From: nitin motgi (nitin.motgi_at_[hidden])
Date: 2005-12-30 13:45:24
> The serialization concept might be a bit harder to validate than it
> seems initially, since there are various ways how serialization can
> be implemented. The ways that come to my mind (skipping over types
> directly supported by the library):
> - a serialize member function
> - a serialize free function
>- load_construct_data and save_construct_data functions for types
>without a default constructor, or types containing reference members
>- a save_override or load_override function in the archive class
In regards to validating the property/functionality of the serialization
library can we not use a signature for each/a property/functionality.
Because of the complexity and the depth of this library this kind of
mechanism where in library tags a property/functionality and concept
checker uses it to validate would be very useful.
When library provider wants to validate some functionalities of
he would be able to do it in parts. IMHO I feel providing flexibility to
allow users to check only for those parts of concepts they are interested in
be largely useful. E.g. NVP, Primitives ..etc.
This will also allow users to use serialization concept check on
older versions of serializable classes.
-- Nitin Motgi NITIN . DOT. MOTGI .AT. GMAIL .DOT. COM
Boost list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk