Very fast, header only, C++ logging library.[![Build Status](https://travis-ci.org/gabime/spdlog.svg?branch=master)](https://travis-ci.org/gabime/spdlog)
## Install
## Install
Just copy the files to your build tree and use a C++11 compiler
Just copy the files to your build tree and use a C++11 compiler
...
@@ -9,7 +9,7 @@ Just copy the files to your build tree and use a C++11 compiler
...
@@ -9,7 +9,7 @@ Just copy the files to your build tree and use a C++11 compiler
* Linux (gcc 4.8.1+, clang 3.5+)
* Linux (gcc 4.8.1+, clang 3.5+)
* Windows (visual studio 2013+, mingw with g++ 4.9.1+)
* Windows (visual studio 2013+, mingw with g++ 4.9.1+)
* Mac OSX (clang 3.5+)
* Mac OSX (clang 3.5+)
##Features
##Features
* Very fast - performance is the primary goal (see [benchmarks](#benchmarks) below).
* Very fast - performance is the primary goal (see [benchmarks](#benchmarks) below).
* Headers only.
* Headers only.
...
@@ -31,7 +31,7 @@ Just copy the files to your build tree and use a C++11 compiler
...
@@ -31,7 +31,7 @@ Just copy the files to your build tree and use a C++11 compiler
## Benchmarks
## Benchmarks
Below are some [benchmarks](bench) comparing popular log libraries under Ubuntu 64 bit, Intel i7-4770 CPU @ 3.40GHz
Below are some [benchmarks](bench) comparing popular log libraries under Ubuntu 64 bit, Intel i7-4770 CPU @ 3.40GHz
#### Synchronous mode
#### Synchronous mode
Time needed to log 1,000,000 lines in synchronous mode (in seconds, the best of 3 runs):
Time needed to log 1,000,000 lines in synchronous mode (in seconds, the best of 3 runs):
...
@@ -43,7 +43,7 @@ Time needed to log 1,000,000 lines in synchronous mode (in seconds, the best of
...
@@ -43,7 +43,7 @@ Time needed to log 1,000,000 lines in synchronous mode (in seconds, the best of
|100| 15.008 |1.139s |4.512s |0.497s|
|100| 15.008 |1.139s |4.512s |0.497s|
#### Asynchronous mode
#### Asynchronous mode
Time needed to log 1,000,000 lines in asynchronous mode, i.e. the time it takes to put them in the async queue (in seconds, the best of 3 runs):
Time needed to log 1,000,000 lines in asynchronous mode, i.e. the time it takes to put them in the async queue (in seconds, the best of 3 runs):