In connection with the open source release of OSS Discovery my friends at OpenLogic have also announced an ambitious new project called 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 enterprises throughout the world. Of course, OpenLogic acknowledges that it will not be possible to count every single instance of open source software in every enterprise in the entire world. They are, however, shooting for a large enough sample to make it representative.
Setting out to count the number of enterprise open source installations in the world is no small task, but OpenLogic has provided the infrastructure to automate much of the process:
— Companies download the open source version of OSS Discovery;
— Use it automatically to scan the machines in their organization (or a sampling of those machines) for open source software;
— Submit those scan results (anonymously) to an online repository on The Open Source Census web site; and
— Starting in Q1 2008, view the results of the Census online.
I have been involved in a number of corporate open source audits — with the goal being to inventory the open source software in use in by a company or within a given unit of the company. If there is one thing all of these audits have in common it is that they reveal there to be far more actual open source usage than was originally thought (note that if there is one other thing these audits have in common it is that they end up involving far more time and resources than was originally thought as well — but with automated tools like OSS Discovery now available free of charge I am hoping that will not be as much of an issue with future audits ;-). The nature of how open source software is acquired (i.e., through downloads instead of standard procurement channels) means that most companies just can’t produce an accurate inventory of their open source software usage. As a result, whenever I read a research report by Forrester, Gartner, IDG or some other group detailing the extent of open source usage in a given industry (or a report by a company indicating that they use “no open source”) I often wonder about the accuracy of the statistics underlying the reports (particularly when they rely on self-reporting). While I am sure the folks at Forrester, Gartner and IDG have methods to address sources of inaccuracy in their data, the combination of OSS Discovery with The Open Source Census appears to present a very simple and straight-forward means to generate accurate statistics on open source usage on a wide basis. I will be interested to see the result.