Re: noncopyable, so why not...

From: williamkempf_at_[hidden]
Date: 2001-03-09 09:50:16

--- In boost_at_y..., Thomas "H." Ptacek <tqbf_at_s...> wrote:
> 
> Am I missing an obvious reason why Boost doesn't include a
> "class heap_only" (that protects the destructor and provides
> a virtual kill() method) and "class stack_only" (that protects the
> class-specific operators new and delete)? Or is it just
> lack of demand?

Personally, I wouldn't want to see such classes.  Why?  Because they 
don't work.  Scott Meyers goes into great detail about these concepts 
in EC++ so I won't restate them here, but the end result is that 
these "techniques" have very limited usefullness.

Bill Kempf