MediaWiki: a case study in sustainability
by Pete Cooper on 2 May 2008 , last updated
MediaWiki is server-based wiki software, designed for use on large-scale websites. It is readily available to download as open source software. It is used widely by a number of high-profile websites including, notably, Wikipedia and Wikimedia Commons. A key feature of MediaWiki is rich content generation and display, including mathematical formulae via LaTeX and an increasing range of media files, including user-uploaded content. MediaWiki can be used both internally within a networked organisation, or externally on the Internet.
MediaWiki is actively maintained and updated. It is released under the GNU General Public License v2, allowing for wide deployment and customisation. It is written in PHP and requires either a PostgreSQL or MySQL database. All of these dependencies are also open source, though the licence specifics differ for each package: PHP is released under the PHP Licence, PostgreSQL is available under the BSD License, and MySQL is dual-licensed as both GPL and a commercial proprietary licence, dubbed MySQL Enterprise, for operators not wishiunrestrictedng to be bound by GPL terms.
The MediaWiki software was originally written by Lee Crocker for use solely in Wikipedia, and was a re-writing of Magnus Manske’s earlier MediaWiki PHP script. Before that, Wikipedia originally ran UseModWiki, a Perl-based wiki engine written by Clifford Adams. The requirement for additional functionality led Wikipedia to adopt Manske’s PHP-based wiki engine in place of UseModWiki. As the popularity of Wikipedia increased, scalability problems led Crocker to re-write Manske’s software to use a MySQL database, allowing for greater scalability and performance. With the formation of the Wikimedia Foundation in 2003, the software became known as MediaWiki. Early stable releases of the software, made available in the second half of 2003, preceded the first official production release, labelled MediaWiki version 1.1, of December 2003. Updated versions of MediaWiki have been released typically every three to six months, incorporating optimised code designed for increased scalability and inclusion of features and functionality driven by user demand. The development and history of MediaWiki has been largely shaped by Wikipedia, Wikia and related large-scale wiki projects using the software, both non-profit and commercial.
Growth in install base and end-user base
MediaWiki is readily available to the general public, has few barriers to entry with regard to hardware and software requirements, and the code is widely considered to be of at least production/stable quality. Combined with the increasing popularity of Wikipedia and associated social and collaborative aspects of the internet, the proliferation of wikis is increasing. Due to the flexible nature of the software and licence, allowing for extensive use on intranets and the Internet, it is not possible to accurately quantify the number of MediaWiki installations or users. Installation of a MediaWiki wiki is relatively quick and simple, typically taking less than an hour. An increasing number of Linux and UNIX distributions are readily available with MediaWiki installers and updaters included, simplifying the installation process and ensuring that maintenance and updating are not avoided or postponed due to an excessively time-consuming or esoteric process.
Significance of Wikipedia, impact on MediaWiki
Wikipedia has used MediaWiki, along with pre-MediaWiki software incarnations, since it ceased using UseModWiki in 2002/3. At the time of writing, Wikipedia ranks in the top ten of most-visited sites on the internet. It includes thirty million articles across its network of sites, nearly two billion words and edits numbering in the hundreds of millions. Much of the development work on MediaWiki stems from feature and additional functionality requests from users and operators. Scalability, especially requirements for database replication across multiple server farms, has turned MediaWiki into an enterprise-capable collaboration tool, ideally suited for knowledge-sharing environments. The deliberate inclusion of methods to extend functionality, known as extensions, are actively allowing content creators and developers to increase the scope and capability of MediaWiki.
At the time of writing, the lead developer of MediaWiki was Brion Vibber, who was then also CTO of the Wikimedia Foundation. As Wikimedia relies heavily on the stability and development of MediaWiki, development is assured, at least at the present time. Although MediaWiki is released by the Wikimedia Foundation, there are over 130 maintainers and contributors at the time of writing. The Wikimedia Foundation facilitates sustainability and growth of MediaWiki by offering extensive documentation and resources for users, operators and developers, giving each group similar weighting with regard to documentation availability. The documentation and resources are themselves available on a wiki, allowing each of the three groups to contribute and improve articles. MediaWiki is written in PHP, a widely used and established scripting language, allowing for code modification in any text editor without the need for expensive or proprietary software tools.
The relatively low barriers to entry, general availability and maturity of the MediaWiki code base indicate a sound future for sustainability of the project. More users are discovering and getting involved with Wikipedia and other wikis. Although generally a small percentage of a wiki audience actively edit pages, familiarisation with a standardised user interface and layout will increase user confidence and ultimately drive future development, helping sustainability. If a user is able to participate actively, whether it’s authoring or editing articles, or simply being able to make minor semantic corrections, they will feel more involved with the community. This, in turn, enhances the reputation and significance of the wiki. If the user feels value and attachment to the wiki, they will continue to contribute where they can, thus building community momentum and adding to the wiki sustainability.
Many MediaWiki installations are left in their default configuration, either deliberately or due to operator oversight. One side-effect of the default configuration is the ability for any user to modify content on the wiki, in some cases without authentication. While some users have entirely good intentions regarding wiki content, other users regrettably abuse this trust. Defacement and vandalism of MediaWiki pages is common on busy sites. Although these actions can be undone be reverting to a previous revision of the page, left untreated they can negatively affect the appearance and perceived value of the wiki, either to individuals within the organisation, or to the wiki user community and other visitors. If wikis are seen to be arduous to maintain, the install base is threatened. If end users regard wikis as low value and not actively maintained, their long-term involvement may be in doubt. In each of these instances, the sustainability of both the local wiki and its software is adversely affected. On a larger scale, widespread dissatisfaction of operators and users will drive the need for significant changes in the software architecture or a need to seek alternative software from another party.
The strength and global reach of Wikipedia makes it a strong leader in the domain of wiki software. The choice of software used to power this and other Wikimedia sites will automatically be seen as a market leader, at least for the foreseeable future. Critics will continue to point out weaknesses in the Wikipedia model, such as ignoring contributor’s credentials (Wikipedia does not use a democratic process, nor does it rely on credentialed experts, but instead operates a consensus model). These perceived weaknesses allow competitors to emerge, but while Wikipedia remains as popular as it is today it would seem that MediaWiki will remain viable.
In the event Wikimedia changes the software behind its wiki websites, the move will no doubt be monitored closely by wiki operators. Regardless of any changes made, many established sites will continue to use MediaWiki due to the significance of the install and end-user bases. Total cost of ownership is also a factor, with both operators and users being faced with learning the configuration and syntax of an alternate wiki platform.
With a mature and respected code base, acceptance from high-profile websites and track record powering such a major site as Wikipedia, MediaWiki is an attractive proposition for the public and private sector. Comparatively low requirements to be a contributor and active development from both Wikimedia and third parties will continue to strengthen MediaWiki’s position as leader in the wiki software marketplace. The global proliferation of MediaWiki and its related dependencies will ensure that, in time, users will come to expect MediaWiki-powered wikis, if not by name, then certainly by appearance.
For a more general discussion of sustainable open source, see our briefing document Sustainable open source.
- MediaWiki [http://www.mediawiki.org]
- Wikipedia [http://www.wikipedia.org]
- Wikipedia: Not a Democracy [http://en.wikipedia.org/wiki/Wikipedia:NOTDEMOCRACY]
- Wikimedia Commons [http://commons.wikimedia.org]
- LaTeX project [http://www.latex-project.org/]
- List of sites using MediaWiki [http://www.mediawiki.org/wiki/Sites_using_MediaWiki/en]
- Wikipedia entry at Wikipedia [http://en.wikipedia.org/wiki/Wikipedia]
- MediaWiki contributors [http://www.ohloh.net/projects/mediawiki/contributors]
- MediaWiki Developer hub [http://www.mediawiki.org/wiki/Developer_hub]
Related information from OSS Watch