The Independent Oracle Users Group (IOUG) has released its latest member survey on the use of open source, with the results indicating not surprisingly that acceptance remains guarded but rising.
"When open source first came out, there were concerns about not having a 'real' company behind it," said IOUG president Ari Kaplan, who added that most have overcome their early scepticism. "What jumped out is that over 60% are relying on open source in the middle tier."
The IOUG survey, conducted by Unisphere Research and sponsored by MySQL, revealed over a third of the 200+ respondents were using an open source database; given that the survey polled Oracle customers, that meant that respondents were deploying MySQL or similar open source alternatives in addition, rather than in place of Oracle.
The survey explored the impact of freebie 'Express' editions that are now being made available by Oracle, IBM, and Microsoft to encourage new customers to kick the wheels. The survey revealed that nearly 80% of 'Express' database users are still using open source counterparts.
Interestingly, the number of organisations running over half of their applications on open source grew from 9% last year to 13% in the current survey. By far, the driver remains cost, with freedom from vendor lock-in being a very distant second.
But adoption of open source remains a mile wide and, at best, a few inches deep. Although over 90% reported using at least some open source software, barely 4% used it for enterprise applications such as SugarCRM or Alfresco portal. Not surprisingly, most open source software was primarily confined to dedicated uses, such as middleware, or for test and development.
And, not surprisingly, most open source database users still view these systems as comparative toys. Only 20% of open source database users stored more than 50 GBytes within them, and only 3% reported storing over a terabyte.
The use of open source, or in this case, Linux, has actually been encouraged by Oracle, which has made Linux is reference development platform. As one survey respondent was quoted, "Oracle 11g beta releases were available for Linux before they were available for other platforms."
The Apache web server proved the single most popular open source middleware component, used by 60% of respondents. It was followed by OSs such as Linux or FreeBSD at 58%; appservers such as JBoss or Tomcat at 45%; databases such as MySQL, PostgreSQL, or EnterpriseDB at 35%; development tools such as Eclipse or NetBeans at 30%; and frameworks such as Spring at 21%. By comparison, only 9% indicated that they did not use open software at this time.
But just over half the group said they would increase use of open source software over the next year, with nearly 70% citing cost savings as the reason for their interest.
The prime obstacles were not that surprising. Just over half the respondents cited poorer support compared to commercial software packages, which comprised an increase over last year when only about a third of respondents voiced these concerns. But there were some interesting patterns here, with larger organisations citing support, while smaller organisations said that security was their prime concern.
It is clear that when it comes to open source, customers are 'buying' it for one plain reason: cost. Consequently, adoption patterns, where open source is deployed for piece of software architecture that are commodity is no longer surprising.
Specifically, they are focusing on software plumbing that is simply supposed to work, and on its own adds little intrinsic value.
One of the obvious places is the operating system, for which customers view as being a means to an end. Conversely, the fact that customers still pay lots of money for Windows clients (the one portion of the tier where open source penetration has remained limited) is not because they are paying for the OS itself. Instead, they are paying for access to Microsoft Office, which has become a de facto standard in the developed world.