A Global Open Source Census

There’s an issue that has always been contentious when it comes to Open Source. How do you accurately and definitively measure adoption. Due to the myriad ways Open Source code can be disseminated, the answer has historically been “you can’t”. Sure, you can use download metrics from mirrors and some educated guessing to come up with decent approximations, but those are neither definitively or truly accurate (and probably almost always underestimate real adoption). Enter the Open Source Census:

The Open Source Census is a global, collaborative project to collect and share quantitative data on the use of open source software in enterprise. In short that means we want to count how many enterprise installations there are in the world for each open source software package. Results of the census will be shared on this website.

Companies can use an open source software tool called OSS Discovery to automatically scan a sampling of machines in their organization. These scans will gather completely anonymous data on which open source packages and versions are installed. The scan results can then be submitted to a repository on The Open Source Census website. Scan contributors can review the actual results of the scans before they decide whether to contribute the data. Contributors will be asked to provide some limited demographic data to enhance reporting of the results.

It’s a lofty goal to be sure, and whether it gains any traction remains to be seen. Even if it does take off, it’s aimed only at the Enterprise which will certainly skew results. That being said, it’s definitely better than nothing and the information gleaned should be quite valuable. Additionally, the OSS Discovery tool should be a great way for an Enterprise to track its Open Source usage. Data collection will begin in Q1 2008.

–jeremy