Re: OOP::problem?

Bertrand, Thanks for the example. This is truely scarry. I come from a Java OOP background and one of the the key atributes of OOP is encapsulation and for now PHP is limited to programmer obeyance of convention. I think PHP would greatly benefit from adhering to some standard OOP definitions: encapsulation, polymorphism, interfaces, etc... Right now, none of these features are enforceable... but I understand PHP5 promises some tighter restrictions...what you've documented below is just another convention to avoid... "Bertrand Mansion" <bmansion@mamasam.com> wrote in message news:BAAF890D.626E%bmansion@mamasam.com... <matthias.nothhaft@gmx.de> wrote : > Hi all, > > this is my first mail to this list, > but I don't know if this is the right one > for my problem, because it is not really depending > on pear, but I hope there are some php devs, who > can help me? > > Ok, I have a little idea to use it in PEAR... > I promise I'll tell it in my next posting ;-) > > > Take a look at my example: > > <?php > > class a > { > function test($s) > { > $this->myvar = $s; > } > } > > class b > { > var $myvar; > > function b($s) > { > a::test($s); > > echo $this->myvar; > } > } > > $b =& new b("hello world!"); > > ?> > > The output is... "hello world!" > > Now my "problem": Is this a bug or a feature, > that I can access to class variables of any object > by using $this->anyvar in a class::function() ? I think it should be considered as a bug, not a feature so don't rely on it. You might get some better (or more confuse) advices in the php-dev archives. Bertrand Mansion Mamasam