geof

Geoffrey Glass, 2004-11-29, updated 2006-05-04

The Internet is a site of conflict between two cultures: corporations, and an older Internet tradition of community. The differences between these two cultures can be seen in their use of time in online media. I will focus on the community culture and its use of time in four phenomena: blogs, wikis, open source software, and the Jargon File (also known as the Hacker Dictionary). Characteristic of these forms is that they embody a concept of a shared information and dialog: a commons.

The commercial sphere includes media conglomerates, the music industry, Hollywood, much of the software industry, some journalists. They see the Internet a source of profit. To their eyes, the non-commercial orientation of the community commons and its associated technologies, such as peer-to-peer networking and blogs, pose a threat to their power. (Many in the community respond that they are a threat to centralized power, yes, but not to profits or commerce.)

The conflict between the commercial and community views of the Internet has manifested in battles between lawyers, businesses, and technologies. The music industry has pursued online music sharing through the courts. The Linux operating system has arisen to challenge Microsoft's Windows monopoly. Bloggers have begun to exert influence on the mainstream media: They were the first to analyze and reject CBS's memos about George Bush's service in the U.S. National Guard (Koman, 2005). When mainstream media ignored comments by Senator Trent Lott supporting Strom Thurmond's segregationist policies, pressure from blogs elevated the story to national prominence; the story led to Lott's resignation (Goodgame and Tumulty, 2002).

Scholars have remarked upon the clash between these cultures. Siva Vaidhyanathan describes this conflict as between two ideologies – oligarchy and anarchy (2004, pp. xi-xii). Andrew Feenberg and Maria Bakarejieva (2004) write about two different models for Internet communication: the consumption model and the community model.

Many fear the consumption model is winning. In a recent journal article in the Canadian Journal for Communication, Catherine Frost expresses fears that "the Internet is becoming a controlled commercial product rather than an open public infrastructure" (2003, p. 19). Innis proposed that communications technologies have a bias either toward time or space. Cultures primarily oriented around space-biased media tend to dominate space and geography, while those oriented around a time bias are better able to preserve their culture and tradition over time. Frost applies Innis's analysis to the Internet and concludes that it is increasingly commercialized and space-oriented.

The contemporary era is characterized by speed and change. Corporations feed on this – they use innovation and fashion to compete and drive consumption. Schumpeter believes the race to innovate is inherent in the capitalist system (Hall, 1999, p. 297). Lawrence Lessig (2004, p. 255) describes how the movie industry fought a change to copyright that would have allowed forgotten films to enter the public domain at virtually no cost to the industry. He suggests the industry fought this (and won) because they didn't want competition from the past. Commerce focuses its consumers on Now; it reinvents itself and breaks with the past.

The community culture of the Internet, however, is more traditional. The earliest Internet culture is hacker culture, now a subset of the community culture (I use "hacker" in its original sense, which means simply a skilled or clever programmer; it applies to many of those who pursue the craft for enjoyment). This tradition is evident in the Jargon File, also known as the Hacker Dictionary.

Jargon File

The Jargon File is a single document; first started in 1975, it calls itself " common heritage of the hacker culture" (Raymond, 2003, "Welcome to the Jargon File"). Eric S. Raymond, who maintains the File, even argues explicitly for the heritage of the document ("Updating JARGON.TXT Is Not Bogus: An Apologia"). The authors of the Jargon File itself state that they "hope and expect" that it will play a "a central role in spreading hacker language and the culture" (Raymond, 2003, Chapter 1).

One might argue that, following Innis, the Jargon File, as a written medium, must have a space bias. But the File is is a living document. Whereas films, television programs, and books are more-or-less unchanging once created, hacker texts are dynamic. Successful texts are added to and modified over time, but they are seldom discarded or replaced. The File itself documents 31 "major" revisions between 1990 and 2004. This practice of updating existing documents rather than creating new ones is repeated in other online texts, including open source software, blogs, and wikis.

Open source software

The original hacker texts were software. At that time, before software became a commodity, the source code for virtually all software was treated like a shared resource, part of the commons – it was shared with other programmers.

This is the defining feature of open source software: the source code is freely available. Anyone who wants to look at the source code can do so; if they want to make a copy and change it for their own use, they can do that too. Most commercial software, by contrast, is not open source – if you buy the software, you can use it, but the code has been compiled into a binary form which is not human-readable and hence is virtually impossible to learn from or change. Many proprietary software companies, such as Microsoft, see open source software as a threat to their concept of software as "intellectual property". The hackers themselves are aware of the potential political implications of open source in areas other than software, as Steven Weber writes in The Success of Open Source:

Eric Raymond [probably the chief spokesperson of hackers in the open source movement] has said more than once that he wants to see open source win the "battle of ideas" in the software world and consolidate its home base before too much energy gets expended anywhere else. (Weber, 2004, p. 271).

Open source software has existed for decades within hacker culture; the concept was originally borrowed from the academic world, where scholars build on the work of others. Its recent successes have gained open source it visibility among businesses, academics, and the wider world. The Linux operating system is the most famous example. Open source is embedded in our technological infrastructure: the most widely-used web server is Apache, the standard email server is Sendmail, the original TCP/IP networking implementation on which the Internet is based – all are open source.

Open source software has a dramatically different relationship to time than does closed source software. The latter is like a book, both legally and practically: once it is published, its form is fixed. Later versions may replace it entirely; famous examples include the move from DOS to Windows 95 to Windows 2000 and Apple's upgrade from OS 9 to OS X. Open source software seldom breaks with the past so dramatically. The source code is available: even if the programmers working on a successful open source project abandon it, others are free to continue developing the older version.

Versioning, a feature of all software development, makes this possible. Version numbers identify a changing work at a particular point in time. Most successful software is continually modified; there is no finished state. Some code is added, other code is altered or deleted. Older versions are retained – both because they may contain code which might again be useful, and because changes can introduce bugs, in which case comparison with older working code can help resolve the problem. This slow evolution and link to the past distinguishes software from written texts, which are frozen and unchanging after publication, but it is similar to the development of oral texts.

In fact, there are several intriguing similarities between open source software and Ong's primary oral cultures – which, if Innis is correct that the bias of oral communication is time, may indicate a time bias in the medium. Ong lists several characteristics of such cultures; I have already argued that one of these – tradition (Ong, 1982, pp. 41-42) – applies to the Internet commons in general.

Ong also says that people in these oral cultures learn by doing (Ong, 1982, p. 43), just as hackers learn to program by writing software (Graham, 2004, p. 26). Hackers often write software as a practical way to solve their own problems: to "scratch an itch". In this sense, although the thinking that produces software is abstract, hacker software itself shares Ong's oral characteristic of being situational rather than abstract (Ong, 1982, pp. 49-50).

As Steven Weber points out (2004, pp. 157-158), versioning also presents a danger to open source software projects. Typically, multiple independent developers work together voluntarily to improve open source software. The risk is that one or more developers will disagree with the direction a project is headed, and will "fork" the code – i.e., create their own version (something commercial software does all the time: witness OS/2 vs. Windows NT, or the diversity of commercial UNIX implementations). When a project forks, the debates and recriminations can be fierce and damaging to the community as a whole. Worse, the split divides efforts to improve the software.

Hackers are aware of this. With no central authority to prevent a fork, there is a strong tendency to build consensus, typically on the basis of peer review and technical merit. Like Ong's primary oral cultures, open source communities must be homeostatic (Ong, 1982, pp. 46-48) and resist centripetal forces.

This need for consensus also drives one of the chief criticisms of open source software: successful projects tend to be conservative, focusing on perfecting existing technologies rather than inventing new ones (open source proponents claim this conservatism will pass as the movement matures). Linux, the flagship open source operating system, is based on UNIX, whose history goes back to 1969. The same phenomenon can be seen in the leading user interfaces for Linux, KDE and Gnome, with their clear roots in Windows and the Macintosh. The broad acceptance of standards – e.g. in the Firefox web browser – is another product of this need to build consensus; it also strengthens links to the past.

Blogs

Blogs (the word blog is an abbreviation of "web log") are a relatively recent product of hacker technology which have exploded in popularity over the past few years – largely due to the ease of creation by non-technical people (addressing a problem noted by Frost (2003, pp. 16, 19)). As of this writing, the blog search engine Technorati lists over 4.5 million of them. While many hackers reject blogs as technically uninteresting, they are deeply rooted in the hacker way of doing things.

A blog, like a diary, is a sequence of entries – "posts" – created over time. Some blogs are diaries, but many are more like newspaper columns. They may cover political issues, technical issues, scholarship, or news from different parts of the world.

By default, the web page for a blog displays these posts in reverse chronological order. As the author of the blog adds new content, the blog homepage changes to display the newest material. In this sense, like a newspaper or a radio broadcast, a blog is focused on Now.

But blog posts do not exist in isolation. Like a diary, they are part of a history – in many cases part of a narrative. Unlike a diary, blog posts are also part of the Web and often have hyperlinks. These links tie blogs together in a web of dialog: one blogger may write a post, then another may comment on it and link to it, and so on. In order to follow the web of discussion, or even understand a single post, a reader may need to able follow the trail of conversation backwards in time in order to understand a post¹. And even if a post is only stating an idea, or pointing to an interesting web site, if the idea or link originated on another blog it is considered good etiquette to link to that blog. These links are what set blogs apart – Technorati indexes 700 million links between blogs posts (unlike web pages, these posts contain only one article and no superfluous content or links, such as design elements or navigation). Blog syndication allows the content of each article to be isolated and referenced individually.

These links are not just references in space – they are references in time. The hackers who work on the technology are keenly aware of this, for the problem with links is that they can break. The solution is URLs (link addresses) which will never change.

Broken links have long been a problem. Tim Berners-Lee (1998), inventor of the Web, wrote in 1998 that web page authors should attempt to make "URIs so that they will still be around in 2, 20 or 200 or even 2000 years"². His favored solution was to categorize by date rather than content, for while other categories may change, the creation date of a page will be forever the same. Many blog technologies have adopted this convention, to the point where it has nearly become a de facto standard – a typical URL might be <http://www.someblog.com/2004/07/30/elections>.

Blog search engines solve another problem with time. Links are unidirectional. Following them can only take the reader from more recent posts back in time to older posts. What if one wants to start with an older post and go the other way? Search engines solves this by indexing links to posts. A search can then produce a list of all blog posts which have referenced a post..

So it is possible to create links that do not change, and which are useful for following the thread of a discussion over time. There is one more risk here: the Web is an infinitely plastic medium. As Frost points out, (2003, p. 21) Web pages are only a sequence of bytes and can be changed at any time. How will a future reader following a link know that the target of that link hasn't changed? Blogs don't normally support version numbers, so there is no way to guarantee this. Etiquette, however, rules that changes to a post must be noted as such (e.g. by prefixing changes with the word "Update"). This is necessary for the authority of the medium; bloggers who violate the rule are heavily criticized by the community.

This criticism is extended to traditional newspapers, and illustrates one of the fault lines between the community culture and the consumption culture. Many online newspapers force users to register to view articles, cutting them off from the flow of discussion and links. Papers may also change or pull stories. Many, such as the CanWest chain in Vancouver, simply allow articles to disappear from the Web, breaking any links to them. This behavior weakens the entire network, as blogs comment upon news stories.

The newspapers, for their part, may not want to participate. Just as software companies see open source as a threat, many journalists dismiss blogs as lacking in authority and objectivity. On the other hand, one survey of 17,00 blog readers found that 80% read blogs for information they could not find elsewhere, and 61% felt that blogs were more honest than traditional media sources (Blogads, 2004). Both sides make good points; it may be that journalists and bloggers will help to keep each other honest. Spammers may not be so benign: blogs already suffer from spam in the form of comments and links from spammers.

Although blog phenomenon is relatively new, businesses are already trying to figure out how to take advantage. Spammers already attack blogs with links and comments, although the medium has proven more resistant than email. More seriously, commercial media may absorb the medium. David Talbot, for example, founder of the independent media site Salon, suggests that Salon will syndicate bloggers (Glaser, 2004). If corporations succeed with such a strategy, the best bloggers may lose their independence.

Wikis

Wikis are another Web-based medium which gives its consumers even more control, and so challenges the consumption model more directly. A wiki is simply a web site that can be freely edited by its users. This appears to be a recipe for chaos, but so far has been surprisingly successful, as popular wikis – such as Wikipedia <http://wikipedia.org> – have demonstrated.

It turns out, in fact, that wikis – like open source software – are homeostatic. In order for content to last, it must conform to the consensus of the community; if it does not, it will be overwritten. Perhaps the broken windows theory of crime prevention applies here: those who might disrupt the community give up when they find the chaos they create doesn't last.

Part of the reason for the stability of wikis is time. Like open source software, they maintain version control for individual pages. If one user disrupts a page, another can simply revert the text to a previous version. As Peter Tupper notes (2004), destruction costs more than creation. Wikis are maintained by a process of peer review.

Wikis also share their aggregative quality with blogs and open source software. They grow and change over time, depending on the demands of their users. Like blogs, they can focus on the now – Tupper suggests they are ideal sources of news and discussion – but unlike traditional media they benefit from links and so maintain a stronger sense of history.

Wikis are relatively new; battles with existing media interests have been rare. Wikipedia, however, has been faulted for its lack of authority: every article is an amalgam of many different authors, who may be anonymous. This may present yet another risk that outsiders will hijack the medium.

Conclusion

The time-based characteristics of these new media demonstrate the conflict between the community-based Internet and the commercial Internet. So far, these media have largely withstood the overt attacks of entrenched interests. As they grow in influence, they may come under more subtle threat from commercial attempts to co-opt them. Some companies already profit by selling open-source software; for example, IBM has a major stake in the development of a number of open source projects; under pressure from Linux, Sun has pledged to release Solaris, its flagship operating system, using an open source license. The conservative nature of open source development is already strongly influenced by commercial innovation; as businesses invest in it, their influence grows.

Blogging is more recent, but there too commerce is having an impact. Many popular bloggers collect revenue from advertising on their pages; this will inevitably influence what some choose to write or report. Some businesses already use blogs – openly and covertly – as marketing tools. Wikis appear even more vulnerable: political and advertising interests could modify articles to suit them; with greater resources at their disposal they might effectively take over services like Wikipedia. They may also own wiki and blog services, effectively controlling the content. One popular blogging service, Live Journal, requires users to sign an agreement surrendering rights to any content they create. If creators do not control their content, the longevity of their work will be threatened by business interests in disposable culture.

Despite the threat of commercial encroachment, these media forms offer alternatives to a commercial Internet. Their values – of tradition, history, authenticity, integration, and collective authorship – contrast sharply with the short attention span of the consumption model, and more closely resemble the time-based traditions of past ages and oral societies. As the conflict between commerce and community continues, we may see evidence in the changing relationship between community media and time.

Notes

¹ This process is intriguingly similar to the evolution of common law.

² URIs are a superset of URLs – a URL can be a clickable link; a URI may not be.

References