From: Carlo Wood (carlo_at_[hidden])
Date: 2002-09-03 20:51:34
On Tue, Sep 03, 2002 at 08:50:13PM -0400, David Bergman wrote:
> One might conceive programmers wanting to use this tool in these
> memory-tight debugging scenarios we have all discussed, without
> necessarily knowing how to implement an allocator that would not cause
> more harm than the debugging process together with Boost.Demangle would
> do good...
You have a point - we might decide not to use std::allocator<char>
as the default. In that case we might start with adding this
allocator to the demangler library and later move it to another
part of boost (as was suggested by Rob Stewart).
-- Carlo Wood <carlo_at_[hidden]>
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk