On Friday, January 25, 2013 11:18:46 AM UTC+8, kira kk wrote:
>
> Thanks *Marc. Before I run some trial tests. May you advise some benefit
> of using protocal buffer instead of a collection of primitive data type.*
>
To name some of the benefits (over using primitive data types):
1. Backward co
Thanks *Marc. Before I run some trial tests. May you advise some benefit of
using protocal buffer instead of a collection of primitive data type.*
*
*
*Kira *
On 24 January 2013 15:40, Marc Gravell wrote:
> Exact performance is very implementation- / application-specific. Yes,
> protobuf will b
Hi,
I am considering to adopt protocol buffer in c++. However, I cannot find
there is any performance about it.
Here is my situation: the applications are distributed and
performance-critical C++ applications. Each application communicates
through message by socket.
So the message is serialize
In 2.3.0, the Python implementation is now 10x-25x faster than it used to
be. However, as the implementation is pure-python, it is still much slower
than C++. Petar (author of much of the Python code) is currently working on
a version of the Python implementation which wraps the C++ library in th
Hi !
I'm interested in using Protocol buffers for serialization of data
passed between Python and C++ processes.
I've seen some posts on performance comparison of different python
serializes, saying that the performance of google protocols
implementation for python is much worse than in C++.
Was
