Re: C++ / multiple instances / iostreams

[Detlef Vollmann]

Hans Aberg wrote:
> I believe that the design criterion that BS expressed was that C++ should
> admit efficient implementations, but by no means as optimized as in C. Sed
> the book DEC++ ("Design and Evolution of C++").
No, Bjarne probably never said or wrote such.  He was one of the main
proposers of the C++ Performance Report
 http://www.open-std.org/jtc1/sc22/open/n3646.pdf
(I worked together with him on some parts of that report.)
There are actually cases where C++ is more efficient as C.
SCNR.

>  class semantic_type {
>  public:
>    long number_;
>    std::string text_;
>    my::object object_;
> 
>    semantic_type() : number_(0) {}
>  };
>  #define YYSTYPE semantic_type
> will suffice as semantic type. One converts data using either
> dynamic_cast<T*>, in which case null pointer must be checked, or
> dynamic_cast<T&>, in which case null pointers case an exception to be
> thrown.
I don't really see how you use this, but for dynamic_cast to work the
classes must be related: the only casts that are allowed with
dynamic casts are
 - downcasts
 - crosscasts
 - upcasts
 - identity casts
(The latter two usually make no sense as dynamic_casts, but can happen
in generic code.)
Some compilers allow (and check) dynamic_casts between pointers to
unrelated classes, some even allow dynamic_cast<T *>(void *), but thats
completely implementation defined.

And (as you can also read in the Performance Report) dynamic_casts
are notorically slow in most implementations.

Detlef

-- 
Detlef Vollmann   vollmann engineering gmbh
Linux and C++ for Embedded Systems    http://www.vollmann.ch/