Re: [boost] Re: Sealed/Final implementation.

From: Jonathan Wakely (cow_at_[hidden])
Date: 2005-01-13 08:16:26

On Thu, Jan 13, 2005 at 01:45:59PM +0100, Daniel Wessl?n wrote:

> Jonathan Wakely skrev:
> >On Thu, Jan 13, 2005 at 12:49:33PM +1000, kalin wrote:
> >
> >This is a simpler way of preventing inheritance:
> >
> >  class Finaliser
> >  {
> >  protected:
> >    Finaliser(){}
> >  };
> >
> >  class Final : private virtual Finaliser
> >  {
> >    // rest of class as normal
> >  };
> >
> >[...]
> >There's another problem with Final that lets you derive from it in a
> >class with user-defined constructors, if you know how. Our wiki where
> >I work offers a chocolate bar to anyone who points it out but noone's
> >claimed it from me yet :)
> 
> class WillCompile : public Final, virtual Finaliser {
> public:
> 	WillCompile() {}
> };

Exactly. I was fairly surprised to see Dr Dobbs publishing it without
that caveat.  I guess it works to tell users they _shouldn't_ derive
from it, but as often happens you can get around it if you really want
to, and you can't always stop people who deliberately try to breakthe
rules.

> Will you mail the chocolate bar to me then?

I didn't expect it to be a challenge for anyone reading this list but 
if you want some british chocolate let me know offlist :-)

jon

-- 
"Do what thou wilt shall be the whole of the law"
	- The Doctrines of Thelema