Good videos. One clarification, in this ex. the pointer variable "p" will be created on stack so its local in main() ,but the chunk of memory will be allocated on heap. So once we get exception , the pointer "p" goes out of scope hence memory leak because we lost pointer "p" and thus lost the access to chunk of memory on heap.
good video explaining the idea behind unique_ptr and how to use it (it's like the non-managed C++ version of garbage collection for raw pointers). The only issues I had were that your examples can be a bit hard to follow at first. You know what you intend to show but it's a little confusing, even for a veteran programmer, to follow what you're doing when passing ownership around, releasing pointers, etc. But then that's generally difficult to follow in C++ at the best of times. Maybe more descriptive variable names, including the initial values, e.g. unique_ptr fooptr1_10 = make_unique(10); Then when you e.g. unique_ptr fooptr2_10 = std::move(fooptr1_10) you can see that the contents of fooptr1_10 were moved. Dumping the values of objects after moves to show that the object really was transferred from one place to another might help people unfamiliar with the concept too.
Dude you are making good videos. Just one suggestion, it is hard to read on mobile screen whatever you have written on the terminal. Better if you can increase the font size or zoom the window.
Sir I want to do specialization in a language. Which language should I prefer for campus placement and for future . Is java better or c++ or other one.
Great video, easy to understand. i still got a problem tough.. i'm trying to allocate a private member vector of unique_ptr to a base class object ( std::vector m_objects ; ) and than to pushback derived classes into it trough a void function that gets nothing and returns nothing but it knows the vector member in the private section of the class because the cpp file belongs to its header file so i use ( m_objects.push_back(std::make_shared (); ) but i always get these linkage errors 2019 and 1120 and i don't know what to do anymore..
I am getting this error on doing this - std::unique_ptr t1 = new Test (2000) ; error : conversion from 'Test*' to non-scalar type 'std::unique_ptr' requested Please help!!!!! Why doesn't = work here?
I thought destructors are called automatically when scope ends. Why is destructor not called in your first run? It has something to do with stack/heap allocation? I'm guessing that when you declare class object with new, only the reference is destroyed at end of scope, not the object. And destructors are called only when the object is destroyed.
universal initialization using {} is the preferred method. The C++ committee has gone through many attempts to get initialization of all types of object correct, and it still has problems in certain circumstances. Nicolai Josuttis (C++ committee member and prolific book writer and conference presenter) informs us that {} should be used everywhere, even to the extent of doing e.g. for (int i{0}; i < limit; ++i) ... I think most people (including me) will continue to write for (int i = 0; ...) in that instance but everywhere else, for local, member, and global/static variable initialization, {} is a good habit to get into.
@sara That's what my point is. In your example 2 your program is not exiting main, while 1 will keep looping indefinitely inside main. My doubt is Int main() { Int *p = new int(10); } Is it a memory leak after program exits? I believe no! OS takes away memory allocated to your a.out
@@13taras The point of calling delete everytime is to make most efficient use of memory allocated to the program. If we free memory, the freed memory can then be reused by our program and we do not run out of memory.
What happen if I do delete p after p3.get(): Foo *p = p3.get(); delete p; what happens with p3? Is it still point to the object in which it was returned?
It still points to the adress in memory, that you deleted. Your program will crash when the p3 gets out of scope and the deleter is called, cause there is nothing left to delete.
@@JayAnAm Did you try to execute? It doesn't crash. p is a pointer of type Foo *. It is not a unique pointer, and it holds the address to the object of p3, so p and p3 are 2 different entities. Did you ever see a destructor called for a object pointer being deleted?
std::move : cast object into rvalue get: returns the raw poiner, wrapped by smart pointer reset: Replaces the object managed by the smart pointer release : Releases the ownership of the managed object
Not bad, but you have to be a bit more careful when explaining things like move, espcially when your viewers are not familiar with lvalues and rvalues. std::move doesnt move anything, it is casting into rvalue. And p1 is not a nullpointer after std::move, it is an empty object, just the raw pointer that is wrapped by p1 is nullptr.
@@JayAnAm I took Rupesh (?) to mean that the contained raw pointer was set to nullptr after the move, but then I know the subject matter. Making it crystal clear and unambiguous to less experts wouldn't hurt.
I think you no need to be over conscious of your voice part..... because of this drama, your video content does not come out well. Concentrate on content and good explanation, instead of making singing kind of voice.
Good videos. One clarification, in this ex. the pointer variable "p" will be created on stack so its local in main() ,but the chunk of memory will be allocated on heap. So once we get exception , the pointer "p" goes out of scope hence memory leak because we lost pointer "p" and thus lost the access to chunk of memory on heap.
good video explaining the idea behind unique_ptr and how to use it (it's like the non-managed C++ version of garbage collection for raw pointers). The only issues I had were that your examples can be a bit hard to follow at first. You know what you intend to show but it's a little confusing, even for a veteran programmer, to follow what you're doing when passing ownership around, releasing pointers, etc. But then that's generally difficult to follow in C++ at the best of times. Maybe more descriptive variable names, including the initial values, e.g. unique_ptr fooptr1_10 = make_unique(10); Then when you e.g. unique_ptr fooptr2_10 = std::move(fooptr1_10) you can see that the contents of fooptr1_10 were moved. Dumping the values of objects after moves to show that the object really was transferred from one place to another might help people unfamiliar with the concept too.
Dude you are making good videos. Just one suggestion, it is hard to read on mobile screen whatever you have written on the terminal. Better if you can increase the font size or zoom the window.
Sure i will try to do something about it. Thanks man..
One of the best explanations!
This channel just get a new subscriber.
Thanks man!!
Obviously nice video but i like this color combination with white and *cppnuts*
Rupesh Bhaiya thanks
Good..😀
Very nice explaination sir thank you so much sir
Thanks man..
Best tutorial on the topic i have ever seen, thank you
Thanks mate..
Best expalanation of smart pointer i've ever seen
Thanks man..
Nice video but it will be wonderful if you can add some example of storing the array of object in unique pointer.
Very nice explanation. Cleared the concept well. Subscribed!
Sir how move and release are different?
please do a video with implementing our own unique pointer
Sir I want to do specialization in a language. Which language should I prefer for campus placement and for future . Is java better or c++ or other one.
Campus placement then java/cpp both are good.
And after that see what companies gives you after you join. And master that language.
@@CppNuts thank you so much 🙏🙏 for your suggestion.
Great video, easy to understand. i still got a problem tough.. i'm trying to allocate a private member vector of unique_ptr to a base class object ( std::vector m_objects ; ) and than to pushback derived classes into it trough a void function that gets nothing and returns nothing but it knows the vector member in the private section of the class because the cpp file belongs to its header file so i use ( m_objects.push_back(std::make_shared (); ) but i always get these linkage errors 2019 and 1120 and i don't know what to do anymore..
Thanks a lot for nice explanation.. Your videos are really good.. Thank you..
Glad you like them!
Thank you so much.
More power to you ✌, keep going.
what all extensions are you using with the Visual studio code editors ? Thanks
std::unique_ptr p4 = p3.release() doesn't work. So release() is for regular pointer only right?
p2.reset(p4),the previous object belonging to p2 is deleted and the p4 becomes a null pointer ,is that correct?
In 9:02, what is the problem if I make single pointer pointed by 2 unique_ptr?
I am getting this error on doing this -
std::unique_ptr t1 = new Test (2000) ;
error : conversion from 'Test*' to non-scalar type 'std::unique_ptr' requested
Please help!!!!! Why doesn't = work here?
I thought destructors are called automatically when scope ends. Why is destructor not called in your first run? It has something to do with stack/heap allocation? I'm guessing that when you declare class object with new, only the reference is destroyed at end of scope, not the object. And destructors are called only when the object is destroyed.
First example will be in heap bcs of dynamic allocation bcs of pointer
I'm confused how were you able to set explicit foo (int x): x{x} {} with curly brackets instead of () ?
This is also one of the way to do it.
universal initialization using {} is the preferred method. The C++ committee has gone through many attempts to get initialization of all types of object correct, and it still has problems in certain circumstances. Nicolai Josuttis (C++ committee member and prolific book writer and conference presenter) informs us that {} should be used everywhere, even to the extent of doing e.g. for (int i{0}; i < limit; ++i) ... I think most people (including me) will continue to write for (int i = 0; ...) in that instance but everywhere else, for local, member, and global/static variable initialization, {} is a good habit to get into.
Can someone explain me what happens to p4 in line 45 with p2.reset(p4); ? p2 has the ownership of the object of p4 then?
Yea you can, Just use move(ptr)😎
Is it a memory leak if main exits? I believe no, because if main exits, the entire program exits and memory is freed and given back to OS.
@sara That's what my point is. In your example 2 your program is not exiting main, while 1 will keep looping indefinitely inside main. My doubt is
Int main() {
Int *p = new int(10);
}
Is it a memory leak after program exits?
I believe no! OS takes away memory allocated to your a.out
@@13taras The point of calling delete everytime is to make most efficient use of memory allocated to the program. If we free memory, the freed memory can then be reused by our program and we do not run out of memory.
can i ask what benefit run code, so importently in linux , as u even using virtual box !
I like to work in linux, it's just a thing.
What happen if I do delete p after p3.get():
Foo *p = p3.get();
delete p;
what happens with p3? Is it still point to the object in which it was returned?
It still points to the adress in memory, that you deleted. Your program will crash when the p3 gets out of scope and the deleter is called, cause there is nothing left to delete.
@@JayAnAm Did you try to execute? It doesn't crash. p is a pointer of type Foo *. It is not a unique pointer, and it holds the address to the object of p3, so p and p3 are 2 different entities. Did you ever see a destructor called for a object pointer being deleted?
What is the difference btw move,get, reset and release..
std::move : cast object into rvalue
get: returns the raw poiner, wrapped by smart pointer
reset: Replaces the object managed by the smart pointer
release : Releases the ownership of the managed object
How come a person forgets to call delete. That is so weird.
Believe me It is good candidate to be forgotten.
Not bad, but you have to be a bit more careful when explaining things like move, espcially when your viewers are not familiar with lvalues and rvalues. std::move doesnt move anything, it is casting into rvalue. And p1 is not a nullpointer after std::move, it is an empty object, just the raw pointer that is wrapped by p1 is nullptr.
I don't think its defined that after performing move what remains.
Actually should be is undefined behavior.
@@CppNuts The state of the p1 after move (which is actually a cast) is unspecified, but it is not nullptr.
@@JayAnAm yaa.. Is it like i mentioned it will be nullptr.
@@CppNuts No. Set a breakpoint and see it by yourself in the debugger - the raw pointer is a nullptr, not the unique pointer.
@@JayAnAm I took Rupesh (?) to mean that the contained raw pointer was set to nullptr after the move, but then I know the subject matter. Making it crystal clear and unambiguous to less experts wouldn't hurt.
If I have a class with unique_ptr as a member, How to write a copy constructor for this class???
class A
{
std::unique_ptr< int > up_;
public:
A( const A& a ) {};
}
get deleter video ?
Many things are wrongly explained(at 5:55) in this video.. So please avoid it.
Like what?
I think you no need to be over conscious of your voice part..... because of this drama, your video content does not come out well. Concentrate on content and good explanation, instead of making singing kind of voice.
How many videos did you watched?
And what is the problem is singing?