Testresults for SCons
Welcome to the SCons testresults. This repository is a collection of several run time analysis and profilings of real-life projects and artificial benchmarks using SCons (http://www.scons.org). The scripts for automatically timing and profiling single builds and update runs, rely on the accompanying scons_testsuite repository which is available at http://bitbucket.org/dirkbaechle/scons_testsuite. Its purpose is to be able to compare different SCons revisions regarding time and memory consumption.
Disclaimer: The testsuite mentioned above was developed during my analysis of SCons' runtime behaviour for the Wiki article "Why SCons is not slow" ( http://www.scons.org/wiki/WhySconsIsNotSlow ). It's currently not in a state of "running everywhere", but especially crafted for my very own Ubuntu Linux machine. So, if you try to start the examples on your own, be prepared to run into some pitfalls. You might have to adapt the controlling scripts, or even need to patch the software packages themselves...including the installation of special packages as prerequisites.
- Based on a test benchmark by Eric Melski, the somewhat ancient SCons v1.2.0 is run and compared against the classical make.
- Here a current SCons v2.3.0 is used, and not only run on a single core/CPU but also in parallel. Additionally, the underlying wonderbuild and Quest for performance benchmarks get patched, such that the CPP compiler gets some real work to do.
- Following the comparison of make vs SCons above, this folder documents an analysis with strace. Three builders and extensions, fastcpp, batchbuild and stubprocess, got developed for speeding up large clean and update builds.
- Results for several runs of the full scons_testsuite against special SCons experimental development branches.
- Detailed tracings of single compile commands for CPython and PyPy.
In each of the above listed result folders, you'll find a reST file named something like README.rst or overview.rst. It explains the structure and contents of each comparison in greater detail.