Re: Serialization (Persistence) library draft submission

From: dylan_nicholson (dylan_nicholson_at_[hidden])
Date: 2002-03-01 00:17:53

--- In boost_at_y..., "rameysb" <ramey_at_r...> wrote:

> Portability and utility demands the numbers be in some sort of 
> canonical form.  I chose text rendition numbers as that canical 
form.
> 
> Why did I choose this one?  Well its the first that came to mind 
but 
> there is a real reason.  This permits us to leverage on the 
> float/text conversion functions already programmed for streams in 
the 
> standard library. What is the point of writing our own?  Are they 
> going to be better than the ones in the standard library? are they 
> going to be faster? use much less space? more robust? I doubt it. 
> 
Actually speed is a really big issue.  We attempted using an LZ 
compressed ASCII format for what are coincidentally enough 
essentially CAD models, and determined that the sheer time it took to 
parse all the ascii representations of floating point numbers was 
unacceptable (the compression time was less significant, but only 
made matters worse.  And the size of the ASCII files made compression 
necessary).  Hence we changed to an in-house binary format (I would 
have preferred XDR, which is what we use elsewhere, but so be it), 
and reduced load and save times by a factor of 10.

Basically what it comes down to it that any serialization that 
doesn't allow you to override how low-level data types are serialized 
is fairly useless.  I suggested elsewhere that the stream class 
itself should be templatized - I have written a binary_stream (it's 
in the boost files section) that is completely compatiable with 
std::iostream and uses XDR, so this is what I would use.

Dylan