Ok I got a bit further on this. In the example I attached earlier, if I make the base class non-abstract and remove the BOOST_IS_ABSTRACT(base) specialization, it "works". However this will kill the use of serialization for me :-( Is is_abstract<base> interacting in a mean way with shared_ptr?

PS: Sorry for the post format, can't figure out outlook webmail

-----Original Message-----
From: boost-users-bounces@lists.boost.org on behalf of Sohail Somani
Sent: Tue 3/27/2007 4:12 PM
To: boost-users@lists.boost.org
Subject: Re: [Boost-users] [serialization] Assertion `new_cid==cid'failed-fails on gcc, works on msvc

> -----Original Message-----
> From: boost-users-bounces@lists.boost.org
> [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Robert Ramey
> Sent: Tuesday, March 27, 2007 2:39 PM
> To: boost-users@lists.boost.org
> Subject: Re: [Boost-users] [serialization] Assertion `new_cid
> ==cid'failed-fails on gcc, works on msvc
>
> Note that serialize a poly morphic shared pointer requires a little
> bit of gymnastics.  Check out the "test_shared_ptr" and the
> source of serialization/shared_ptr.hpp

I think I've followed the documentation, but the attached test still
triggers an assertion when deserializing on gcc 3.2.2:

$ ./build/stage/bin/test_serialize
Serialized form: 22 serialization::archive 3 0 1 2 7 derived 1 0
0 1 0
1 10 derived id
test_serialize: basic_iarchive.cpp:466: const
boost::archive::detail::basic_pointer_iserializer*
boost::archive::detail::basic_iarchive_impl::load_pointer(boost::archive
::detail::basic_iarchive&, void*&, const
boost::archive::detail::basic_pointer_iserializer*, const
boost::archive::detail::basic_pointer_iserializer*(*)(const
boost::serialization::extended_type_info&)): Assertion `new_cid == cid'
failed.
Aborted

This assertion does not trigger occur on Visual C++ 7.1. Can you spot if
I'm doing something obviously silly?