WALTHAM, USA: Beyond user surveys, there is little actual data on the dynamic open source software industry. As a result, some inaccuracies have prevailed, including the belief that open source developers have created only a few billion lines of code.

Today, Black Duck Software released research based on the Black Duck KnowledgeBase that dispels this and other common “myths”, five in all, about open source in application development. These findings were derived from actual analysis of open source software rather than just user perception.

Black Duck Software actively spiders the Internet collecting downloadable code into a giant repository which is known to Black Duck customers as the KnowledgeBase. This core information repository contains more than 170,000 open source projects from nearly 4,000 unique web sites. It is the largest collection of open source software in the industry. Black Duck mines this valuable resource for information about the development of the open source industry.

Black Duck’s recent findings dispel the following misconceptions about open source software (OSS) in application development:

Myth: Open Source is just source code

Source code is actually only 15 percent of what is released by open source developers. There are four times as many binary files as source files in open source releases. In addition to binaries, open source projects are packaged with scripts, markup language files, graphics images, documentation and many other artifacts.

Myth: Open Source adoption is mostly application infrastructure

There is a tendency to focus on adoption of monolithic applications in IT infrastructure, for example, Linux or MySQL. The Black Duck KnowledgeBase shows that the open source world is dominated by components, not fully formed applications, and these are being reused from project to project in hundreds and even thousands of instances. One example, Apache Log4j is reused by over 5500 projects. Java developers, in particular, have taken tremendous advantage of code reuse. There are 14 times more files distributed ending in the .class file suffix (binaries) than .java (source files). A major reason is that Java components are built once and reused and redistributed by many other projects in binary form.

Myth: There are a few billion lines of code out there

This figure is an order of magnitude too low. There are tens of billions of lines of open source code available on the Internet. In addition, twenty-three percent of all downloadable code was released or renewed in 2008. Over 90 percent of open source code is written in the major languages: C, C++, Java, Javascript and C#, however, dozens of languages are used.

Myth: Real programmers do NOT comment

Open source developers create about one comment line for every four lines of source code. The most commented programming language is Java with more than one comment line for every two lines of code. The least commented language is Boo; a python-inspired programming language that operates within the .net framework.

Myth: GPL Version 3 is being ignored

First released in June 2007, GPL version 3 has grown from zero to over 6,300 projects. In terms of project adoption, it has surpassed the CPL, Mozilla, MIT and Apache licenses. GPLv3 is now the fifth most chosen license in the open source community and if the current trend continues, it will surpass BSD for the number four spot in a year or two. About 70 percent of all open source projects use a variant of the GPL license.