Five months after OpenLogic Inc. spearheaded a census of open source software use, the Broomfield, Colo.-based...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
open source aggregator has yet to learn much about what's running in corporate data centers.
The goal of the voluntary ongoing survey was to provide open source software companies with a more accurate gauge of the amount of their downloaded products that are actually in production. In addition, it was supposed to help participating companies identify which open source applications they run. This, in turn, supposedly would speed problem-solving and provide a benchmark of their open source usage versus the portfolios of competing firms.
Unfortunately, however, the survey's relatively low participation rate -- with only 50% more machines scanned over the past five months than in the first 45 days in which the survey was available -- and its preponderance of desktop data have yielded no quantifiable information about what's running on corporate servers. Over the past five months, the number of scanned machines grew from only 1,270 to 2,217, and the number of open source software copies from only 224,370 to 303,426, indicating that the largest influx of respondent data occurred during the survey's initial release.
Kim Weins, OpenLogic's senior vice president of marketing, said, desktop focus aside, that the survey represents the open source market as a whole, of which commercial software is "not the most prevalent."The findings: Under the hood
"Analysts believe the results are representative because RHEL [Red Hat Enterprise Linux] and SUSE [Novell's SUSE Linux Enterprise] run on production machines, which is still a small slice of the [open source] market," Weins said.
The universe of open source software is much larger than the commercial, for-profit share, which gets more visibility, she said. Less than 10% of OpenLogic's stock of 450 open source applications are backed by a commercial vendor, she noted. Besides, commercial open source applications such as SugarCRM are still early in their adoption cycles to show up in inventories and, in any case, may never appear if they are run primarily as a hosted service, she added.
Red Hat, the most popular open source commercial distro, did not appear in the survey results which were overwhelmingly dominated by Ubuntu. Fedora, Red Hat's community-based software, SUSE (probably referring to openSUSE), Ubuntu, Debian and Gentoo ranked the highest in operating systems while Firefox, zlib compression library, the Xerces interface to Apache ranked tops in applications; overall, OpenOffice slid from 10th to 11th place. Popular open source business applications such as SugarCRM customer relationship management and Alfresco content management software were nowhere on the landscape.
Nevertheless, the survey did reveal noteworthy trends. Open source adoption, for example, appears to be somewhat more prevalent in U.S. government and financial circles than in U.S. businesses overall and more widespread in Europe than in North America. In addition, lots of open source applications are running on Windows desktops.
Matt Lawton, director of open source software business strategies at Framingham, Mass.-based IDC, which is a sponsor of the OpenLogic study, said the response rate has not increased as much as organizers had hoped but that over time it should increase as awareness of the project and comfort with the anonymity and security of the inventory process grows.
Weins said the survey is a long-term effort that should grow significantly over the next two years. OpenLogic will continue to work with project sponsors to encourage survey participation by customers and observe how the data changes over time, she said.
For the survey results and the Discovery software inventory engine go to the Open Source Census website website.