Re: [Boost-users] [serialisation]Portable Binary Archives

From: Scott McMurray (me22.ca+boost_at_[hidden])
Date: 2007-06-27 16:08:21

On 27/06/07, Cliff Green <cliffg_at_[hidden]> wrote:
> > I don't think binary archives are going to be smaller
> >than the equivalent
> > text archive.  native binary archives are built for
> >speed, text for
> > portability, and xml to satisfy those who feel they need
> >it.
>
> I generally agree, but I wouldn't be surprised if binary
> archives are smaller than text archives for everything but
> unusual cases. It may not be a significant percentage
> difference, and of course without some objective test
> cases I'm just guessing.
>
FWIW, I'm currently using binary files for some work because the text
is huge and slow.

For the naiive, punning-to-char* approach, my files go, for example,
go down to 1.2 GiB instead of the 2.1 GiB for text, despite storing
slightly more data per entry (quaternion instead of euler angles, to
be precise).  It also makes my load time drop from a few minutes to a
few seconds for the smallest of the files (150 MiB of text), which
adds up very quickly.

Now this is nowhere as general as a proper portable
Boost.Serialization-style archive, but I consider it an important
usage of a possible binary library, whether that be some sort of
boost.endian (to just make what I'm doing port portable), a binary I/O
library (as another thread is discussing), or a portable binary
archive Boost.Serialization library (which obviously couldn't get as
much space reduction, but the speed boost should still be
considerable).

~ Scott McMurray