Thread: Tell me about C/C++

Originally Posted by Ynot

There's certain things that are the "right way" in C, but the "wrong way" in C++ (and vice versa)
Arrays being the biggy

Arrays are exactly the same in C and C++.

He's right about the right and wrong ways being different in C and C++

Critical section in C

Code:

void CriticalMethod(void){
   pthread_mutex_lock(somelock);
   //some critical stuff
   if (boolean){
      pthread_mutex_unlock(somelock);  //have to unlock in every exit path
      return;
   }
   //more critical stuff
   pthread_mutex_unlock(somelock);
}

Critical section in C++

Code:

//Autolock.h
struct AutoLock{
   public:
      AutoLock(pthread_mutex_t * lock){
         pthread_mutex_lock(lock);
         mylock = lock;
      }
      ~AutoLock(void){
         pthread_mutex_unlock(mylock);
      }
   private:
      pthread_mutex_t mylock;
}

Code:

void CriticalMethod(void){
   AutoLock lock(somelock);  //will automatically unlock when it returns
   //some critical stuff
   if (boolean){
      return;
   }
   //More critical stuff
}

***********************


What exactly would you like to know guitarboy, honoured you came to me

I learned C first, then taught myself C++. I would recommend just learning C++, most programming is done in C++, and it's a lot easier. It's also a good thing to understand object oriented programming.

Here's some things to understand, in order
Primitive types: int, long, float, char
Scope
methods and functions
pass-by-reference vs pass-by-value
const
Pointers
Objects (classes and structs)
Constructors and destructors

Feel free to IM me about anything

Originally Posted by ninja9578

Arrays are exactly the same in C and C++.

I know they are the same
but they're hailed in C, and frowned upon in C++

From the man himself
http://www2.research.att.com/~bs/bs_faq2.html#arrays

Oh, okay, you meant that in C++ you're supposed to use STL containers We still use arrays for various things though.

Actually I have a question...

I'm learning C++ and right now I'm reading about ++ and --

The example it gives me is this...

Example 1:

Code:

B=3;
A=++B;
// A contains 4, B contains 4

Example 2:

Code:

B=3;
A=B++;
// A contains 3, B contains 4

I don't really understand the whole ++ and -- and how it affects the A and B variables. Can someone explain how it adds up?

Edit: Also, how does all this make up a video game engine?

increment and decrement
(Ie. add one, or subtract one)
used all the time when looping and keeping count of things

a++ means "return the value of a, and then increment it"
++a means "increment a, then return the value"

subtle (but important) differences, however the differences between the pre & post incrementation are only important if you care about the value of a variable mid-operation

Code:

for (i = 0; i < max; i++)
{

}

and

Code:

for (i = 0; i < max; ++i)
{

}

are identical in operation
(because we do not care about the value of 'i' mid-operation

however,
say we are keeping a count of something that we later need to step back through in reverse

Code:

void increment_something(int var)
{
    old_var = var++;
}

here, we do care about the value of var mid-operation, therefore we store the "old value" of var and increment var

You can of course do this

Code:

void increment_something(int var)
{
    old_var = var;
    var++;
}

then we've split the operation into two
and now we no longer care about the return value of the increment

Originally Posted by Ynot

Code:

for (i = 0; i < max; i++)
{

}

and

Code:

for (i = 0; i < max; ++i)
{

}

are identical in operation
(because we do not care about the value of 'i' mid-operation

Almost correct. In for loops you want to do the latter because in order to return, then increment, it has to keep a copy of it. Not a problem with ints, but with objects it can.

++ and -- are weird to learn.

b = ++a; is like this:
a = a + 1;
b = a;

b = a++; is like this:
b = a;
a = a + 1;

Err...you're using some things in there I haven't learned yet. So I think you just made it more confusing for me...

Edit: Ok that makes it a little easier to understand...

So would it be something like this for the 2nd part?

b = 1;
b = a++;

Would the result be: b = 1, a = 2? Because I already defined b to equal 1 and then a equals b which is 1, then I add a to a to get 2, right?

follow a tutorial on C++
I've always liked this one
http://www.cprogramming.com/tutorial/lesson1.html

It'll take you through the basics, and slowly build up to more complex things

*edit*
Just realised you're not the OP.....

Edit: Never mind. Note to self: Learn to see if there are more pages.

But yeah, I also support you doing some smaller projects first.

Have you given any thought yet to C#?

C is pretty much dead except for embedded solutions, C++ is still a great language and supports object oriented design, but with modern hardware, all but the most time critical tasks can be handled with C#. You also have the ability to call C++ unmanaged code from within wrappers if you have a section that needs to execute at it's fastest.

My suggestion for a compiler is Visual Studio 2005 or 2008 (2003 is OK, but doesn't support the .NET 2.0 framework well). I've been using the Microsoft Cxx compilers since they were introduced and really like it. In fact I'm using VS2005 as we speak.

Originally Posted by Seeker

Have you given any thought yet to C#?

C# is generally just used for Windows exclusive applications, though, and they require the .NET network to run. It's harder to write cross-platform applications. It's not as elegant at making native applications, though you're saved from hunting for external libraries or writing system-call wrappers yourself.

Originally Posted by Seeker

C is pretty much dead except for embedded solutions, C++ is still a great language and supports object oriented design, but with modern hardware, all but the most time critical tasks can be handled with C#. You also have the ability to call C++ unmanaged code from within wrappers if you have a section that needs to execute at it's fastest.

I thought C was commonly used for time-critical operations still, such as graphics modules.

Originally Posted by Seeker

My suggestion for a compiler is Visual Studio 2005 or 2008 (2003 is OK, but doesn't support the .NET 2.0 framework well). I've been using the Microsoft Cxx compilers since they were introduced and really like it. In fact I'm using VS2005 as we speak.

I do however agree that Visual Studio is an excellent set of coding tools. I use VS2008 professional myself (no, it's not pirated... but neither is it bought).

Even unsafe C# code doesn't run as quickly as C++.

It's also very heavily dependent on the Windows runtime libraries. I think C# is an atrocious attempt at making C++ more like Java. C and it's derivatives are not meant to be safe, they're meant to do exactly what you tell them to do.