Open Options | Features & quality

While the open source movement is partly a response to poor quality proprietary software, it's unwise to make generalizations about proprietary or open source software. Cost is often a critical criterion for schools, so quality really means value. The best software has all the essential features a school needs and the lowest TCO. Open source software is considered the best choice for many "backend" solutions, while many of the best "frontend" solutions are still proprietary. The open source community creates excellent tools and components, while proprietary interfaces and environments are more user friendly. Neither the proprietary nor the open source model produces many simple, turnkey solutions.

Specific programs should be compared by features, reliability, security, power, networking, and ease of customizing the software for specific needs. The use of open formats is especially important. Schools should also consider how software can support general learning as well as technology curricula.

Features & interfaces
Reliability
Security
Power: Speed & scale
Networking
Customizing
Open formats
Software for schools

Features & interfaces

Proprietary software has more features.

Proprietary software is more user friendly.

On the frontend, proprietary software offers more necessary and desirable features. For example, it may be easier to use peripherals like digital cameras with proprietary software. Solutions like Microsoft Office have matured through years of added features, interface improvements, and usability studies. Some open source projects are overtly cloning proprietary products (e.g. OpenOffice.org is clearly influenced by Microsoft Office). A cloned program may be just as user friendly as the original (such as it is). Of course, while most programs are flush with features, most users (including schools) only need a handful of features.

Schools may not find mature open source solutions for certain needs. The open source community (including universities) is targeting these needs. The results remain to be seen. On the backend, both open source and proprietary solutions often have rich features but include a significant learning curve. Fortunately, most backend users are more comfortable with a command line or other "user hostile" interfaces.

For more about user friendliness, see: "Users & migration"

[back to top]

Reliability

Open source is more reliable.

Open source software may be more reliable than proprietary. It may not make as many errors or crash as often. (e.g. Linux is famous for not crashing.) Since any programmer can find and fix bugs, software may be repaired and improved more quickly. The initial program may not be more reliable than a proprietary alternative, but it may mature faster as hundreds or thousands of programmers correct mistakes and add features. Eric Raymond calls this Linus's Law: "Given enough eyeballs, all bugs are shallow." (Raymond, Cathedral) Some people think of this as permanent beta testing. The open source community can endlessly troubleshoot and improve software as needed or desired.

However, this advantage depends on the participation of enough competent programmers. Just like proprietary software, the reliability of an open source program depends on clear feedback after rigorous use in a variety of environments. Without enduring, sufficient, talented interest, an open source project fails, and many do. In contrast, proprietary software companies may create and support necessary programs that no one would enjoy working on. Some companies are starting to blend the best of both models, by employing a core group of programmers while attracting volunteers from the open source community.

[back to top]

Security

Open source is more secure.

The transparency and reliability of open source may mean better security. Linus's Law describes parallel debugging: multiple programmers independently finding and fixing the same program to discover the best solution. Security threats like viruses and worms exploit bugs to damage computers. Eric Raymond explains:

One advantage of parallel debugging is that bugs and their fixes are found / propagated much faster than in traditional processes. For example, when the TearDrop IP attack was first posted to the web, less than 24 hours passed before the Linux community had a working fix available for download. (Raymond, Halloween)

With both open source and proprietary software, effective security depends on thoughtful deployment, regular monitoring, and timely upgrades or other modifications. Some proprietary solutions offer potentially robust security, but Linux leads the industry in defensive design. (This is a critical reason why business and government are interested in open source.) For example, Linux isolates most users from the code viruses and worms need. These threats need unchecked access to execute, replicate, and deliver a payload (e.g. delete all files). In a properly configured Linux environment, an email attachment in a user's inbox can't spread destruction to the whole system. For schools, security also includes Web filtering. Open source offers popular and effective solutions (e.g. SquidGuard).

Transparency is sometimes misrepresented as a critical security threat. Transparency may make it easier for a criminal gain access to a program. However, the open source model leverages a community of programmers to maintain and improve security. This collective benefit seems to outweigh the danger of transparency. Also, the security of a system depends much more on careful deployment and maintenance. For example, the source code for Linux is publically available. But a shrewd system administrator will deploy a Linux server with passwords, firewalls, additional software, and other defenses. Thus, transparency in the "virgin" source code is usually a trivial concern.

The open source movement is partly a response to safety and security concerns with proprietary software. Open source offers powerful potential security advantages by preventing spyware and promoting encryption. However, some of these issues are fairly remote from schools.

[back to top]

Power: Speed & scale

Open source is more powerful.

Some open source programs are generally faster and scale larger than proprietary alternatives. Where power means speed, projects like Avalon and Stone SouperComputer have used Linux to build supercomputers. (Cook, 2000) Where power means multitasking, the K-12 Linux Terminal Server Project (K12LTSP) creates whole computer labs using a single Linux server and a network of thin clients.

Linux is famous for its ability to scale: to run large networks with many users. For example, the Beaverton School District in Portland, Oregon serves about 33,000 students. The district wanted to create a single virtual hard drive: a file server with space for every teacher and student. They tried several solutions, but they all failed; Linux was the only software that could scale successfully. They are "really pleased with the result." (Kelly Kuntz, personal communication, January 30, 2003)

The second-most popular open source operating system is the Berkeley Software Distribution (BSD). BSD is equally impressive in speed and scale. "BSD servers power everything from Yahoo to --until very recently-- a website set up by Microsoft and Unisys to encourage businesses to stop relying on UNIX." (Caulfield, 2002)

However, proprietary solutions will continue to offer comparable speed and scalability. For example, Sun offers both open source and proprietary solutions (Linux and Solaris) and is continuing to develop proprietary technologies for the future. While open source deployments like the Stone SouperComputer or Yahoo may demonstrate relatively inexpensive power computing, proprietary software will continue to offer adequate power for most education needs. For example, there are proprietary solutions for thin client networks.

[back to top]

Networking

Open source is more network friendly.

Much of the popularity of open source comes from its performance on the Internet and lesser networks. Open source software is often very networkable. For example, more than half the World Wide Web runs on Apache, an open source solution. Apple builds its OS X on BSD, an open source operating system. The Internet is a critical reason: Apple recognizes they can't privately innovate Internet functionality as well or as fast as the open source community. Most open source networking solutions are compatible with proprietary software. For example, a lab of Linux computers can be seamlessly nested inside a larger proprietary network using an open source intermediary (e.g. Samba).

[back to top]

Customizing

Open source can be more customized.

By it's very nature, open source allows any user or organization with enough expertise to tailor software to their needs. The diversity of Linux distributions reflects this flexibility. Each distribution offers a customized operating system targeting a specific market. There are even two competing graphical desktop interfaces: KDE and GNOME. On the backend, the open, modular nature of open source solutions allows advanced users extraordinary power to customize any aspect of a network (e.g. firewalls, spam filtering, email filtering).

Beyond customization, anyone can make significant changes to open source code. Of course, most educators don't have the time or expertise to do this. However, they see significant benefits from others' access and customizing. For example, the K12 Linux Terminal Server Project (K12LTSP) is based on the more general Linux Terminal Server Project. Eric Harrison customized K12LTSP to the needs of schools. Many educators value Harrison's work even though they never "hack" the source code themselves.

The car analogy illustrates this potential advantage. Not everyone has the desire or ability to customize their car and the same is true for software. But Eric Raymond asserts that "Every good work of software starts by scratching a developer's personal itch." (Raymond, Cathedral) Some automobile enthusiasts enjoy customizing cars; some programmers enjoy improving code. A specific program may be the best solution for a school in ways the original programmer never imagined. Anyone can adapt the software without asking or paying the programmer or a company. Such adaptations lead to innovations which eventually benefit the whole software-using community. Proprietary software may offer new features or flexibility in each new version. But open source will continue to offer greater ability to customize through the so-called right to fork: anyone can take the source code and develop it in a new direction.

[back to top]

Open formats

Open formats are better.

Open source usually uses open formats. As a result, open source programs are generally better at working together and sharing files. For example, the Free Standards Group oversees the Linux Standard Base and the OpenI18N standard, which are designed "to make it easier for software to run on different companies' versions of [Linux]." (Shankland, 2002, Verizon)

Data management offers another example. Many states are trying to integrate their data management systems throughout all schools and districts. Open formats make integration much easier. For example, suppose different districts in the same state have different database programs, and each program has a proprietary format. To integrate the data, the state would have to select a program that licenses all those formats, or convert to the data to the new format, or abandon some data. Using open formats forecloses on these problems and may curb opportunity costs.

The greatest potential advantage of proprietary formats is quality control. By controlling and licensing a format, a company should be able promote smooth file exchange between programs (e.g. the different programs in Microsoft Office). Open formats may mean a lower TCO, however, especially since the opportunity cost is high if a school locks data into a proprietary format.

The open source movement is partly a response to incompatibility in proprietary software.

[back to top]

Software for schools

Open source is not mature enough for schools.

Open source is not a new model and much of the software has been used and improved for years. Linux is a UNIX-like operating system and UNIX is mature and reliable. The Internet depends and thrives on open source software. Business and government are deploying open source solutions. Some open source solutions are mature enough for schools. As more educators join the open source community they are creating education-specific programs.

There are no open source solutions for some school needs.

Some curriculum software is incompatible with open source.

However, critical programs may not run under Linux (even with emulation). Microsoft Windows has become the dominant platform for education software like Reader Rabbit. The community may never deliver this kind of specialized software, and existing companies may resist developing for multiple platforms. Current users in K-12 acknowledge there are no viable open source solutions for payroll, finance, or student information systems. There are few open source solutions for some needs, especially assistive technology for physical or leaning disabilities. It may be easier to develop such solutions since Linux has a command line and open standards, but development and support of any K-12 solutions will be proportional to interest and use. Some proponents argue that's an acceptable trade.

Open source supports better curricula in technology.

For basic productivity tasks like word processing, open source offers attractive alternatives to proprietary software. Furthermore, technology literacy should mean more than skill in specific applications. As future employees and consumers, students need be taught general skills to avoid unnecessary dependence on any specific program (e.g. Microsoft Office). Students will see software change many times before they graduate, so they need to be lifelong learners of technology. However, students approaching the job market (e.g. high school juniors and seniors) need to be taught the software employers want, and this will usually include proprietary software (e.g. Microsoft Windows and Microsoft Office).

For advanced curricula in technology, students will benefit from studying open source, especially if they want to pursue careers in technology. Just as writers benefit from studying literature, programmers benefit from studying open source code. (Lessig, 2001) Students should be exposed to the technical and policy issues of open source (e.g. by studying Web sites like this). Unfortunately, K-12 education is currently struggling to teach core curricula while integrating new standards and triaging budget cuts. Schools may not have the resources to experiment with open source, much less create the advanced curricula open source can support. And migration is a daunting barrier.

[back to top]

Open Options is a product of the Northwest Regional Educational Laboratory. These materials are in the public domain and may be reproduced without permission. The following acknowledgment is requested on materials which are reproduced: Developed by the Northwest Regional Educational Laboratory, Portland, Oregon.

This Web site was developed and maintained by the Northwest Educational Technology Consortium. The federal funding for the regional technology consortia program ended on September 30, 2005, and no further updates are planned unless additional funding becomes available. However, much of the content is still useful and NWREL will continue to provide access to this site to support educators and to meet its own technical assistance needs.