User:Eloquence/State of the Wiki: February 26, 2005

A lot has happened since my last State of the Wiki summary. First and foremost, there are now 10 Wikinews editions: English, German, Dutch, Swedish, French, Spanish, Bulgarian, Portuguese, Polish, and Romanian (launched in this order). This report is, by necessity, somewhat focused on the English edition. I encourage editors to submit interesting developments in their local Wikinews community to the international Wikinews mailing list.

Multilingual progress
Let's start with some statistics. Here is the number of pages in the article namespace, as of today, on the different editions (any page without a namespace prefix containing at least one link):

[*] includes many index pages which are not articles

English and German are the leaders in terms of Wikipedia article count, and they were launched first, so their high position is to be expected. Swedish is surprisingly large, which is owed to a very dedicated core team of editors. Polish, only launched very recently, is off to a very promising start and has already taken the fifth position. The biggest surprise in the list, however, is French, which was launched together with Swedish and Spanish. The French Wikinews is still very disorganized, and there is no regular output of stories yet.

It is notable that, in the global Wikinews vote on whether to launch the project, there was actually a majority against the project on the French voting page. However, since the vote was global, the same standards are applied to all Wikinews editions (with the possible exception of Chinese). When the required number of pledges by potential participants was reached, the project was launched in accordance with the agreed upon policy for new languages. This led to some protests from the French language community, who felt that they should have had the opportunity to decide (by vote or otherwise) whether they want a Wikinews project in their language.

I don't think that is the right approach, but I do agree with those who have argued that simply pledging to participate in a future language edition is not enough. The amended policy now requires a certain minimum number of edits on existing Wikimedia projects for at least some of the pledges. In addition to or instead of that, I believe that the requirement to translate or create some key Wikinews documents (mission statement, neutrality in news articles, article tags, etc.) before a new language is launched would make sense. This way, the actual desire to work on that language edition could be put to the test before we create the wiki. More on what to do when a project has become temporarily inactive below.

Logo, portal and design contest
After a long voting process, we finally have a Wikinews logo. Some people have expressed concern that the logo looks remarkably similar to that of a certain global organization, and that the light blue color does not go well with the gray background. David Vasquez, who created the logo, has made a few logo variants and is open to suggestions to refine his work. I think he has done an excellent job so far. Critics should be silenced by the history of the Wikipedia logo. The first logo used by Wikipedia was an American flag.

Like Wikipedia, Wikinews now has a central portal page at www.wikinews.org. The components of the page can be edited on Meta-Wiki: Www.wikinews.org portal (the main body) and Www.wikinews.org template (the surrounding HTML, only editable by sysops).

Unfortunately, Wikinews currently looks exactly like Wikipedia except for the logo. I believe we should attempt to create a unique visual identity for the project. To this end, I have launched a Wikinews design contest for the portal and stylesheets (MediaWiki:Monobook.css). Please promote this contest in your local Wikinews edition and elsewhere on the net.

Content license and fair use
Wikinews content is currently still in the public domain. Please make sure that your local MediaWiki:Copyright and MediaWiki:Copyrightwarning are clear about this. The main reason for this is that the public domain allow us to switch to any other license, or combination of licenses, in the future, so we can gather data about the licensing requirements for Wikinews before making a final decision.

The key underlying question is: Is it more important to us to maximize the copyleft effect -- that is, derivative works of Wikinews articles have to be under a free license -- or is our primary focus to achieve maximum distribution? Right now, under the public domain, the New York Times could take any Wikinews story and make it the basis of their own reporting without giving us any credit (unless one of the contributors holds moral rights to the work, which includes the right to attribution, though this may be difficult to enforce across national boundaries). A license like Creative Commons Attribution-Only (CC-BY) would require attribution, but nothing else. The GNU Free Documentation License (GFDL) on the other hand, has a number of very onerous requirements, which has led some prominent open source and open content projects to reject it,. If we choose another license than the GFDL, this also complicates combination of Wikinews and Wikipedia article content.

My personal opinion is that copyright doesn't make sense, and that if we want to build a society without coercion and restrictions on information use, we should set a good example by using the public domain. Richard Stallman's copyleft approach is essentially a "trick" to use copyright against itself; this sounds like a good idea in theory, but in practice creates the false and dangerous perception that open source software and open content require copyright law to exist. On the other hand, I also think the issue is vastly overrated in importance, and should not be used to drive a wedge into the free content community. I am happy to accept any decision we can come up with.

When is this decision going to happen? Unless someone else steps up to the plate, I will probably organize a vote in early or mid-March. This is not entirely trivial as we have to properly distinguish between different dual licensing options (primarily CC-BY/GNU FDL and CC-BY-SA/FDL).

A related issue is the question of fair use on Wikinews. Most Wikinews articles are summaries of stories from various outside sources. This is, or should be, perfectly legal and does not fall under "fair use". It is not possible to copyright information, only a specific expression thereof. Or at least that's what intellectual property lawyers keep telling people.

In practice, the distinction between copyrights and patents is continually being eroded through the idea of "derivative works". An example of this was the attempt to halt the publication of The Wind Done Gone, a parody of the book Gone With the Wind. The copyright industry has very actively tried to make it possible to protect information itself. So, at some point, news publishers may accuse us of creating derivative works of their stories.

Our goal is to become a player in the global multi-billion-dollar industry of newsmaking, and as such, we have to expect that these companies will try to kill us when we become relevant. Besides a tremendous power to manufacture consent, Murdoch, Bertelsmann & Co. have an army of lawyers at their disposal, and they will not hesitate a split second to use it. This is why I believe we must be extra careful when resorting to copyright exemptions like "fair use", especially for photos.

We directly compete with news organizations whose photos we could potentially use. Regardless of whether or not that makes our fair use claims less legally plausible, I believe it increases our risk of being targeted. For this reason, fair use of photos from other publications is currently not allowed on any Wikinews edition. Local uploads have been disabled on all Wikinews editions, and uploads are only possible to the Wikimedia Commons, which does not allow non-free content.

That does not mean that we will not allow any fair use whatsoever -- screenshots and logos are fair game, for example. However, every Wikinews edition needs to formulate a clear policy on the matter before local uploads can be re-enabled.

Wikinews and blogs
On February 5, the first open Wikinews Chat took place on IRC. The topic was potential present and future interaction between Wikinews and the blogosphere. Thanks to a lot of publicity in various blogs, the chat attracted plenty of high profile people, including Joi Ito, Rebecca MacKinnon, Kevin Marks of Technorati, Pete Carr of Chatmag.com, Susan Crawford, Nelson Pavlosky of freeculture.org, Adina Levin, Jonathan Dube of CyberJournalist.net and MSNBC, and many others. Wikipedia founder Jimmy Wales and elected trustee Angela Beesley were also present. At peak times, we were chatting at a speed of about 20 lines per minute!

My primary goal with this event was to get bloggers excited about Wikinews. One idea which has been discussed was the creation of a global real-time forum for citizen journalists from various different projects (individual blogs, Indymedia, OhMyNews, etc.). This would be quite a challenge, but something I would like to pursue further. In order to get an RSS feed in spite of our wiki software not yet having the necessary functionality, Dan100 started a Wikinews blog which links to the latest English stories (though it does not include their actual content). We're still quite a long way from real RSS feeds (see below), but I think the connections we have established with this chat will help us when we can integrate Wikinews better with the blogosphere. I also encourage all Wikinewsies to participate in building an index of relevant blogs to observe, and to directly invite bloggers to participate.

Thanks to everyone who made the chat possible, especially Amgine, who did most of the grunt work of collecting email addresses and writing invitiation emails.

Submit a story
In order to ease participation for people unfamiliar with wikis, I came up with the idea of a simplified story submission screen, which is now at Submit a story and has been translated into various languages. The idea is simple: To submit a story, you append it to the page using the "post a comment" feature. The community of existing Wikinews editors will look over these stories and decide which ones to turn into real articles and which ones to reject. Rejected stories are collected in the Submission log.

The feature has been used very actively, as the page history shows. Key to making it work is prominent exposure on the Main Page as well as the www.wikinews.org portal. It's also important that the instructions on the page are kept brief -- it is not intended for people willing to read policy pages. It's perfectly OK if people use it to submit letters, POV essays and similar unacceptable material as long as there is an active team of editors weeding this out. If your edition doesn't already have this page, I encourage you to create it. It is promising to become a key entry point for new stories.

Software needs
I am observing a general problem with the Wikimedia projects, in that the needs of the projects are not met by the open source development process of the underlying software, MediaWiki. The English Wikipedia alone has 200,000 registered users, yet there are only a handful of people actively doing work on the software. MediaWiki now has about 150,000 lines of code (including the language code, but not including many extensions), and is turning into an immensely complex piece of software.

A developer coding a new feature will not only have to make it work in their environment, they will have to make sure that it interacts well with the Squid proxy cache, that it doesn't interfere with future plans such as ESI support, that it works with MySQL replication, that it is absolutely secure against outside attacks (as a single such attack could do a lot of damage), and that it can scale to Wikipedia's size (one of the largest 100 English-language websites in the world). Coding conventions and documentation standards are increasingly enforced, and the feature also needs to be properly tested before it is put into use in order to avoid pissing off a community of hundreds of thousands. This is very different from when Wikipedia was hosted on 2 servers with very limited caching and little concern for system security.

I see this as a main cause for relatively little new feature work happening on MediaWiki at the present time. It's no longer fun and easy to do. See it like this: When you write a Wikipedia article, it doesn't stop working if it still has bugs. Nobody will delete your article just because you didn't follow a convention or made a small mistake. You don't have to test the article with an audience before being allowed to publish it. It's OK to submit something which is incomplete and buggy. And of course the only thing you need to be able to speak is the language of the wiki project you're working on -- no need to be able to grok other people's PHP code with very incomplete documentation.

Yet, projects like Wiktionary, Wikispecies, Wikicommons and Wikinews are in dire need of additional functionality. In the context of Wikinews, RSS support for categories would be one essential feature, automatic building of indexing pages another. The latter is currently stalled because of concerns about the scalability of the current implementation.

While the Wikimedia Foundation has now hired a very talented developer, he will have his hands full just keeping the codebase secure, fixing bugs, coding absolutely critical features and managing incoming patches. I don't see how this problem can be solved without a dedicated effort to research the project needs, to prioritize tasks, and to assign them to full-time or part-time developers. It is hard to exaggerate the relevance of the software issue. Wikinews would already have grown at a much faster rate if it was fully integrated with the blogosphere, for example. The same is true for other projects which are only slowly plodding forward due to missing functionality.

I think even the word "software" shows how much we still underestimate its relevance. "Code" is a much better word. Code is complex. Code is law. Code determines what we can and cannot do. And as long as we don't have all the code we need, Wikinews can and will not be able to fulfill its mission.

What to do when a project is inactive?
One notable difference from Wikinews to other projects is that a quick glance at the frontpage gives you an impression of the level of activity: if you only see old news, obviously the project is not very active. This is good if it is seen as an incentive to work on it, bad if it frustrates people and makes them give up on the project. Wikipedia, Wikibooks, Wikisource, Wiktionary in the different languages all have their phases of activity and inactivity, but nowhere are those phases more visible than on Wikinews.

So how do we deal with small language editions that have not yet reached critical mass? One idea would be to put a template on the Wikinews Main Page when an edition has become temporarily inactive:

The French edition of Wikinews is currently not very active. Please participate in building this website into a useful resource! Ideas for increasing activity are collected at Building Wikinews.

This helps people to understand that such phases of inactivity are natural in a small community and no reason to shut down a project. Here are a few specific suggestions to increase activity:


 * 1) Translate. It is much easier to translate a well-written article than to write a new one from scratch. Wikinews articles have to grow to publishable quality very fast, "stubs" aren't helpful, and references are a strict requirement. As such, writing them takes a lot of effort. But the English Wikinews, for example, is now publishing articles at a reasonable rate. As an author of that edition, I also get great satisfaction out of knowing that my work is going to be translated into languages I do not speak.
 * 2) Prominently link the equivalent of Submit a story - this helps to attract people who do not know about the Wikinews process and just want to get their feet wet.
 * 3) Have brief news items on the frontpage which aren't fully developed stories yet. This should only be done until real stories are being published regularly, so as to not conflict with the Current events pages on Wikipedia.
 * 4) Edit MediaWiki:Sitenotice and point people to the IRC channel, mailing list and Village Pump (or Water cooler was we like to say). Building a community helps to create a feeling of identity and the desire to protect that community, which is an incentive to do work on the site.
 * 5) Organize and participate in events like the design contest or the Wikinews chats. Point people on Wikipedia to these events to get them interested in the project.
 * 6) If you are a software developer, help us to implement MediaWiki features like category RSS feeds and dynamic article display (see above).

Simon says ...
Simon Waldman, Director of Digital Publishing for Guardian Newspapers, has written an interesting blog entry about Wikinews titled "Wikinews...dare I say what I think?" Apparently there is an insidious wiki mafia that will violently silence all critics. In all seriousness, in my experience the Wikimedia community is one of the most thoughtful and most critically thinking groups of people I have ever met. Under the banner of neutrality, people from all ideological persuasions are united -- and, of course, constantly bickering about everything. The very idea of Wikinews was (and continues to be) highly controversial and was intensely debated on the mailing lists and the Meta-Wiki before a solid majority voted for the project to be launched.

That being said, Simon makes a few good points, which I would like to briefly respond to.


 * o Most Wikinews articles have already been covered elsewhere. That's certainly true. Most articles in your daily newspaper have already been covered elsewhere. They come from a news organization like the Associated Press, Reuters, AFP, dpa, or ITAR-TAS, which distributes them to many thousands of clients. These stories have to be licensed. If you distribute them illegally, you can be thrown into jail. Copyright is a powerful instrument to control the distribution of news. Want to start a local newspaper? Good luck, if you can't afford to license content from the news agencies.


 * Simon also speaks enthusiastically of a "world of free news sites, blogs, newsreaders and rss feeds". That's great, but how many of these news sites are truly free? How many, on the other hand, require you to fill out a complex registration form before you can even read their stories -- including, for example, the New York Times Story about Wikinews? How many of them move their content into commercial archives after a few weeks? How many blogs actually do real reporting or summarization of news stories instead of just linking to what others have published?


 * How difficult is it to build a meaningful aggregate of blogs that report interesting stories? What rights do I have, as a reader, to make modifications to the content of blogs and news sites, or to distribute it? How many news organizations (let alone blogs) follow a strict policy of neutrality? How many cover highly regional stories from the entire planet in as much detail as global events? How many allow anyone to fix errors, to question sources, and to contribute material?


 * Free is not free, Simon. Wikinews content is currently in the public domain. You can do with it whatever you wish. You can start up your own newspaper anywhere in the world and use Wikinews stories as a basis to work with. Already, sites like The World Forum are redistributing Wikinews stories and enriching them in different ways. Wikinews should not be seen strictly as a reader-oriented site at this point. It is also useful for anyone who makes use of the content to save money on licensing stories from elsewhere.


 * Wikinews doesn't need to be complete. It doesn't need to be original. It needs to be free, and it is. To criticize Wikinews for being "what others are already doing" is like saying we should get rid of Linux because there are so many good operating systems out there.


 * o Wikinews should do only original reporting. This misses the point above, that free content has intrinsic value. Wikinews allows original reporting (see also below). Sadly, we do not have money to hire full-time investigative journalists, so we have to rely on the kindness of strangers. The original reporting space is also firmly occupied by activist groups like Indymedia at the moment -- obviously, it is easier to motivate people to write propaganda than to give neutral accounts of what happened.


 * I think this will change when Wikinews becomes more useful to readers (through category RSS feeds for different topics, for example) and more comprehensive. At this point, people will want to publish on Wikinews simply to reach its audience, which ideally should be far larger than that of Indymedia and similar propaganda websites.


 * o Wikinews needs to focus. We cannot force our volunteers to write about particular topics. It makes a lot more sense to categorize articles properly. Some categories will then develop naturally into more useful sources than others. Again, with proper software support, people will be able to subscribe to just the topics they are interested in.


 * o Wikinews needs an editor. I will try not to be too dismissive of that suggestion. Many people don't know that Wikipedia has had an editor (or "chief organizer"), Larry Sanger, for some time (until his resignation in March 2002). I think everyone would acknowledge that Larry has played a very important role in the early phase of the project, and that it would have evolved very differently without him.


 * However, in my experience, it is better to let authority develop naturally rather than to assign it. People like Amgine and Ilya on the English Wikinews enjoy a lot of respect because of the work they are doing. As such, other people will seek their guidance in controversies. Yet, they do not have any real power which they could abuse to get what they want even when the community disagrees with them. Their authority fades when their activity goes down. This is how it should be.


 * o Wikinews needs to join forces with the blogosphere. This is happening and we are actively pursuing it -- see above.


 * o Stick to Wikipedia. This was one of the points frequently brought up when the Wikinews project was originally launched. I can thus simply quote from the Wikinews FAQ:

Wikinews articles provide very detailed information about specific events ("Allende toppled: Pinochet new ruler"). Wikipedia articles provide condensed information about a series of events (History of Chile, Augusto Pinochet, etc.). Also, Wikinews allows original research, while Wikipedia does not (Wikinews may eventually become reliable enough to be cited as a source on Wikipedia).

Let's take as a specific example the article Hurricane Charley on the English Wikipedia. In Wikinews, we would not just publish a brief timeline as in Wikipedia, but a detailed article about
 * what the forecast of the hurricane path and strength for the next day is
 * according to different experts
 * what preparations are being made by the local authorities (evacuations etc.)
 * what the situation is like for the people living in the storm area (live wiki-report)

Where Wikipedia would just summarize something like [http://www.nhc.noaa.gov/archive/2004/dis/al092004.discus.067.shtml? this report] in one sentence, Wikinews might well reproduce (if in the public domain) the whole thing in order to give people as much useful "live" information as possible.

So the two projects are very different, and in many cases, the work of summarizing the Wikinews article in Wikipedia will not be much different from the work of summarizing other sources like CNN or MSNBC.


 * Of the Wikinews articles which have been written so far, only a very small fraction has been covered on Wikipedia, and if so, usually only in a few sentences on the current events page or a relevant article. Major events like elections or disasters are a special case, as they are deserving of individual Wikipedia articles; yet, in those cases, Wikinews will ideally provide much more detailed coverage over a longer period of time. It does not make sense to compare the two in that department yet, because Wikinews is much, much smaller than Wikipedia.

Though I disagree with almost everything Simon wrote, he has clearly spent some time thinking about the issues, and I'm very happy that people are beginning to analyze Wikinews.

Accreditation and original reporting
Ilya of the English Wikinews has begun work on a page for requesting credentials to do original reporting. The really cool part is that he has set up a voicemail account which organizations can call to verify whether a person is an accredited Wikinews reporter. Incoming voicemails are forwarded as audio file attachments (WAV files) to an internal mailing list, where volunteers can then proceed to answer the verification requests by phone or otherwise.

Accreditation was one of the ideas I collected on the think tank page when I proposed the project. The first Wikinews edition to experiment with it, to my knowledge, was the German Wikinews: de:Wikinews:Akkreditierung. I would be interested in hearing a report from that community on how the process has worked out so far.

Original reporting on the English Wikinews is collected in Category:Original reporting, though so far examples are hard to find. The Unrest in Belize scoop has already been done to death, so I won't spend more time talking about it. Iraqi elections kept low-key, but secure, in Paris is a nice example of a Wikinewsie going out and taking photos to write a report about a local event.

Hopefully, we will move towards a more systematic process of sending volunteer reporters to events in all world regions. One advantage we have is that we're very good at translating -- once it's in English, German or Spanish, chances are pretty good that it will spread into the other editions.

Review
The original multi-stage review process of the English Wikinews has been scrapped in favor of a much simpler tagging model. If there are problems with an article, it will be labeled as such, and not prominently featured on the Main Page until the objections are resolved. This brings up an interesting point.

The exposure on the Main Page is itself a form of community review which does not exist like that on Wikipedia. Every published English Wikinews article will be shown there, and this temporary burst of attention leads to problems with the article being recognized relatively quickly.

Though a single-page model can obviously not scale, it is a process which has, in my opinion, greatly helped the quality of our articles so far. Certainly we will want to move towards a very rigid fact checking process, but right now, we have to get the momentum going and reach critical mass, in certain topic areas at least (on that I agree with Simon). I have also come to the conclusion that any advanced review process will require software changes in order to be maintainable.

Final words
As an example of where we might be headed, let me cite the somewhat silly story "Pfizer and Microsoft team up against Viagra spam". Pingswept and I wrote that article using a tool called MoonEdit, a collaborative real-time editor for multiple platforms. It's a clone of SubEthaEdit. Editing in real-time with someone else is quite a lot of fun, though MoonEdit is a bit painful to use (clumsy GUI, network problems) and proprietary.

Another example of how code determines what we can and cannot do, tools like this can massively increase the productivity of citizen journalists and the quality of their output. Imagine multiple such real-time editing sessions going on at any given time. There's more to it than just the shared workload -- it also makes the process of writing a more social, fun experience. I hope that eventually organizations like Wikimedia will take the lead in funding development of open source tools that can be used for this purpose.

As you can see, a lot has been happening in our little world, and I hope that if you aren't already involved with Wikinews, you will decide to give it a try or submit a story from time to time. If your language doesn't yet have its own edition, you can petition for it to be created on Wikinews/Start a new edition.

I apologize for this summary being a bit English-centric, and I hope that we can read similar reports for the various other editions soon. It is incredibly exciting and rewarding for me to see the project grow in many different languages, and it has always been one of the great strengths of Wikimedia to embrace internationalization, where most online communities don't. May we infect each other with the best ideas.

Erik Möller

Berlin, February 26, 2005