[boost] Re: [policy_ptr] multi-inheritance failure

From: Larry Evans (cppljevans_at_[hidden])
Date: 2003-12-12 09:57:59

On 12/12/2003 08:25 AM, Larry Evans wrote:
 > On 12/12/2003 07:57 AM, Peter Dimov wrote:
 >
 >> David B. Held wrote:
[snip]
 > but then the whole point of using smart_ptr's is that
 > the programmer does NOT KNOW which pointer will be the
 > last to be destructed.  Of course maybe it wouldn't be

But I can see another point.  Maybe David's implicit point was that
the programmer shouldn't have to pay for what he doesn't use, i.e. the
extra memory in shared_ptr, and the programmer is willing to impose on
himself the discipline of either using virtual DTOR's or assuring the
last ~smart_ptr call is for the most derived class.  This of course is
precisely the advantage of policy pointers, i.e. the programmer has a
greater ability to tailor the smart_ptr to his needs.