Re: Some articles on misconceptions about C++
On Mar 6, 9:23 pm, Stefan Chrobot <jan...@op.pl> wrote:
James Kanze napisa?(a):
Maciej Sobczak wrote:
Sushrut Sardeshmukh wrote:
C++, for example, has better support of OO than does Java,
despite that being the only paradigm Java claims to support.)
I have no intensions to challenge you.
However, I am curious to know how C++ supports OO better than Java.
Cannot speak for James, but my guess is multiple inheritance.
Exactly.
And it would be really useful in Java when coding GUI by hand. If you
want your JPanel to accept mouse and keyboard events, some class has to
implement MouseListener and KeyboardListener interfaces. But say you are
interested in "mouse moved" and "key pressed" events only. Too bad, you
have to implement all 8 methods, even if some are NOOPs. True, there are
MouseAdapter and KeyboardAdapter classes that do that NOOPs and which
one can new with anonymous classes, but I'd prefer - what I would call
"classic OO" - to use MI with those adapters.
If I understand you right, what you want is mixins. A useful OO
technique, that Java explicitly doesn't support.
To be fair to Java, however, your problem is relativly easily
solved with Java's inner classes. You register an inner class
derived from the Adaptor, rather than "this". It's a bit more
work, but not the end of the world.
Could C++ also be modified to support something like inner
classes? Probably, but I'm not sure I'd favor it. My
experience is that in all but the simplest cases, they lead to
confusion, since functions in the inner class logically belong
to two different classes (and have two "this" pointer). I
imagine that with "lambda classes", or whatever they are called,
we could end up with the advantages, without the inconvenients.
Also, IMHO, the C++'s ctor-dtor class model is more natural.
It doesn't change much when you only consider "OO" objects,
which are dynamically allocated. It just means that destruction
has a special syntax (as opposed to being just a conventionally
named function, like "dispose()"). Where the C++ model shines
is when the destructor is automatically called, for local
variables. (But objects on the stack isn't "OO"---lucky for us,
C++ is multi-paradigm, and not pure "OO".)
But
probably we're going to have a GC in C++09. Oh dear.
What's the relationship? GC is just another tool in your
toolkit. It certainly doesn't change the C++ ctor-dtor class
model in any way.
--
James Kanze (GABI Software) email:james.kanze@gmail.com
Conseils en informatique orient?e objet/
Beratung in objektorientierter Datenverarbeitung
9 place S?mard, 78210 St.-Cyr-l'?cole, France, +33 (0)1 30 23 00 34
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]