Billy O'Neal

http://stackoverflow.com/users/82320/billy-oneal

Malware Response Instructor - BleepingComputer.com

On Sat, Nov 9, 2013 at 8:56 AM, MANTOSH = KUMAR <mantosh4u@gmail.com> wrote:

Hi,

Frankly Speakin= g i also did not understand your points regarding the removing the std::lis= t<T>::size() method from the standard.

However here= is my opinion on your points:

1. It makes most of the operations of=A0std::list slower= ..
How std::list<T>::size() makes most of the operation slowe= r? I mean i did not get how a method (like inline size() )of the class can = affect the other operation.

2. std::forward_list don't have this method. This makes peopl= e very=A0confused
size() is not defined for this container class b= ecause purpose of std::forward_list is different and it should be used for = very small or empty situation. And to optimize and to save one extra word o= f size of object of this class, it only store a single base pointer.

3.the method=A0splice() becomes much slower.
Again i did= not understand how size() can impact the performance of splice() method in= this std::list<T>

4.We know our librar= y couldn't make people very confused. That was=A0why we didn't defi= ne the operator[] in the std::list<>;

operator[] is not defined for std::list<T> not becau= se of confusion and some other reason. Its not there because its not natura= l for std::list<T> and if someone would need this operator[], it woul= d be terribly slow as operator[] is natural for random iterator(std::vector= <T>).

5.The function of this method can=A0be easily implied by programmers=20 themselves. We know this function had been defined clearly=A0in cpp11,=A0bu= t in the cpp98, its time-complexity is not defined clearly. For this=20 reason,=A0in the past we=A0almost didn't=A0use it. So =A0I think this c= hange=20 won't make so many codes invalid.

If you are talking = about the std::list<T>::size() time-complexity then i think its inlin= e function and even it do not require the cost of a simple function call.

I agree with the Daniel comment that its there in the st= andard since the inspection of STL library in C++. So no way it should be r= emoved from the standard.

Alexander Stepenov and Bjr= ane both advocate to use std::vector<T> instead of any other containe= r unless you have very solid reason to do it otherwise. May be you should s= tart thinking to use the std::vector<T> and should measure the perfor= mance of your program.

Thanks
Mantos= h Kumar

=A0
On Sat, Nov 9, 2013 at 10:03 PM, <euloa= nty@live.com> wrote:

1.It makes most of the operations of= =A0std::list slower;

2.std::forward_list don't have this method. This makes people very= =A0confused.

3.the method=A0splice() becomes much slower.
4.We know our l= ibrary couldn't make people very confused. That was=A0why we didn't= define the operator[] in the std::list<>;
5.The function o= f this method can=A0be easily implied by programmers themselves. We know th= is function had been defined clearly=A0in cpp11,=A0but in the cpp98, its ti= me-complexity is not defined clearly. For this reason,=A0in the past we=A0a= lmost didn't=A0use it. So =A0I think this change won't make so many= codes invalid.

--
=A0
---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposals+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/gro= up/std-proposals/.

--
=A0
---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposals+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/gro= up/std-proposals/.

--

---
You received this message because you are subscribed to the Google Groups &= quot;ISO C++ Standard - Future Proposals" group.
To unsubscribe from this group and stop receiving emails from it, send an e= mail to std-proposals+unsubscribe@isocpp.org.
To post to this group, send email to std-proposals@isocpp.org.
Visit this group at http://groups.google.com/a/isocpp.org/group/std-proposals/<= /a>.
--e89a8ff1ce0257ee5d04eac24a1f-- .