Oracle is very odd company, they own Open Source projects like Java, MySQL and OpenOffice but instead to develop them they do anything to shut them down.
First they try to cash Java by suing Google for using Java in Android and failed miserable, stopped the development of MySQL and OpenOffice, so MySQL was forked to MariaDB and Open Office was forked to LibreOffice and Apache OpenOffice.
Now in this white paper they spread FUD for Open Source trying to convince US government to use their expensive products.
In this document you can read such brilliant statements like:
Open source is not free, nor is it easy to understand the strict legal terms and conditions associated with its use.
Because labor is the largest portion of the costs of an IT project and COTS software is such a small component of the overall lifecycle cost of a program, government officials should carefully assess claims by IT services organizations that promote open source adoption over commercial alternatives.
That is correct, but does the commercial software require less labor to implement project? No! Actually one of the comments below this article in Reddit clears the picture what in practice happens when you use Closed Source software in real project:
Let me preface this by stating that I have nothing inherently against commercial or closed software. It's a different business model. I prefer Open Source, for a variety of reasons, but I don't consider commercial software as much of a sin as people, such as RMS, do. I also don't traffic in the broad statements of code, documentation, or security quality comparisons between Open and Closed source. "Both" sides have examples of warts. That said, I intensely dislike Oracle (although I do like VirtualBox, so there's that). Setting aside their continuing skull-fsck of Sun's corpse and sins against Java, my professional interactions with them have not been positive. This is purely anecdotal. So take it FWIW. I used to work for a DoD contractor. On one project, the requirements all-but-mandated the use of Oracle. Technically, we could've used Open Source solutions, but the government would've immediately red-flagged it for extra scrutiny. Not unreasonable for the type of project we were working on. Our project manager's assessment was that, given that we received 1/6th of the original budget and 1/3 of the original time to develop the system — from architectural specification to prototype — avoiding the extra time for evaluation was prudent. So we had to use Oracle. The big tech we used were Oracle WebLogic, Oracle Service Bus, Oracle Database, and, of course, Java. There were other subsystems we used, but those were noise in the system. We determined that we needed some expertise to help us with the system, so we started asking for Oracle consultants. It took, and I am not exaggerating, 4 months to get Oracle consultants to meet with us. And when they were sent over, they were more salesmen than experts in the technology stacks. They basically tried to sell us more Oracle tech and failed to answer our detailed technical questions. My colleagues and I derived 0-benefit from that meeting. So, we set about learning the tech ourselves — we had deadlines to meet. We'd actually been doing it for those 4 months before the Oracle meeting, but it would have been much faster had we been provided with someone to initially hold our hands. It would have saved me hours of time tracing through WebLogic internals to discover bugs. I can't recall how many conflicting pieces of documentation I came across trying to get Oracle's "Integrated Stack" to... uh... integrate. It was also fun when my colleague compiled his classes with Java 7 (WebLogic was only running on Java 6 at the time). That was a day well spent — not Oracle's fault, but better error messages would've helped. In the meantime, we continued to try to get some expert Oracle support. The government finally provided us a guy from a consulting company that came highly recommended for Oracle work. My colleague and I, the two leads on the project and customer facing employees, corresponded with him remotely at first. We communicated what we needed help with, and planned to meet up on our next trip out to our sponsor. In the meantime, he produced weekly reports to the government sponsor indicating that he was progressing on the stuff we needed. When we arrived, he asked us to clarify a few things for him. So we sketched out the ideas for him, he drew some pretty diagrams — anyone who's worked with OSB would probably recognize the big horizontal cylinder in the middle that says "OSB (magic happens here)" — and told us he'd come back the next day with some examples. The next day, he was hours late. When he arrived, he didn't have the examples he promised. To cut out a bunch of hemming and hawing, and to paraphrase him, "He had to learn the technology first." This was a guy billed to the government as an expert consultant in OSB. He should've been able to whip something out in his sleep. At this point, we'd had enough with trying to get support. We'd already gotten major components of the system working through painful efforts wrestling Oracle technology stacks into submission. The only thing we needed we OSB to communicate between a couple of components. My colleague, who was an Aviation PhD, but not a computer scientist, software engineer, or programmer, was so pissed off that he sat down for 8 hours straight and taught himself how to get OSB doing what we needed it to do. This was in a lab with no outside internet connection, so this was no small feat. This Oracle consultant continued to produce reports and continued to get paid, because "he was contracted to do it." We largely ignored him for the remainder of the project. We wound up delivering the project — minus some features — on time and on budget. At least, according to the government. I now work in the private sector. I am the tech lead for my project at work, and I have integrated many of the lessons learned from those days. But, when I think back to that project, I am convinced that we could have used Open Source software to achieve the same — or better — end result with a fraction of the ass-pain, in less time, and at a lower cost. Oracle is right, Engineering time is the single most costly part of a project. Licensing costs, even at Oracle scale, are a pittance compared to the cost of a Engineering time. So, from a direct cost perspective, the cost savings from open source vs commercial source are simply noise. But the engineering time needed to work with commercial technologies — specifically Oracle — probably far outstrips that required for Open Source solutions. For every single piece of Oracle tech we used, there was an Open Source alternative. They had shortcomings, sure, but those tend to be acknowledged in Open Source projects. Rather than the "Rah-rah, we are the solution to everything" style of Oracle. As I said, this is anecdotal. You can disagree with it if you want. I'm sure there are nits to pick. TL;DR — Used Oracle tech working for DoD contractor. Had bad experience. info source: http://www.dcdata.co.za/public/oracle-attacks-open-source/