes, explicitly converting those constants to doubles is a good idea:

#include <iostream>
#include <cmath>

using namespace std;

int main()
{
  cout << "sine 63 = " << sin(double(63)) << endl;
  cout << "log to the base 10 of 54 = " << log10(double(54))
<< endl;
  double quick = (sin(double(53)) * sqrt(double(14.3443)) 
    + tanh(double(22))) / log(double(61.444));
  cout << "quick = " << quick << endl;
}

which then compiles cleanly on my FC2 install (gcc 3.3.3).

Ed

-- 
Edward H. Hill III, PhD
office:  MIT Dept. of EAPS;  Rm 54-1424;  77 Massachusetts Ave.
             Cambridge, MA 02139-4307
emails:  eh3(a)mit.edu                ed(a)eh3.com
URLs:    http://web.mit.edu/eh3/    http://eh3.com/
phone:   [phone]
fax:     [phone]

[Robert L Cochran]

On Mon, 2004-08-30 at 21:17, Ed Hill wrote:
>  On Mon, 2004-08-30 at 21:04, Robert L Cochran wrote:
>  > > Does this compile for you
>  > > 
>  > > #include <iostream>
>  > > #include <cmath>
>  > > 
>  > > using namespace std;
>  > > 
>  > > int main()
>  > > {
>  > >   cout << "sine 63 = " << sin(63) << endl;
>  > >   cout << "log to the base 10 of 54 = " << log10(54)
> << endl;
>  > >   double quick = (sin(53) * sqrt(14.3443) + tanh(22)) / log(61.444);
>  > >   cout << "quick = " << quick << endl;
>  > > }
>  
>  
>  Yes, explicitly converting those constants to doubles is a good idea:
>  
>  #include <iostream>
>  #include <cmath>
>  
>  using namespace std;
>  
>  int main()
>  {
>    cout << "sine 63 = " << sin(double(63)) << endl;
>    cout << "log to the base 10 of 54 = " << log10(double(54))
> << endl;
>    double quick = (sin(double(53)) * sqrt(double(14.3443)) 
>      + tanh(double(22))) / log(double(61.444));
>    cout << "quick = " << quick << endl;
>  }
>  
>  which then compiles cleanly on my FC2 install (gcc 3.3.3).
>   
[rlc@bobcp4 cpp]$ emacs tsine2.cpp
[rlc@bobcp4 cpp]$ g++ -Wall -pedantic -o tsine tsine2.cpp
[rlc@bobcp4 cpp]$ ls -al
total 24
drwxrwxr-x  2 rlc rlc 4096 Aug 30 21:19 .
drwxrwxr-x  6 rlc rlc 4096 Aug 30 20:58 ..
-rwxrwxr-x  1 rlc rlc 7739 Aug 30 21:19 tsine
-rw-rw-r--  1 rlc rlc  342 Aug 30 21:19 tsine2.cpp
[rlc@bobcp4 cpp]$ ./tsine
sine 63 = 0.167356
log to the base 10 of 54 = 1.73239
quick = 0.606956

:-)

[Paul F. Johnson]

Hi,

>  Yes, explicitly converting those constants to doubles is a good
> idea:
>  {
>    cout << "sine 63 = " << sin(double(63)) << endl; 
Really speaking for C++, that should be

cout << "sine 63 = " << sin(static_cast<double>(54)) <<
endl;

>  which then compiles cleanly on my FC2 install (gcc 3.3.3). 
But even that shouldn't be needed to compile. It wasn't from 2.96 up!

double sin(double x);

TTFN

Paul
-- 
"Our enemies are innovative and resourceful - and so are we,"
"They never stop thinking about new ways to harm our country and our
people - and neither do we." - George W. Bush, Aug 2004

[Paul F. Johnson]

Hi,

>  [rlc@bobcp4 cpp]$ g++ -Wall -pedantic -o tsine tsine.cpp
>  tsine.cpp: In function `int main()':
>  tsine.cpp:8: error: call of overloaded `sin(int)' is ambiguous 
Without seeing the source, how is a chap to know what you're doing?

TTFN

Paul
-- 
"Our enemies are innovative and resourceful - and so are we,"
"They never stop thinking about new ways to harm our country and our
people - and neither do we." - George W. Bush, Aug 2004

[Robert L Cochran]

If you look again at my original post on this, you can see the source.
It's your source. 

Explicitly converting constants to doubles per Ed Hill's suggestion
allowed the code to compile without error and run.

Bob

On Tue, 2004-08-31 at 02:44, Paul wrote:
>  Hi,
>  
>  > [rlc@bobcp4 cpp]$ g++ -Wall -pedantic -o tsine tsine.cpp
>  > tsine.cpp: In function `int main()':
>  > tsine.cpp:8: error: call of overloaded `sin(int)' is ambiguous
>  
>  Without seeing the source, how is a chap to know what you're doing?
>  
>  TTFN
>  
>  Paul

[Paul F. Johnson]

Hi,

>  If you look again at my original post on this, you can see the
> source.
>  It's your source.  
Right - sorry. It's been a long day trying to get some bloomin' code to
compile.

>  Explicitly converting constants to doubles per Ed Hill's
> suggestion
>  allowed the code to compile without error and run. 
To be honest, that shouldn't really have made a damn of difference -
unless you're using some flag or other when compiling (such as -
pendantic).

TTFN

Paul
-- 
"Our enemies are innovative and resourceful - and so are we,"
"They never stop thinking about new ways to harm our country and our
people - and neither do we." - George W. Bush, Aug 2004

[Robert L Cochran]

>  To be honest, that shouldn't really have made a damn of
> difference -
>  unless you're using some flag or other when compiling (such as -
>  pendantic).
>  
>  TTFN
>  
>  Paul 
> From 'man gcc': 
-pedantic 
Issue all the warnings demanded by strict ISO C and ISO C++; reject all
programs that use forbidden extensions, and some other programs that do
not follow ISO C and ISO C++.  For ISO C, follows the version of the ISO
C standard specified by any -std option used.