Month: March 2004

There’s been an ongoing conversation about wiki standardization. The conversation includes proposals for three different types of wiki exchange. Standards trace the shape of a community, like iron filings on a piece of paper over a magnet. This insight helps to explain how these proposals fit together in the world.
a) standard markup syntax. The Tiki crew have an RFC for an international wiki markup standard. This would mandate things like ** for bold and ” for italic.
The wiki world isn’t one big web — it’s composed of numerous creative communities, where people collaborate intensely together within the community, and have weaker ties with other communities. Therefore, attempts to impose a standard wiki markup esperanto are ambitious. They may be worth some attention, but will take a long time if ever to implement.
The fact that wiki markup is arbitrary will not make a standard easier to achieve. Setting a standard implies that some people will be able to continue with current behavior, and others will need to change behavior. There’s simply not enough internal incentive to cause people to change behavior, and there isn’t the external incentive of domininant market share.
b) interchange standard. These proposals don’t mandate wiki markup, but provide a means to exchange formatted pages. Since wikis all export HTML, there are proposals on the table to use a subset of XHTML as a lingua franca. This method would be somewhat “lossy” — specialized features like Purple Numbers and Twiki variables woudn’t get through. But text with the basic HTML formatting and structure would come through fine.
A wiki exchange standard fits nicely with the current pattern of wiki collaboration. People collaborate in their small and mid-sized communities, and want to exchange their results with others. A wiki exchange standard will take some negotiation, but seems feasible and worth attention, and valuable when achieved.
c) metalanguage. These proposals don’t mandate input or output formats —
instead, they abstract the general principles of formatting; parsing blocks and phrases, so as to be able to implement or emulate any arbitrary markup standard. WAFL is one such meta-language, recently implemented as part of Kwiki. There’s another, similar proposal at Meatball
In the mean time, before all of the standards efforts take hold, we’re living in a world with multiple markup flavors, and no defined output standard. In this motley world, there’s a pragmatic benefit in being able to emulate multiple sorts of markup. A metalanguage enables a wiki community to build readers or translators for other dialects, while still developing specialized vocabulary for technical documents, paragraph footnotes, or other nuances important to the subcommunity.

The design of online spaces reflects personal identity, danah boyd said in her excellent talk on the SXSW panel on the Esthetic of Social Networks. The “Fakesters” on Friendster didn’t mean that users were trying to fake out the system — they were expressing their identity by affiliating with icons, in the same way that kids put up dormroom posters, restaurants put up signed celebrity photos, and people put family photos on their desk at work.
There are analogs in the world of games, but much less in the realm of other social software. This insight points to an opportunity in social software design, and an opportunity to push the limits of some bad laws, too.
In the 3d world, people have many ways to decorate shared spaces — interior design in homes shared with family and guests; and exterior design of houses and gardens.
Today, flickr lets people share photos with IM buddies and others in social networking groups. This is cool as a feature, and would be even more powerful integrated with other online social spaces.
Music is a universal means of expressing shared identity. Today, there are widgets to publish a personal playlist on a weblog. There ought to be similar group tools to play music for online groups. Individuals could vote about choices, to maximize collective preferences. “Off” would be an important standard option.
This can be done legally with Creative Commons-licensed music from Magnatune and other sources of open-licensed music. And it should be legal — there shouldn’t be any difference, legally, between Joi Ito playing music for friends in his living room, and the #joiito IRC channel sharing tunes.
These features will be subject to intense negotiation, just as they are in 3D, where home decoration and neighborhood zoning are fraught with negotiation and conflict. The benefit of online spaces is that they’re not constrained to four walls, one set of color choices, and one playlist. Individuals should be able subscribe to some of the group’s choices, but not all. The consequence of semi-personalization could be greater tolerance and diversity, lower levels of affiliation, or some combination of both.
It will be interesting to see how danah’s insight plays out in the evolution of social software to reflect more of the cultural affiliation patterns of humans in groups.

Reflecting on two changes in web architecture to combat spam.
* LOAF is a new proposal for social spam filtering.
* AOL has been quietly blocking the websites of spammers
I ought to like the first (which uses a social network) and dislike the second (which relies on a centralized power). But my reaction is the other way around.
The blocking of spammers sites makes me want to cheer. This uses an age-old technique of punishing the anti-social with ostracism. The problem with blacklisting in general is lack of accountability. In order for the system to be fair, AOL should provide a test site for users to determine if the sites are being blocked, and there should be a public appeals process where people could get their cases reviewed and decided.
On the other hand, the blocking or slowing of all emails from strangers makes me sad. One of the beauties of the internet is the ability to meet new people with common interests. The restriction of social networks to people already in one’s social circle contradicts a core value of the net. This solution would take us back to the bad old world where you needed to know someone to get an introduction. The LinkedIn model becomes mandatory, not optional.
Blocking email from new people would seem to punish the innocent much more than blocking the websites of spammers. Mail from innocent new people will languish and die in the company of viagra ads. The senders will never know.
By contrast, if the spammer blacklist had diagnostic and appeals, the innocent could free themselves from the blacklist. The guilty would be ostracized, and the innocent could speak freely.
I’d love comments from sensible readers.