In an effort to gauge the use of open source software on the enterprise level, Alfresco turns to its rapidly growing developer community to create a biannual report of user preferences with respect to: operating systems, application servers, databases, browsers, and portals.
This report, forever known as the Open Source Barometer, is generated by collecting opt-in survey data from Alfresco community members.
For the sake of clarification, we are not talking about a couple of hundred people here. Alfresco received responses from 10,000 of their approximately 15,000 community members from countries including: the US, the UK, Germany, Italy, France, Spain, India, Canada, Brazil, the Russian Federation, along with other countries from Europe, Asia, and South America.
The survey contains some surprising results: Operating Systems Alfresco Community users evaluate Alfresco as much on Windows (60.3 percent) as they do on various Linux distributions (34.6 percent). However, when it comes time to deploy Alfresco, 52 percent of users choose to deploy on Linux versus 29 percent on Windows.
Part of this has to do with the fact that Windows remains more prevalent across the desktops of organizations, including the development staff. Yet when it comes to production deployment and maintenance, IT managers are looking for something secure, stable, and openly manageable - three attributes that simply do not describe Windows.
For the Mac fanboys out there, Mac OS X clocks in at 12 percent when it comes to deployment - up from 3.8 percent during the evaluation phase. As for Linux, Alfresco deployments are spread almost equally between Debian, Fedora Core, RHEL, SUSE, Ubuntu, and Other. However, if Fedora Core and Red Hat Enterprise Linux are combined they take up 37 percent of the reported deployments.
An interesting side note contained within the survey results indicates that Red Hat Enterprise Linux adoption took off just after Novell SUSE signed the patent agreements with Microsoft. The lesson to be learned there is never sign a deal with the devil. Application Server From an application server perspective, it is not even close. Alfresco community users prefer Tomcat (72.1 percent) and JBoss (19.8 percent) to BEA Weblogic (2.5 percent) and IBM Websphere (2.3 percent) for evaluation.
For deployment, BEA and Websphere make up some ground combining for 15 percent of the responses. However, the open source competitors remain firmly in the lead with Tomcat and JBoss making up a total of 70 percent of the responses.
This data lends evidence to the viewpoint that there simply is not that much difference between an open source and a commercial application server. When licensing and professional services costs are factored in, it is obvious that a large number of organizations figure that an open source solution can help them achieve their goals faster and cheaper. Database Much like the application server data, MySQL is used in the evaluation phase 61.9 percent of the time with Oracle coming in a distant second at 14.3 percent. Furthermore in the deployment phase, MySQL is the database of choice with 40 percent while Postgres shows up at 28 percent, with Oracle ticking up slightly to 19 percent. Microsoft's SQL Server weighs in with a whopping 4 percent.
It seems like a long time ago when MySQL was not considered an enterprise level database. It is obviously good enough for the enterprises who responded to this survey. Browsers and Portals For accessing and interacting with the Alfresco ECM repository, the surveyed users preferred browsers with Firefox being the most popular - neither of which are surprising. Yet when it comes to accessing Alfresco via a portal application, JBoss Portal leads with 51 percent and the somewhat surprising Liferay Portal comes in second at 32 percent. The portal applications from BEA, IBM, Vignette, and Oracle combine to make up 15 percent of the respondents' choices.
For the needs of organizations evaluating and deploying Alfresco for enterprise content management, it is obvious from this data that open source solutions are more than adequate.
Should IBM, Microsoft, BEA, and Oracle be worried? Until they begin to see legitimate impacts on their bottom lines, probably not.
However, it would certainly be foolish to continue to turn a blind eye to the open source software movement that is showing no signs of slowing down.
To review the results yourself, the initial version of the Open Source Barometer can be found here.
Where do you fall on the open source versus commercial software argument? Is your organization only comfortable with software they have paid for or are they willing to spend money that would have gone to licensing fees on professional services instead?
Let us know what you think in the comments below.