&gt; Is there an implementation where the underlying base type for ssize_t
&gt; is not the same as for ptrdiff_t?

x86-16 with the large memory model.  Pointers are far (32-bit), but individual objects are limited to one segment (so `size_t` is allowed to be 16-bit).

The Open Group Base Specifications Issue 7, IEEE Std 1003.1, 2013 Edition, description of [`&lt;sys/types.h&gt;`][1] says:

&gt;  The type ssize_t is capable of storing values at least in the range [-1, SSIZE_MAX].

In other words, `ssize_t` is signed, _but_ the set of negative values it can represent may be limited to just {-1}.

A `ptrdiff_t`, on the other hand, is guaranteed to have a more symmetric positive/negative range.

I admit that in practice, it doesn&#39;t seem likely that `ssize_t` would be this limited in the negative range, but it is possible.

Of course, another difference is that `ptrdiff_t` is available whenever you&#39;re programming in standard C or C++, but `ssize_t` may not be available unless you&#39;re targeting a standard POSIX system.

[1]: http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_types.h.html


When I was working in enterprise, we used multiple projects in the solution - projects for UI, business logic, data access, database and printing.  Now I&#39;m in a new enterprise and the manager tells me that I don&#39;t have to make all these projects, but that I have to make them into seperate directories in one project in the solution.

I just want to know if I have to convince him to use multiple projects! 

What are the benefits of multiple projects and one solution?

Are there any lossless compression methods that can be applied to floating point time-series data, and will significantly outperform, say, writing the data as binary into a file and running it through gzip?

Reduction of precision might be acceptable, but it must happen in a controlled way (i.e. I must be able to set a bound on how many digits must be kept)

I am working with some large data files which are series of correlated `double`s, describing a function of time (i.e. the values are correlated).  I don&#39;t generally need the full `double` precision but I might need more than `float`.

Since there are specialized lossless methods for images/audio, I was wondering if anything specialized exists for this situation.

**Clarification:**  I am looking for existing practical tools rather than a paper describing how to implement something like this.  Something comparable to gzip in speed would be excellent. 

Compressing floating point data

The C standard (ISO/IEC 9899:2011 or 9899:1999) defines a type `ptrdiff_t` in `&lt;stddef.h&gt;`.

The POSIX standard (ISO/IEC 9945; IEEE Std 1003.1-2008) defines a type `ssize_t` in `&lt;sys/types.h&gt;`.

* What is the difference between these types (or why were both deemed necessary)?
* Is there an implementation where the underlying base type for `ssize_t` is not the same as for `ptrdiff_t`?

What is the difference between ssize_t and ptrdiff_t?

<p>The C standard (ISO/IEC 9899:2011 or 9899:1999) defines a type <code>ptrdiff_t</code> in <code>&#x3C;stddef.h></code>.</p>
<p>The POSIX standard (ISO/IEC 9945; IEEE Std 1003.1-2008) defines a type <code>ssize_t</code> in <code>&#x3C;sys/types.h></code>.</p>
<ul>
<li>What is the difference between these types (or why were both deemed necessary)?</li>
<li>Is there an implementation where the underlying base type for <code>ssize_t</code> is not the same as for <code>ptrdiff_t</code>?</li>
</ul>


I am writing a cross platform program. I want this one program to run under both Windows and Linux, so I have two different code segments for the two platforms. If the OS is Windows, I want the first code segment to run; if it&#39;s Linux, then I want the second code segment to run.
  
So I wrote the following code, but it gets an error while building both on Windows and on Linux. What should I do to solve it?

    #ifdef __unix__                    /* __unix__ is usually defined by compilers targeting Unix systems */

        #define OS_Windows 0
        #include &lt;unistd.h&gt;
        #include &lt;stdlib.h&gt;
        #include &lt;stdio.h&gt;
        #include &lt;string.h&gt;

    #elif defined(_WIN32) || defined(WIN32)     /* _Win32 is usually defined by compilers targeting 32 or   64 bit Windows systems */

        #define OS_Windows 1
        #include &lt;windows.h&gt;
        #include &lt;stdio.h&gt;
        #include &lt;tchar.h&gt;
        #define DIV 1048576
        #define WIDTH 7

    #endif

    int main(int argc, char *argv[])
    {
        if(OS_Windows)
        {
            MEMORYSTATUSEX statex;
            statex.dwLength = sizeof (statex);
            GlobalMemoryStatusEx (&amp;statex);

            _tprintf (TEXT(&quot;There is  %*ld %% of memory in use.\n&quot;),
                        WIDTH, statex.dwMemoryLoad);

        }

        else if(!OS_Windows) // if OS is unix

        {
            char cmd[30];
            int flag = 0;
            FILE *fp;
            char line[130];
            int memTotal, memFree, memUsed;

            flag=0;
            memcpy (cmd,&quot;\0&quot;,30);
            sprintf(cmd,&quot;free -t -m|grep Total&quot;);
            fp = popen(cmd, &quot;r&quot;);
            while ( fgets( line, sizeof line, fp))
            {
                flag++;
                sscanf(line,&quot;%*s %d %d %d&quot;,&amp;TotalMem, &amp;TotalUsed, &amp;TotalFree);
            }
            pclose(fp);

            if(flag)
                printf(&quot;TotalMem:%d -- TotalUsed:%d -- TotalFree:%d\n&quot;,TotalMem,TotalUsed,TotalFree);
            else
                printf(&quot;not found\n&quot;);

        }

        return 0;
    }


Detect Windows or Linux in C, C++

I&#39;m trying to make a simple fibonacci calculator in C but when compiling `gcc` tells me that I&#39;m missing the pow and floor functions. What&#39;s wrong?

Code:

    #include &lt;stdio.h&gt;
    #include &lt;math.h&gt;
    
    int fibo(int n);
    
    int main() {
            printf(&quot;Fib(4) = %d&quot;, fibo(4));
            return 0;
    }
    
    int fibo(int n) {
            double phi = 1.61803399;
    
            return (int)(floor((float)(pow(phi, n) / sqrt(5)) + .5f));
    }

Output:

    gab@testvm:~/work/c/fibo$ gcc fib.c -o fibo
    /tmp/ccNSjm4q.o: In function `fibo&#39;:
    fib.c:(.text+0x4a): undefined reference to `pow&#39;
    fib.c:(.text+0x68): undefined reference to `floor&#39;
    collect2: ld returned 1 exit status



Undefined reference to `pow&#39; and `floor&#39;

I was just reading 

**ISO/IEC 9899:201x  Committee Draft — April 12, 2011**
 
in which i found  under 5.1.2.2.3 Program termination

    ..reaching the } that terminates the main function returns a value of 0. 

 
it means if you don&#39;t specify any return statement in `main()`, and if the program runs successfully, then at the closing brace } of main will return 0.

But in the following code i don&#39;t specify any return statement, yet it does not return 0 

    #include&lt;stdio.h&gt;
    int sum(int a,int b)
    {
    return (a + b);
    }
    
    int main()
    {
        int a=10;
        int b=5;
        int ans;    
        ans=sum(a,b);
        printf(&quot;sum is %d&quot;,ans);
    }

   

compile 

    gcc test.c  
    ./a.out
    sum is 15
    echo $?
    9          // here it should be 0 but it shows 9 why?

Why main does not return 0 here?

What would be the easiest way to create a C compiler for a custom CPU, assuming of course I already have an assembler for it?

Since a C compiler generates assembly, is there some way to just define standard bits and pieces of assembly code for the various C idioms, rebuild the compiler, and thereby obtain a cross compiler for the target hardware?

Preferably the compiler itself would be written in C, and build as a native executable for either Linux or Windows.


**Please note:** I am _not_ asking how to write the compiler itself. I did take that course in college, I know about general compiler-compilers, etc. In this situation, I&#39;d just like to configure some existing framework if at all possible. I don&#39;t want to modify the language, I just want to be able to target an arbitrary architecture. If the answer turns out to be &quot;it doesn&#39;t work that way&quot;, *that* information will be useful to myself and anyone else who might make similar assumptions.

How to create a C compiler for custom CPU?

I have the following program 

    #include &lt;stdio.h&gt;
    
    int main(void)
    {
        unsigned short int length = 10; 
                
        printf(&quot;Enter length : &quot;);
        scanf(&quot;%u&quot;, &amp;length);
    
        printf(&quot;value is %u \n&quot;, length);
    
        return 0;
    }

Which when compiled using `gcc filename.c` issued the following warning (in the `scanf()` line). 

`warning: format ‘%u’ expects argument of type ‘unsigned int *’, but argument 2 has type ‘short unsigned int *’ [-Wformat]`

I then referred the `C99 specification - 7.19.6 Formatted input/output functions` and couldn&#39;t understand the correct format specifier when using the length modifiers (like `short`, `long`, etc) with `unsigned` for `int` data type.

Is `%u` the correct specifier `unsigned short int`? If so why am I getting the above mentioned warning?! 

EDIT: 
Most of the time, I was trying `%uh` and it was still giving the warning. 

What is the format specifier for unsigned short int?

I want to kill all processes that I get by:

    ps aux | grep my_pattern

How to do it?

This does not work:

    pkill my_pattern

 

How to kill all processes with a given partial name?

I&#39;m a little confused by &quot;C standard lib&quot; and &quot;C POSIX lib&quot;, because I found that, many header files defined in &quot;C POSIX lib&quot; are also part of &quot;C standard lib&quot;. 

So, I assume that, &quot;C standard lib&quot; is a lib defined by ANSI C organization, and there are different implementation on different platforms (Win32/Unix-like), and &quot;C POSIX lib&quot; is just a implementation for &quot;C standard lib&quot; on Unix-like OSes, right?

But &quot;C POSIX lib&quot; contains some headers not specified in &quot;C standard lib&quot;, such as `&lt;sys/types.h&gt;`, `&lt;sys/wait.h&gt;`, and `&lt;pthread.h&gt;`.

Take `&lt;pthread.h&gt;` as an example, I presume its &quot;C standard lib&quot; counterpart is `&lt;threads.h&gt;`, then if I want to write a multi-threaded program on Linux, which header file should I include, `&lt;pthread.h&gt;` or `&lt;threads.h&gt;`?

Difference between C standard library and C POSIX library

When loading shared libraries in Windows, `LoadLibrary()` call causes `DllMain` in library to execute for each new process and thread library attaches to, and for each process and thread library deattaches from.

Is there similar mechanism for Mac OS X, Linux and possibly other POSIX-compatible OSs?

Automatically executed functions when loading shared libraries

I wrote a program for an assignment which is supposed to print its output to stdout. The assignment spec requires the creation of a Makefile which when invoked as `make run &gt; outputFile` should run the program and write the output to a file, which has a SHA1 fingerprint identical to the one given in the spec. 

My problem is that my makefile:

    ...
    run:
         java myprogram

    


 also prints the command which runs my program (e.g. java myprogram) to the output file, so that my file includes this extra line causing the fingerprint to be wrong.

Is there any way to execute a command without the command invocation echoing to the command line?


Suppress echo of command invocation in makefile?

I found this piece of code in /etc/cron.daily/apf    

    #!/bin/bash  
    /etc/apf/apf -f &gt;&gt; /dev/null 2&gt;&amp;1  
    /etc/apf/apf -s &gt;&gt; /dev/null 2&gt;&amp;1  


It&#39;s flushing and reloading the firewall.   
I don&#39;t understand the `&gt;&gt; /dev/null 2&gt;&amp;1` part.  

What is the purpose of having this in the cron? It&#39;s overriding my firewall rules. 
Can I safely remove this cron job?  

Content Type	Original Author	Original Content on Stackoverflow
Question	Jonathan Leffler	View Question on Stackoverflow
Solution 1 - C	dan04	View Answer on Stackoverflow
Solution 2 - C	Adrian McCarthy	View Answer on Stackoverflow

What is the difference between ssize_t and ptrdiff_t?

C Problem Overview

C Solutions

Solution 1 - C

Solution 2 - C

Compressing floating point data

What are the benefits of multiple projects and one solution?

Attributions