Re: [boost] [valid_ptr] proposing a new type of pointer object that allows validity tracking without memory management and object ownership

Subject: Re: [boost] [valid_ptr] proposing a new type of pointer object that allows validity tracking without memory management and object ownership
From: Gruenke, Matt (mgruenke_at_[hidden])
Date: 2011-03-17 12:30:25

-----Original Message-----
From: boost-bounces_at_[hidden] on behalf of Dan Walters
Sent: Thu 3/17/2011 8:21 AM
To: boost_at_[hidden]
Subject: Re: [boost] [valid_ptr] proposing a new type of pointer object that allows validity tracking without memory management and object ownership
 
On 17 March 2011 03:33, Gruenke, Matt <mgruenke_at_[hidden]> wrote:
...

> Matt <mgruenke_at_[hidden]> wrote:
> the same thing could be accomplished by using weak pointers to an
> object whose refcount had been manually incremented by 1.  When you
> want to kill it, one of the referers could then manually decrement the
> refcount, so that it gets killed when it goes back to idle.
>
> valid_ptr should not be able to release it's target object by design -
> it is not a memory management pointer.

It doesn't matter whether valid_ptr can be used to delete the object.  The point was that in order to fulfill the contract of ensuring that the pointer is valid while used, in a threadsafe fashion, it would need to either hold a lock or a reference count preventing the object from being deleted while an expression involving one of the access operators is being evaluated.

As you point out, valid_ptr doesn't need to solve the problem of thread synchronization in order to be usable in a multithreaded environment.  However, if it doesn't, the question is whether it provides enough value that people would use it.

Matt