Re: [boost] Re:Serialization Library Draft # 10

From: Jeff Garland (jeff_at_[hidden])
Date: 2003-09-14 11:13:41

On Sun, 14 Sep 2003 08:13:05 -0700, Robert Ramey wrote
> Jeff Garland wrote:
> 
> >Of course there are many great new features that I had to try out and that ran
> >me into a couple issues with the xml output.  The first is that I can't seem
> >to output a std::string name value pair:
> [snip]
> 
> As described in the "Release Notes" of the documentation, XML 
> archives don't work with gcc 3.2 due to a bug in that version of 
> gcc.  I'm now working on draft # 12 and its possible that this 
> problem may be worked around.  It doesn't appear in gcc 3.3

Ah, thanks I didn't catch that....

> >  I notice that trying this with strings results in a compilation error.  I
> >  assume the intent is to make all types fail to compile with archives that
take
> >  name value pairs?
> 
> You assumption is correct. It is my intention that the xml archive 
> trap at compile time any attempt to serialize anything not  wrapped 
> as a name-value pair.  I'll look into this.

Great.

BTW, I look forward to a description of the requirements for the creation of
new archives.  I believe the new design very nicely supports user written
archives and in my experience this becomes a very important capability (your
requirement #8).  As an example, on a recent project we had to write an
'archive' (we split archives and call them 'Reader' and 'Writer' classes)
because we needed to support certain binary byte orderings for sending /
receiving messages.  We use this type of archive on the boundary of the system
and then used o/istream-based archives for communications within the system. 
All the while, the message objects are oblivious to these different formats.

Jeff