[boost] Re: serializatation const change

From: Robert Ramey (ramey_at_[hidden])
Date: 2005-05-28 17:51:47

Peter Dimov wrote:
> JOAQUIN LOPEZ MU?Z wrote:
>
>> For the record, not that I'm in love with the stricter
>> const rules either, but I must confess that, when hit
>> by them the first time, they actually uncovered a (potential)
>> problem in my test code. So...
>>
>> I've been thinking about this issue, and my impression is
>> that regular code won't be hit as much as one might think.
>
> All will be fine in simple cases, where all you do is recursively
> serialize members. To give a data point, in one of my recent projects
> one out of the nine classes does not serialize its members directly.
> This is the
> unversioned case. Once you add versioning, you almost always have to
> serialize a temporary object to take care of the previous versions,

I'm not sure I'd agree with that.  But

> but I'm not sure whether this is an issue if one only supports
> versioned loading (and not saving) with Boost.Serialization.
>

I'm not even sure what versioned saving would mean.  If it means something
and its supported by the library, its would be totally unintentional on my
part.

This issue doesn't arise during load - unless one is loading into a const
member.  This case would trap anyway and require a cast - as I believe it
should.

Robert Ramey