CategoryWikis

Semi-automated deletion nomination at Commons

I just saw that the JavaScript wizards at Wikimedia Commons came up
with an impressive new tool – if you look at an image description page, you will now see “Nominate for deletion” link in the bottom
right corner. If you follow that link and give a deletion reason,
everything – the tagging of the image, the listing on the Deletion
requests page, and the notification of the uploader – is done
automatically using JavaScript.

This is quite impressive. I’d love to see more of this kind of
automation enabled by default, at least for users in the
“autoconfirmed” group (registered accounts older than X days). Think about it:

  • Deletion, peer review, featured article status nominations
  • Speedy deletion with auto-notification to the affected users
  • updating news pages & portals with important announcements

I’m sure there are countless scenarios where this might come in handy.
I can see the dangers, but I think the benefits justify some more
experiments. Any takers & other examples of similar semi-automated
tools?

Citizendium is not Free Content

Nearly a month after its public launch, Citizendium (a new wiki-like encyclopedia that positions itself against Wikipedia) still has not figured out its licensing policy. While the project has no choice but to follow Wikipedia’s GNU FDL when it imports articles from there, its own pages are still under undefined terms. Contributors are asked to wait while (someone) figures out the licensing terms: “All new articles will be available under an open content license yet to be determined.” It does not say who will make that determination and on what legal basis they even have a right to do so. Unless CZ decides to ask every contributor for permission to relicense, authors would have a very good claim to question a licensing decision they do not agree with.

I hope that Citizendium will become free content eventually, instead of adopting odious restrictions like “no commercial use” which would make subsets of it incompatible with Wikipedia and other free knowledge resources, not to mention making an awful mess of the editing process. Meanwhile, the content created by CZ contributors is completely proprietary and not usable by anyone beyond its publication to the CZ website. The few high quality articles they have developed and which could potentially be merged back into Wikipedia are non-free. I would caution anyone who contributes there to at least explicitly license their content (for example, by putting a licensing template on their user page).

An Adventurer is You!

I love it when I discover utterly bizarre and wonderfully unique worlds on the Internet previously unknown to me. And of course, as is so often the case, I did so when browsing Wikipedia. Specifically, in the article about NetHack which I check occasionally for updates to this deeply fascinating comptuer game classic, I found a reference to another game called Kingdom of Loathing I head never heard of. The article is pretty informative, though I think the intro written by the game’s designers gives you a better feel for the game. 😉

Essentially, KoL is a browser-based role playing game and online community, but instead of fighting giant rats or hordes of the undead, your enemies are sabre-toothed limes, ninja snowmen, and fluffy rabbits. Items (“filthy corduroys”) and character classes (“disco bandits”) are equally bizarre. But what is the most surprising (and perhaps concerning) is the number of active players. The KoL community is huge, with more than 2 million messages posted to the game’s forums, and thousands of players logged in at a time. The community is further sustained by large fan websites, “clans”, and frequent real-life meetings. The game is financially supported by donations and merchandise.

What makes it, to me, more fascinating than other similar browser-based game communities is the incredible level of surrealism and satire. It is in some ways a complete abstraction of certain RPG core principles like quests, skills, levels, magic, all these elements being replaced by jokes and nonsense. The visuals are literally doodles and stick figures, and interactivity is limited by the minimal browser interface. Still, in spite of the lack of an environment that could possibly be immersive without additional drug use, all the core RPG mechanisms seem to be as addictive as ever to its user base (though I would imagine that the humor also helps).

Within Wikipedia, factions often dispute the usefulness of articles about “non-notable” web phenomena like this one, because they tend to not receive significant coverage outside the web’s micromedia. I’m glad Wikipedia has an article about KoL, especially because no other place would provide me with a neutral, comprehensive summary of such a bizarre subculture. Indeed, I hope that the wikisphere will encourage and drive original research into these topics — not in Wikipedia itself, but in other spaces like Wikiversity and Wikinews. Even within Wikipedia, I hope the bias against using primary sources in documenting projects like KoL will decrease. Indeed, as I mentioned previously, I think wikis have the potential to take referencing to new levels.

And why, you might ask, is it even important to understand such an obscure, silly phenomenon? Why is it important to understand gaming culture, furries, or TV fandom? Should we not dismiss such embarrassing cultural idiocy, and lead humanity towards a golden age of a new enlightenment? I believe in the latter, but not in the former. If we want to advance as a species, we must understand what makes us tick. We must develop models that help us to explain why people form online communities around the idea of hunting menacing citrus fruits. If we can accurately predict these motivations and their underlying patterns, we can make use of this knowledge to build sustainable communities dedicated to human progress. Should, for example, a project like Wikipedia make use of RPG-like mechanisms to build motivation for routine tasks? Probably not, but right now we are stumbling in the dark when it comes to predicting the effects particular mechanisms might have, because we have no empirically sound framework to place them in. We can use trial and error, but the more errors we make, the harder it gets to justify more trials.

Information gathered about a project like KoL should eventually be part of a massive database with an overlaid ontology which allows us to compare it to similar communities (online and offline), analyze growth patterns, see relevant case studies of conflicts and procedures, and so on. Psychologists, economists, sociologists, historians, neurobiologists, and researchers from many other disciplines ought to work together in developing unified models we need to engineer the rules and structures of networked communities systematically towards certain ends. That will require science itself to mesh into a networked community, independent of institutions and disciplines. We see the early beginnings of this in the wikisphere, but also in the open access movement with PLoS leading the way in web-based innovation of the scientific process. But there are still great challenges to overcome, ranging from proprietary licensing and closed data over institutional vanity and academic arrogance to short-sightedness in policies for communities like Wikipedia.

And that’s why you should care about ninja snowmen. :-)

Zotero & Wikipedia

… perfect together.

Zotero rocks. It should be part of the toolset of any serious Wikipedian.

Video game sequences as sources

People often make fun of Wikipedia’s obsession with pop culture. While it’s mildly amusing to look at the wiki bureaucracy surrounding, say, Pokémon related articles, this obsession has made Wikipedia a most unusual publication that takes a scholarly approach to topics that are not typically treated that way.

Currently, the article about the video game Devil May Cry is a candidate for featured article status. Looking at the article, one thing I had never noticed before is that dialogue from the game is used in footnotes to back up particular statements about the plot.

How does one reference a segment within an interactive resource, taking into account also the typically increasing difficulty and dependency on prior interactions in video games? This demonstrates the degree to which Wikipedians explore scholarly approaches even in areas of culture that are shunned by traditional academia (but which will without doubt be of great interest to the historians, sociologists and psychologists of the future). I suspect that the ideal solution for this would involve savegames or even memory dumps. Certainly, it is an easier problem to solve in the context of open source games, where the reference can go directly into the source code. I haven’t seen examples of that yet, though I would expect the NetHack article to eventually be full of such source code references, perhaps directly to the code line numbers in a particular version.

By permitting anonymous contributors and adopting an egalitarian editing model, Wikipedia has forever condemned itself to researching and documenting even the tiniest facts in its most obscure articles. Credentials count for nothing, and unreferenced statements can be removed. The implicit commitment to collaborative research by volunteers on an unprecedented scale will surely bring interesting new challenges and could redefine what it means to cite a source.

Wikimedia’s Open Source Toolset

There is a rarely explored relationship between the world of open source tools and free knowledge collaboration. This relationship developed naturally and quietly. Some open source tools have become quite essential to projects like Wikipedia, and I’ve started a page on our Meta-Wiki called Open Source Toolset to document the use of open source/free software tools in the Wikimedia Foundation projects. Others have quickly made useful additions (if you see any glaring omissions, please do not hesitate to edit).

Inkscape is an example of a mainstream open source tool that has become essential, even though it has not reached 1.0 yet. It has been used to create thousands of vector drawings in Wikimedia projects. But there are much more specialized tools, such as Hugin (used for stitching panorama pictures) or PP3 (used for celestial charts). The availability of these tools is incredibly empowering. Anyone with the necessary skills and interest can use them to immediately contribute their knowledge; there is no charge, and the quality of the software almost always increases over time.

The importance of this open source ecosystem of tools can hardly be overestimated. Every new tool, every new feature, directly feeds back into the quality of content that is being generated. Therefore, I strongly believe that we must find ways to support them. Google has an annual Summer of Code, through which it spends a lot of money on student projects. This is very worthwhile indeed. We do not have a lot of money, but we do have global website exposure. Perhaps the Wikimedia Foundation should support its own “Autumn of Collaboration”, providing learning resources and guiding volunteers to work on the projects that make the greatest difference in the collection and development of human knowledge.

A bunch of buttons

The Definition of Free Cultural Works has been officially adopted by the Wikimedia Foundation. This gives us some major visibility, and we want to make use of that visibility by encouraging people to adopt a new set of free culture buttons, designed by amymade.com. Here are some examples:

I asked Amy to make three different colors for different types of works (such as music, scientific papers, or weblogs). I’m very happy with how they turned out — if you have a CC button for a free license on any of your works, you may want to consider replacing it with one of these buttons. I hope they will be adopted within Wikimedia as well.

Phyllobius argentatus

A lovely new featured picture candidate on Commons:

It’s used on Wikispecies. You can vote voice your opinion here.

Eventually, Wikimedia Commons will become the real ARKive.

Wikipedia Gets Advertising

Wikipedia has finally seen the light and started putting on some animated banner ads. Moreover, they’re not limited to the article space, but placed where editors are most likely to see them: on the pages showing information about users, and the internal communication system (“user talk pages”). It seems that the Wikimedia cabal even recruited the users themselves to add the ads to the pages (possibly through some kind of revenue sharing arrangement). Here are some examples:







It’s good to see that Wikipedia has given up its communist pinko resistance against advertising. Now that they’re raking in the money, I hope they’ll finally stop asking for donations like some maggot infested hippie. Jason Calacanis was right again!

(More ads here)

DBPedia: Querying Wikipedia

Ever since templates were added to Wikipedia, many people have thought about ways to make all the exciting data from the various infobox templates added to articles searchable in some way. DBPedia has done it: they have converted all the template data into RDF and made it browsable and searchable. Using, for example, the Leipzig Query Builder, you can ask questions like “which film composers were born before 1965” or “which soccer player with tricot number >11 from club with stadium with >40000 seats has scored more than 100 goals”.

Of course, it is desirable to store this data in a structured form in the first place, and to make it searchable in real-time. This is what Semantic MediaWiki and OmegaWiki are aiming for, using different strategies (one provides you with a syntax to annotate wiki pages and stores all annotations in a machine-readable fashion; the other focuses only on the structured data itself and treats it separately from a wiki page). I believe that, after we’ve made some progress on quality annotation (priority number 1), adding content structure to Wikipedia should very much be the next priority for the Wikimedia Foundation.