Re: [boost] Proposal: Monotonic Containers

Subject: Re: [boost] Proposal: Monotonic Containers
From: Christian Schladetsch (christian.schladetsch_at_[hidden])
Date: 2009-06-15 16:36:22

Yes, I was going to suggest boost::allocator::monotonic, rather than
boost::monotonic::allocator.

More on this later.

On Tue, Jun 16, 2009 at 5:26 AM, Ross Levine <ross.levine_at_[hidden]> wrote:

> On Mon, Jun 15, 2009 at 12:11 PM, Simonson, Lucanus J <
> lucanus.j.simonson_at_[hidden]> wrote:
>
> >
> >
> > I think an allocator that uses your chain buffer for storage and is safe
> > could be a good thing.  I think that the allocator can be implemented in
> a
> > way that it doesn't need non-static data members.  The chain is in and of
> > itself not interesting as a replacement for deque, but as a memory buffer
> > for a good allocator I think it could find its niche.  I'd like to see
> the
> > memory buffer implement object deallocation and memory reuse, and perhaps
> > thread saftey (maybe optional thread safety enabled by default).  I
> almost
> > coded up a proposal for such an allocator based on a deque when I last
> > replied to you, but I see you've thought way further along those lines
> > without my prompting.
> >
>
> Now that is a great idea! A boost::allocators library would be a great
> addition. Then everyone wins, because we could have a cross-platform
> stack_based_allocator and monotonic_allocator and the like.
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>