Ok... I must have unlearned something.
Why is that the following code gives me an error pointing to the vector declaration?
Code:#ifndef COUNTER_H #define COUNTER_H #include <vector> #include <iterator> class counter { public: /*...*/ private: /* ... */ static vector<double> ravg_; //error C2143: syntax error : missing ';' before '<' };
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
Perhaps you should use the fully qualified std::vector<double> ?
Look up a C++ Reference and learn How To Ask Questions The Smart WayOriginally Posted by Bjarne Stroustrup (2000-10-14)
oh my!
Thanks laserlight
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
Alternatively, you could just declare "using namespace std" in your header file, so you don't need to prefix everything with std::
Bad idea, the contextual clues (i.e. header guards) tell us this is a header file we are dealing with.Alternatively, you could just declare "using namespace std" in your header file, so you don't need to prefix everything with std::
Look up a C++ Reference and learn How To Ask Questions The Smart Way
So? Why does that prevent you from declaring a namespace? I'm not asking to be rude, I'd actually like to know because whenever I use a std::string or std::vector, I just declare the namespace so I don't need the std:: prefix all the time.
because whatever file includes that header file forces the using declaration, you should use only use std:: explicitly in the header file instead of blanketing it.
But is that really a problem? I mean logically if you're using std::strings or whatever in a header file, you're also using it in the accompanying source file, in which case declaring the using namespace std in the header saves you from having to declare it in the source file.
Even if you're not using it anywhere else, are there problems caused by saying you're using the std namespace?
Well, a using declaration does not declare a namespace, but rather brings the names in the namespace (or the given name, if it covers only one name) into the current scope.Why does that prevent you from declaring a namespace?
This is problematic in header files, since it means that files that include the header file would also have these names introduced. This pretty much defeats the purpose of namespaces.
A possible workaround is to have the using directive in a local scope, e.g. in a function.
Look up a C++ Reference and learn How To Ask Questions The Smart Way
but what if you don't use "using std::string;" in the main file that also includes the header file. Just because A uses B doesn't mean C has anything to do with it.
Yes. the using directive is best left outside header files. It brings into scope the whole std namespace, which is bound to sooner or later create name collisions. Even if it does not, it is bad style. Scope shoud always be restricted to the best of our abilities to the explicitly declared names in our code.
I prefer to use the scope operator as laserlight suggested. But a viable option for header files can also be the using declaration. An using declaration such as using std::cout is much safer.
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
Imagine an header file that brings into scope the whole of std by means of a using directive (using namespace std).Even if you're not using it anywhere else, are there problems caused by saying you're using the std namespace?
Now... this header file defines a class you use on several projects at your workplace. One of your workmates just created a piece of code with a variable name cout (for instance long cout; ). He then goes on to include your header file. Can you see the problem?
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.
I forget which thread it was but there was also something that said if you use a namespace declaration in the parameter list of a function, then the return type and other parameters are automatically declared within a particular namespace. Or something like it. It had some name.
7. It is easier to write an incorrect program than understand a correct one.
40. There are two ways to write error-free programs; only the third one works.*
I dont think indigo0086 is referring to that, since this mechanism apparently involves using "a namespace declaration in the parameter list of a function", and that certainly isnt ADL.Koenig lookup ?
Look up a C++ Reference and learn How To Ask Questions The Smart Way
