THE QUALITY of open source code has overtaken that of proprietary code for the first time, according to a survey.
In its annual survey published on Tuesday, development testing service Coverity revealed that of the 750 million lines of code it scanned during 2013, the errors in proprietary code exceeded those of open source code for the first time, on all code base sizes.
The open source defect density measure was .59 defects per 1,000 lines of C/C++ code, compared with .72 for proprietary code. Both source code populations were anonymously submitted to assist the project.
Other findings include the fact that Java developers were less likely to fix faults with their code, with 13 percent of identified errors, compared with C/C++ programmers, and that the Apache Hibase, the database module of Apache Hadoop, had the highest number of fixes with 66 percent compared to the average of 13 percent.
Zack Samocha, senior director of products at Coverity, told The INQUIRER, "There are areas in which open source is starting to control the market. It's very hard for a company to come and say that they want to create an operating system now, from scratch, and as far as we are concerned, Linux is leading the charge."
The Coverity Scan service was initially a product of the US Department of Homeland Security before it was taken over by Coverity. Earlier this year, the company was acquired by electronic design automation specialist firm Synopsys.
Although primarily aimed at developers wishing to debug their code, for the first time users can now become observers of the project if they simply wish to study the error finding process of code, to learn from it for their own projects.
Microsoft has recently become the latest organisation to make a partial move to open source, after it announced that it would release source code from its .NET libraries to the community for the first time. µ
Sign up for INQbot – a weekly roundup of the best from the INQ