Vlad,

I never compared OOP with C++, on the contrary, I'm saying that C++ violates OOP concepts, and the use of pointers and structures is one example (they are not objects, as many other C++ "components"?). (For me to say that they are pure objects is like saying that a brick is a pure house).

And we should not forgot the the coiner (is that the word for the person who invented a phrase? of the OO term, Allan Kay is also one of the creators of Smalltalk, so I fail to see how can Smalltalk had failed in implementing it :)

Talking about Allan Kay, I'll quote him talking about C++:
"I invented the term Object-Oriented, and I can tell you I did not have C++ in mind."

Hugo




>Hi!
>
>OOP is not a language so you cannot compare OOP and C++ language. OOP is a concept, and C++ is a realisation of that concept. It does not matter. You can use Assembler language, but still work in OOP bounds and follow all OOP concepts, writing OOP applications. So pointers, structures (that are nothing else but pure objects) and other C stuff in C++ don't spoil OOP in it.
>
>What I'm talking about is just a realisation of OOP concepts in different languages. OOP language provides a helpful tools to model most OOP concepts with a few of declarations and lines in the language. This is not possible with any language we have now, that is why OOP have no full representation in any language even such as SmallTalk. You can only compare a quality of realisation between different languages.
>
>>Vlad,
>>
>>> C++ is the best for OOP, not Delphi.
>>
>>Personally, I don't beleive that any language that makes the use of structures and pointers (among other things, but specially pointers) relevant can be considered OOP. For a pure OOP implementation look at Smalltalk, Eiffel and a few others (Python?).
>>
>>Of course I'm not making any judgement about the power or the use of such elements, but they are definetely not OOP and its use encouraged by the language itself.