Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Updated serialization performance results

1 view
Skip to first unread message

Brian Wood

unread,
Aug 24, 2010, 8:03:39 PM8/24/10
to

I've recently updated this page --
http://webEbenezer.net/comparison.html --
with results based on Boost 1.44. The previous version
of that page used Boost 1.41 and can be found here --
http://webEbenezer.net/comp141.html.
To summarize, the results on Windows were not
noticeably different, but the results on Linux changed in
favor of the C++ Middleware Writer.

Here's a link to one of the Boost based programs in
case there are suggestions to improve it's
performance --
http://webEbenezer.net/comp/bser.cc.


Brian Wood
Ebenezer Enterprises
http://webEbenezer.net
(651) 251-9384

--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]

Mathias Gaunard

unread,
Aug 25, 2010, 1:26:05 PM8/25/10
to
On Aug 25, 1:03 am, Brian Wood <woodbria...@gmail.com> wrote:
> I've recently updated this page --http://webEbenezer.net/comparison.html--
> with results based on Boost 1.44. The previous version
> of that page used Boost 1.41 and can be found here --http://webEbenezer.net/comp141.html.

> To summarize, the results on Windows were not
> noticeably different, but the results on Linux changed in
> favor of the C++ Middleware Writer.
>
> Here's a link to one of the Boost based programs in
> case there are suggestions to improve it's
> performance --http://webEbenezer.net/comp/bser.cc.

Have you tried writing a Boost.Serialization archive using the same
code you use in your own serialization solution?
The difference probably comes from there, not from the Send/Receive vs
template serialize with &.

Brian

unread,
Aug 26, 2010, 11:08:48 PM8/26/10
to
On Aug 25, 12:26 pm, Mathias Gaunard <loufo...@gmail.com> wrote:
>
> Have you tried writing a Boost.Serialization archive using the same
> code you use in your own serialization solution?
> The difference probably comes from there, not from the Send/Receive vs
> template serialize with &.
>

No, I haven't. I agree with your assessment for the
tests mentioned. I'd just point out that I'm using
the two as is without having to make significant changes
to either one.

Brian Wood

0 new messages