In 1999, Tim O’Reilly, founder of a popular open source-oriented publishing house, gave a keynote speech to an audience of Fortune 500 executives called “ten myths about open source software”. As those myths are still perceived today, as shown by recent reports [CIO 07, ED 05, Forr 07], and are still perceived as a barrier towards FLOSS adoption, we will try to provide here a SME-oriented and pragmatic answer to all of them.
* 1 Myth #1: It’s a Linux-vs-Windows thing.
* 2 Myth #2: FLOSS is not reliable or supported.
* 3 Myth #3: Big companies don’t use FLOSS.
* 4 Myth #4: FLOSS is hostile to intellectual property.
* 5 Myth #5: FLOSS is all about licenses.
* 6 Myth #6: If I give away my software to the FLOSS community, thousands of developers will suddenly start working for me for nothing.
* 7 Myth #7: FLOSS only matters to programmers, since most users never look under the hood anyway.
* 8 Myth #8: There is no money to be made on FLOSS.
* 9 Myth #9: The FLOSS movement isn’t sustainable, since people will stop developing free software once they see others making lots of money from their efforts.
* 10 Myth #10: FLOSS is playing catch-up to Microsoft and the commercial world.
Myth #1: It’s a Linux-vs-Windows thing.
Most recent debates about FLOSS were focused on an all-or-nothing perception. For example, when introducing FLOSS in a company, a full software migration is often considered as necessary. This, and the fact that there is limited knowledge of FLOSS projects except for a few very widely known ones (like Linux, Apache, OpenOffice.org), created the perception that most FLOSS is designed and targeted as a direct competitor of Microsoft products. The reality is that there is an enormous number of active projects in practically any IT field, including business-specific (such as ERP systems), being most of them cross-platform, capable of running Microsoft Windows, Apple’s OSX (which is itself based on more than 300 open source projects) or Linux. As can be found in Appendix 1, there are more than 18,000 FLOSS projects that are stable and mature for adoption by SMEs.
Myth #2: FLOSS is not reliable or supported.
This myth is based on a common perception that FLOSS is exclusively developed by volunteers in a non-coordinated or unstructured way. There are many errors in this view:
* the volunteer perception: while volunteer contributions are a significant part (and sometimes the majority) of large scale projects, around 50% of developers have received a financial compensation for working on FLOSS projects, either directly paid to improve the projects or paid to support them. This has been shown in recent studies [Gosh 05, Gosh 06] and can be inferred directly by the fact that in the software industry at large, 68% of software products include directly FLOSS-derived code.
* paid programmers are better: even for the percentage of contributions that are coming from volunteers, it is commonly perceived that those should be of inferior quality, as there is no financial incentive to produce quality software. This ignores the fact that intrinsic incentives are in many cases more effective than monetary compensation [Gosh 06], and the fact that sometimes users are interested in improving the software that they are using [VH 03]. This second effect, called user-driven innovation, has been shown in past research to be a significant force. For example, around 25% of innovations in fields like software security, printed circuit boards CAD systems and library software were designed and introduced by advanced users. The same effect provides a fundamental design feedback, as large project collects both good and bad experiences in using the software (for example, the Ubuntu Linux “Testimonial and Experiences page” that allows for a form of user-driven “steering” of the project and the identification of trouble points.
* there is no support: most large scale project are related to companies that provide paid-for support, in a way similar to that of proprietary software companies. The availability of source code and the modification rights gives also the additional advantage that support can be obtained even for projects that are no longer active, in stark difference with proprietary software where no code escrow clause was included in the acquisition contract.
* FLOSS is inherently unreliable: many believe that FLOSS, as developed in an open and unstructured way, is inherently of lesser quality when compared to proprietary software. The reality is that most FLOSS projects are organized in a semi-strict structure, and only very modular projects are inherently “bazaar-style”, allowing for large scale internal decoupling. In any case, the impact of FLOSS-style development has been assessed in several research papers, and for example in [Suc 04] we found: “The hypothesis that open-source software fosters more creativity is supported by our analysis. The growing rate, or the number of functions added, was greater in the open-source projects than in the closed-source projects. This indicates that the open-source approach may be able to provide more features over time than by using the closed-source approach. Practitioners interested in capturing market share by providing additional features should look to the open-source methodology as a method to achieve this. In terms of defects, our analysis finds that the changing rate or the functions modified as a percentage of the total functions is higher in open-source projects than in closed- source projects. This supports the hypothesis that defects may be found and fixed more quickly in open-source projects than in closed-source projects and may be an added benefit for using the open-source development model.” This is consistent with results from vendors of software defect identification tools, such as Reasoning, that found that while the bug density ratio in initial project releases is on par with proprietary developments, it improves rapidly and for some projects defect densities are significantly lower than that of the average proprietary code [Reas 06a, Reas 06b]. This was confirmed by other studies like the reports from Coverity.
The fact that FLOSS is overall reliable can be also inferred by surveys like [CIO 07], where 79% of respondents answered positively to the question “My company’s experience with open source products other than Linux has been so good we plan to expand their use”.
In this sense, it should be no surprise that several FLOSS projects have received safety certifications, or have been used in medical devices, control systems and avionics. For example, the VISTA system is a large scale electronic health care system, developed by the US Department of Defense for its own veteran hospitals, and now used in more than 1000 hospitals and clinics in the US alone, along with many other installations across many countries. Other examples include the use of Linux in Siemens Magnetic Resonance Imaging systems used in diagnostics, the use of the open source ADACORE environment in in-flight avionics, the FIPS-140 certification of two of the most important encryption toolkits (OpenSSL and NSS), and many more.
If we take as an example the IEC 61508 safety integrity levels [Daf 06-2]:
the UK Health and Safety Executive, in a study from 2002 [HSE 02] found that Linux was robust enough, and that it could be certified up to SIL3 with limited effort. This would make it amenable for use in air traffic control displays, railways control systems and process plant control.
Myth #3: Big companies don’t use FLOSS.
The easiest myth to dispel: apart from the large IT companies that are actively promoting open source software like IBM, HP, Sun, Oracle, and others, about 86% of Fortune 1000 companies are deploying or testing FLOSS, and a similar percentage is found in Europe [Aug 04]. Of those, 35% or more are deploying more than 20% of their systems as FLOSS, and 11% of companies report more than 20% of their applications as FLOSS. While usage in server-centric and IT infrastructure is more common, around 26% of large companies are mentioning the use of Linux on the desktop, and a much larger percentage are reporting the use of some other FLOSS packages, such as OpenOffice.org and Firefox on Windows desktops. A curious fact also evident from other surveys is that many companies and public administrations are not aware of their internal use of FLOSS, sometimes for simple ignorance of the licensing terms and sometimes because the product is offered or embedded in what seems like a traditional proprietary offering (for example, many security and networking products, or enterprise products like VMware ESX server, use FLOSS internally).
Myth #4: FLOSS is hostile to intellectual property.
There are several aspects that are referenced to this myth:
* The GPL license is “viral”: the most widely used license does have a specific clause which mandates that when a software product that is derived from GPL software code is redistributed, the entire product must comply with the conditions of the GPL. This has prompted some companies to claim that “the viral aspect of the G.P.L. poses a threat to the intellectual property of any organization making use of it”. The reality is that for most scenarios, this clause simply provides a way to prevent appropriation of code without giving back contributions or credit, which is one of the reasons why many developers prefer the GPL to other licenses. Simple use of FLOSS in itself does not require any change to the license of internally developed software, and most companies routinely run proprietary software on top of GPL-licensed code like the Linux kernel.
* The free software community steals the intellectual property of other companies: this is mainly the byproduct of a legal case, in which the SCO company claimed in 2003 that IBM improperly included copyrighted material in the Linux kernel. In the original claim, it was mentioned that IBM “put SCO’s confidential and proprietary information into Linux, the free operating system” and that several millions of lines of code of the Linux kernel were stolen from SCO’s Unix source code. Now, four years later, the judges have thrown out most of the claims, leaving less than 300 lines of code (mostly standard interface code) still under evaluation, out of more than 6 million lines of code of a modern Linux kernel. Recently Microsoft issued similar statements, with Microsoft’s CEO Steve Ballmer claiming that “that product (Linux) uses our patented intellectual property”, and later numbering how many patents Linux and other FLOSS products were infringing Microsoft’s intellectual property. The reality is that structured FLOSS projects do have strict policies for accepting patches and external contributions. As an example, the Eclipse project has a strict due diligence process, that covers external contributions, code rights assignments, code review and license compatibility. The Eclipse Foundation also uses automated tools to check for code copying, keyword scanning for words with legal significance and a controlled release review prior to updating the code [Cam 06]. Similar processes are in place in other FLOSS projects [Rig 06].
Myth #5: FLOSS is all about licenses.
While in a strict sense a FLOSS project is defined by its license, most aspects of open source are really related to the openness and collaborative aspects of the development, as described in chapter 1.
Myth #6: If I give away my software to the FLOSS community, thousands of developers will suddenly start working for me for nothing.
There is no guarantee that simply “dumping” source code on the web will make a FLOSS project appear, and there have been several examples of such behavior to be even negative (because the community may see this as “garbage dumping”). The reality is that for some collaboration to happen, there must be first of all a good communication, interaction strategy and effort in place. In addition, investing in community creation and dissemination efforts increase the probability of a bidirectional effort sharing. It is important to mention that surveys like OSSWatch or [CIO 07] found a significant proportion of companies and public administrations (between 14% and 25%) contribute back patches or participate actively in FLOSS communities.
Myth #7: FLOSS only matters to programmers, since most users never look under the hood anyway.
The fact that most users are not interested in the source code does not imply that having the source code available in itself is useless. Several positive aspects can be identified:
* The availability of the code allows end users to eventually pay someone for modifications or continuing maintenance even when the original FLOSS project disappears or becomes inactive.
* “Under the hood” there is not only code, but many non-code artifacts that are vital to a project, like translations, documentation, examples, etc. Many users can contribute in such aspects even if they are not programmers.
* For some projects, having the code available allows for a significant cost reduction or increases dramatically the flexibility of the offered solution. For example, in a project called MuleSource (a sophisticated middleware system) it was found that 64% of users perform at least one source code modification.
Myth #8: There is no money to be made on FLOSS.
Even many researchers have proclaimed in a way or the other that the freely available nature of the code precludes any potential commercial exploitation. For example, in [Hahn 02]: “The GPL effectively prevents profit-making firms from using any of the code since all derivative products must also be distributed under the GPL license”. This of course collides with the economic results obtained by companies like HP (that in 2003 reported more than 2.5B$ in Linux-related revenues), or the 400M$ revenues reported in 2006 by RedHat. In [Gosh 06] it is evaluated that:
* Defined broadly, FLOSS-related services could reach a 32% share of all IT services by 2010, and the FLOSS-related share of the economy could reach 4% of European GDP by 2010.
* FLOSS directly supports the 29% share of software that is developed in-house in the EU (43% in the U.S.).
* FLOSS potentially saves industry over 36% in software R&D investment that can result in increased profits or be more usefully spent in further innovation.
* The notional value of Europe’s investment in FLOSS software today is Euro 22 billion (36 billion in the US) representing 20.5% of total software investment (20% in the US).
Similar measures are predicted by independent consulting groups like Gartner: in [Gar 06] it is predicted that two years from now, around 25% of the total software market will be FLOSS-based (either through external providers, or by internal developments).
Another relevant aspect is that since most companies adopting FLOSS report significant cost savings, these can be directly transferred to external professional services or incorporated as additional profit margin. For example, in [Inf 07]:
In a survey of 800 IT managers, InfoWorld found that of all the FLOSS adopters, those collecting the most significant benefits are those that deploy more open source products, with 24% of the “large users” (more than 100 products) reporting savings of more than 60%. It is also interesting to notice that only a very small percentage (<9%) reports that there are no savings or that costs have increased compared to proprietary software.
Myth #9: The FLOSS movement isn’t sustainable, since people will stop developing free software once they see others making lots of money from their efforts.
This is connected to the view of myth #2, the idea that FLOSS is developed by volunteers, and that companies can only profit in a parasitic way from the code that is developed for free. As discussed in that part, the reality is that in most projects companies and volunteers participate in a collaborative and non-competitive way; also, the most widely used license (the GPL) forced companies to reciprocate their efforts by making dissemination of the source code mandatory whenever there is dissemination of code derived from GPL projects.
Myth #10: FLOSS is playing catch-up to Microsoft and the commercial world.
The concept of software innovation is really rooted in two different aspects: technical innovation and field innovation. While technical innovation is mostly invisible to the user, “field innovation” (for example a new kind of application) is highly visible. Maybe because of this it is widespread the perception that most FLOSS software is sort of a copy of some other (desktop) oriented proprietary application.
The reality, on the contrary, is that most proprietary software is non-innovative in this aspect. While very few examples of new concepts (like Dan Bricklin’s spreadsheet idea) can be found, most applications are matched to the tasks that people performs daily, and as such there is a strong disincentive to innovate. There are very few studies comparing FLOSS with proprietary software in a replicable and objective way, and one of those is [Kli 05]:
The end result is that from a field innovation point of view, around 12% of the projects in the sample are considered innovative, a percentage that is comparable to that of the proprietary software market. As for the technical innovativeness, the already cited [Suc 04] found that “The hypothesis that open-source software fosters more creativity is supported by our analysis. … This indicates that the open-source approach may be able to provide more features over time than by using the closed-source approach.”
Other research, like [ARC 07], found “the enterprises using office packages alternative to Microsoft are more innovative (average index 0.2) than those using Microsoft only (average index 0.15). The correlation between innovation and the use of free/open-source software at the corporate level is confirmed also in international comparative studies conducted by researchers from Harvard University”.
1. ↑ http://ubuntuforums.org/forumdisplay.php?f=103
2. ↑ “At a defect density of 0.09 defects per KLOC, the version of MySQL we inspected has a defect density that is about six times lower than the average of comparable proprietary projects.”
3. ↑ As mentioned by Craig Mundie, Microsoft’s vice president, in a talk at New York University’s Stern school of Business in 2001. Other representatives of Microsoft like Bill Gates said that “[the GPL] it makes it impossible for a commercial company to use any of that work or build on any of that work”, and Steve Ballmer “Linux is a cancer that attaches itself in an intellectual property sense to everything it touches … if you use any open-source software, you have to make the rest of your software open source” (interview at Chicago Sun-Times, 2001).
4. ↑ The transcript of the initial complaint and a full list of case documents (along with significant analysis) can be found in the GrokLaw site, at http://www.groklaw.net
5. ↑ http://blogs.zdnet.com/hardware/?p=154