[boost] Re: Re: Re: shared_ptr and serialization

From: Robert Ramey (ramey_at_[hidden])
Date: 2004-11-20 12:33:25

"Dave Gomboc" <dave_at_[hidden]> wrote in message
news:Pine.LNX.4.44.0411200014060.15771-100000_at_opal.cs.ualberta.ca...
> > Honestly, I don't need to serialize std::tr1::shared_ptr - why would
> > any of us ever use it as long as we have boost::shared_ptr which (with
> > serialization is a superset of std::tr1::shared_ptr.)
>
> That's a completely bogus argument.

what's bogus about it?  As a practical matter if one has boost serialization
one has boost::std_ptr.

A question - is it possible to add a facility that would convert between
std::tr1:shared_ptr and boost::shared_ptr via their public interfaces?  My
guess its that its not.  I'm just curious.

> > I understand your objection to boost::serialization depending upon the
> > implementation details of boost::shared_ptr.   I just don't see any way
to
> > address this other than making big changes to either the serialization
> > libary or shared_ptr.  Even then, its not obvious to me that its even
> > possible.
>
> Doesn't Peter's own serialization library succeed in not requiring special
> access to boost::shared_ptr? If so, it is an existence proof that there's
> a way to do it,

I recall only seeing portions of this library in this discussion thread.  I
don't know how it compares to the current boost serialization library.  So
I'm not sure that constitutes an existence proof.

>  which I think means that big changes may need to be made
> to Boost.Serialization.

Hmmm - at what cost? what if this requirement conflicted with other
requiremetns of the serialization library? For example, what if serializing
though the public interface of shared_ptr required eliminating ability to
serialize polymorphic pointers? would this be worth it?  How can one reach
such a conclusion without considering issues such as this?

> It will be a reasonable expectation for a
> Boost.Serialization user to serialize std::tr1::shared_ptrs in the future.

This is not in dispute - the question is it doable iand if so, at what cost.
So far this is the only std::library component that cannot be serialized via
it's public interface. I doubt that any serialization library available
today (and there many) could implement serialization of std::tr1::shared_ptr
from its pubilc interface.  Perhaps this suggests that there is something
missing from std::tr1::shared_ptr ?

> Has Peter described the algorithm he used?

we've seen some code snipets on this thread.  I don't know what else may be
available.

Robert Ramey