Comments by Joe Woodbury (Top 29 by date)

The stipulation was that the file name not be hard coded; presumably it is user entered in some manner. This solution doesn't solve the conversion issue.

How does cout << *num; even compile?

First, this is C++; declare variables just before using them (in the right scope).

Second, arrays are zero based, not one based.

One issue is predecrementing n. If you have a list of 3 nodes. n is 1. Predecrementing it means you reverse the entire list. (And if you have a length of 1, the while loop runs twice whereupon temp is nullptr the second pass and it throws an exception.)

marks in main() is not an array.
i is undefined before the for loop; if marks was an array it too would be undefined.
mx and m are undefined.
Why have both max and mx and min and m?
Also note that in C++, you should define variables as close as possible to where they are used.
I also suggest gathering the marks, passing them to Student and have Student do the calculation (separating the interface from the rest.)

This is valid for C99, but not C++. gcc and clang allow it as a non-standard extension. VC++ does not.

In C++ a scope is the logical space where a named item is visible. The double colon ('::') is a scope resolution operator. In this case, PLAIN_TEST_CLASS is in the scope of CAT_btscanner, so to use PLAIN_TEST_CLASS outside of CAT_btscanner, you need to tell the compiler what scope to look in.

unsigned selected_colour[colours]; is an error. I prefer a vector here.

Putting aside that the code is broken (where does bullet2 come from?) consider that bullet 2 will start at ANIME_END. Further consider that instead of isActive, you'd have: int remaining;

Go from there.

As is, the code doesn't even compile (first::test and second::Create are private).

I fixed that and added a call to Update1 in first::test and it worked.

Is the assumption that grades are capped at 100?

isspace includes CR and LF. Is this intended?

Also note that scanf_s should have an added parameter: scanf_s("%c", &t, sizeof(t));

What is there to figure out? It's a Qt vector class.

Much better.

Small point:

class Test
{
static void main(std::wstring args[])
{
(new CheckNumbers())->check(L"010100111");
}
};

Appears to be C#-like. It leaks. In this case, I suggest declaring an instance of CheckNumbers and call the method.

For std::stack<wchar_t> stack = std::stack<wchar_t>(); drop what's from the equal sign to the semi-colon.

The push loop could be done:

for (wchar_t ch : s)
{
stack.push(ch);
}

It appears correct. I ran it in VC++ 17 and it worked as expected. (I vaguely wonder if there was an issue with the order of execution of cout. Try splitting the output of num to its own line and see.)

I'm puzzled why you keep using 'new' instead of creating vectors of std::string.

There are other issues:

    
    while(std::getline(file, line))
    {
        words[i] = new char[line.length()];
        for(int j = 0; j < line.length(); ++j)
        {
            words[i][j] = line[j];
        }
        words[i][line.length()] = '\0';
        ++i;
    }

You are allocating line.length() and then assigning a terminating zero past the end of the allocation.

bool compairing_words(char * guess, word )
{
   for(int i=0; i < word->length(); i++ )
   {
      if(word[i]==guess)
      {
         return true;
      }
      return false;
   }
}

You return false after first check.

It solves the problem he presented; he wants 20 unique random numbers in an array in the range 1 to 20.

(BTW, I got to that link after reading about over-shuffling, which I'd been doing. Here's one: https://www.i-programmer.info/programming/theory/2744-how-not-to-shuffle-the-kunth-fisher-yates-algorithm.html)

Given your requirements, you may want to consider using the Fisher-Yates inside-out shuffle. (https://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle#The_%22inside-out%22_algorithm)

With C++, take advantage of RAII which helps prevent leaks due to losing track of exit points.

My own preference is to create a class encapsulating the thread and transferring ownership of the data to that class/thread. I'd duplicate the event handle (in a class) in the calling function so I don't need to worry about the lifetime.

Further if hEvent is NULL, you may use strArgument in thread after deleting it in the calling function. (I've only seen CreateEvent fail in stress code designed to use up all handles and the thread creation would likely have failed in that case, but strictly speaking....)

delete strArgument; <==== delete
CloseHandle(hThread);
CloseHandle(strArgument->hEvent); <==== use

Reason for my vote of 5 \n Great code; saved me lots of time and helped me understand the HID interface.

Reason for my vote of 1 \n This is just code without any explanation. It's completely useless.