User:Eloquence/State of the Wiki: April 28, 2005
|State of the Wiki|
|These are Erik Möller's semi-regular summaries of the state of the Wikinews project. Please feel free to edit the articles for spelling, style and clarity.
Editions so far: December 12, 2004 - February 26, 2005 - April 28, 2005.
It's hard to put into words all the fascinating things that have happened in the different editions since my last report from late February. In this issue: new editions, growth across languages, translations, papal veneration, contests, statistics, automation, original reporting, Wikinews TV, Wikinews Radio, chat, licensing, RSS, content partnerships, the French, and praise and criticism for Wikinews from blogs and established media.
I would like to use this introduction to give a big thanks to Datrio for writing an English language report from the Polish edition. While I'm trying to keep my reports international, it would be nice if more regulars from the other editions would write such summaries. I will try to feed the most interesting developments back into the next edition of this report.
- 1 New editions
- 2 Growth across languages
- 3 Original reporting
- 4 Copyright and translations
- 5 Neutrality and balance
- 6 Translations
- 7 Licensing of images and text
- 8 Contests
- 9 Building a free archive
- 10 Predicting the news
- 11 Partnerships
- 12 Wikinews radio and TV
- 13 Dynamic lists and RSS
- 14 Chat
- 15 Wikinews in the media
- 16 Alex says ..
- 17 The Wikinews community
Since we officially launched Wikinews in December 2004 in English and German, the number of Wikinews languages has grown to 12. The two latest additions are Ukrainian (launched on March 19) and Italian (April 1). These two languages are the first whose launch followed an improved process for new languages: contributors are asked to translate or create certain key pages before the edition is set up (an idea I stole from Memory Alpha).
This seems to work well to make sure that there really is an interested community before setting up a new language. Both Italian and Ukrainian are doing fine (see below), and I intend to use and refine this process for all coming editions. I don't think we need to slow down the process further: most editors from other editions I've talked to are happy about the speed at which their new editions have been set up, and that was before the new policy has gone into effect.
So, who's next? Judging from the support on m:Wikinews/Start a new edition, Serbian and Catalan have a pretty good shot at being created soon. Chinese has the required support, but is faced with the threat of censorship, and many contributors fear that a Chinese Wikinews would also endanger all other Wikimedia projects in China.
My personal view is that we should not make the availability of a Chinese Wikinews dependent on China's national policies. There are many Chinese speakers not residing within China's borders. More importantly, if the censorship is so far reaching that Wikipedia itself would be blocked because of another project, then I have little hope that a neutral encyclopedia can be created at all under such conditions. A community whose mindset is always focused on the threat of censorship is not a healthy community.
I believe Wikimedia should keep doing what it has been doing, encourage people to write neutrally about touchy subjects, and deal with censorship as it arises. One way to deal with it is to provide users with the tools (gateways and anonymizers) to keep accessing and editing our content. Activism beyond our mission? Not anymore than making printed copies available to people without computers, in my opinion.
What will actually happen? The vote on what we should do ended 50/50: half of the people who voted said the Chinese community should decide, the other half said the project should go ahead without special treatment. A tentative local vote on the Chinese Wikipedia didn't help either -- exactly half of the people there said the project should go ahead, and the other half was against it. I asked the Wikimedia Board of Trustees to comment, and Angela wryly responded:
- There's no consensus within the board either. In a discussion last month, Anthere said the Chinese community should decide, I said the whole community should decide, Tim Shell said it should be started, Jimbo said it shouldn't, and Michael Davis didn't respond to my email about it.
Jimbo Wales clarified his position, stating that he feels there should be greater consensus in the community before a Chinese edition is launched. This seems unlikely without a major change in China's censorship policy, so the project continues to be on hold. Fortunately, this is the exception, and most other languages should be set up smoothly.
Meanwhile, some speakers of the Scandinavian languages have been pushing for a united Scandinavian Wikinews edition. I don't know how feasible it is to handle different languages in one wiki or to maintain a shared portal, and therefore remain skeptical about the idea. I am certainly not going to make that decision by myself, so I think that those who are interested in it should try to get a vote within the Scandinavian community organized; hopefully, it will yield a clearer result than the Chinese one.
As an additional point of interest, it is worth noting that no other Wikimedia project has been as careful with setting up new language editions as Wikinews has. As of my last count, there were:
- 87 Wikiquote editions
- 121 Wikibooks editions
- 173 Wiktionary editions
- about 200 Wikipedia editions
Unfortunately, many of these are inactive or even attracting spam. I do not regret for a moment that Wikinews has adopted a more rigorous approach, as it has given us time to welcome new editions into the global community and led to a healthy competition of both quality and quantity between the editions (see more on this below). It also should be noted that Wikinews is different from other projects, in that an inactive frontpage makes the entire project look dead, while a reference work like Wikipedia does not have to be constantly edited to be useful.
Growth across languages
Let's start with the basics - the current article count across editions as of April 27, 2005:
- English: 1712 articles
- German: 1292 articles
- Polish: 494 articles
- Swedish: 303 articles
- Portuguese: 292 articles
- Spanish: 181 articles
- Romanian: 112 articles
- Italian: 111 articles
- Dutch: 86 articles
- French: 73 articles
- Ukrainian: 65 articles
- Bulgarian: 3 articles
Note that article counts by necessity includes index pages in the main namespace, though in most cases, that is only a slight distortion. This goes for all the article count statistics that follow.
Both Ukrainian and Italian show healthy growth, given that they are still very young. Bulgarian remains inactive and will likely be shut down if this continues to be the case. But we don't have to limit ourselves to a superficial count comparison, as the inimitable Erik Zachte has made his excellent collection of statistics known as Wikistats available for Wikinews.
The charts of the number of articles are certainly interesting, but they, too, only provide a piece of the narrative. In the context of a news site, and also when looking at such a young project, we have to play attention to what is happening right now. The table of new articles per day provides some insights which reflect my own perception of what's going on. Here's are the 5 Wikinews editions with the current highest rate of articles per day:
English is leading, which shouldn't be surprising. More interesting is the fact that the English story output keeps growing steadily, which is pretty much a test case for whether Wikinews as a whole can scale. After all, if we want to become a seriously competitive news source, we need to write at least about 3 to 4 times as much as we do now on a daily basis. The writing contest (see below) has certainly helped to boost our article output, but many more innovative ideas are needed.
Polish is currently more active than German, which is quite a surprise, given that German is the second largest Wikipedia language. I asked Sblive, a Polish user, how he felt about his edition and the project as a whole, and he responded: "I think that the Wikinews project is doing well at all, but some editions are still too poor to consider them as a good resource of information. [The] Polish edition, which is one of [the] biggest Wikinews editions, is developing into [a] good resource of news. One day it might become the leading press agency and news service in [the] Polish Internet." This optimism was shared by Datrio, who said: "We're aiming to become one of the main news sources in the world, and so far, it turns out pretty well."
The reaction from the Portuguese community to my inquiries was somewhat more cautious. Osvaldo Gago wrote that "we expected more editors", and Get_It, while remarking that the quality "is better than I could imagine", also said that the project needs more writers, ending his e-mail with a call for help: "We need more troops! Send backup :>" I'll see what I can do.
Nevertheless, Polish, Portuguese and Italian all prove through their steady growth that the number of speakers in a language is secondary: what matters is whether there is a motivated community. Even a handful of active editors can write enough stories every day for the site to be useful.
Curiously missing from the list is Swedish, which used to be one of the most active Wikinews editions. Unfortunately, the entire site has been pretty much frozen since April 2, 2005. I am told that the two main editors have moved on to other projects, and no one has taken up their work. We'll have to see if we can get some members of the Swedish Wikipedia excited about Wikinews.
On the positive side, the French edition is beginning to see some real growth. For a long time, it consisted only of brief news summaries in the style of Wikipedia's Current events directly put on the frontpage. I have begun to initiate a dialogue with the French Wikipedia community on the project. Some members of the community have long been opposed to the project, which, I believe, is due to the misunderstanding that Wikinews is meant to replace Wikipedia's Current events. I have tried to clarify that any such use of Wikinews contradicts its original mission statement.
Real articles are now being written -- some recent examples:
- Élection présidentielle contestée au Togo
- Silvio Berlusconi présente un nouveau gouvernement
- Jiří Paroubek est nommé Premier ministre de la République tchèque
I cannot judge the quality of the text, but the level of detail as well as the formatting seem appropriate. Some members of the English edition have also tried to help infusing the French edition with life -- notably CGorman, who started Google traduit Gmail dans douze nouvelles langues et demande à des volontaires de créer 144 nouvelles versions linguistiques and Zacarias Moussaoui admet la conspiration terroriste du onze septembre. I hope there will be more such cross-project collaborations soon.
The Romanian edition is using an approach I haven't seen before to attract new contributors: Its frontpage consists primarily of red links to stories that do not yet exist. While this will probably lead to lots of undesirable pages being created, it might encourage visitors to jump in and get their feet wet -- similar to Wikinews:Submit a story. Romihaitza commented that the Romanian edition "started with [a lot of enthusiasm] from many Romanian Wikipedia editors, but just 2-3 editors really worked [on] Wikinews pages to implement, maintain and create the 'newspaper'." This is an experience perhaps shared by some other small editions.
An interesting question arises: How many contributors does a Wikinews edition require in order to reach the critical mass beyond any initial enthusiasm where it can no longer easily fall back into a deep sleep? One obvious observation is that, when the work is shared among many shoulders, even if the individuals do less work, the risk of the site becoming inactive decreases. It therefore seems like a good recommendation to new editions to begin by growing the community and not just the number of articles.
I would say that English has probably reached the critical mass. Polish and German may be close to it. Portuguese and Spanish are presently reasonably active, but I'm slightly concerned that they may face a similar fate as Swedish if key contributors leave.
If you haven't seen enough statistics, here are some more yet:
- Romihaitza maintains a daily page with cross-project statistics which includes the count of new articles per edition. This could lead to some healthy competition (as is already happening between English and Polish).
- Wikinews:Awareness statistics compiles a number of key data points in regular intervals, such as the number of Google and Technorati hits for "Wikinews" and the Alexa traffic ranking. The latter has been somewhat erratic, but we seem to be heading in the right direction.
- Davodd is maintaining his own batch of stats at User:Davodd/Wikistats. It compares the growth of Wikinews with the other Wikimedia projects. As of his last update, we are leading in the rate of monthly article growth, and we come in third in user growth.
- CGorman is trying to model Wikinews' future growth. By his conservative estimate, we should see about 50 articles per day in the English edition in December -- let's hope he's right!
On March 13, I reported that the English edition had reached a milestone of 10 stories containing original reporting, and described them individually to illustrate what original reporting means in practice. Since then, about 35 additional stories in that category have been written. Describing all of them would be impractical, so I'll just pick some of the most interesting examples:
The story that started the now-famous chili finger saga which has reached its conclusion (so far) with Wendy’s finger pointer gets fingered for crime. Initiated by Ilya, DV quickly took over as our field reporter. This illustrates how Wikinews can stay on a case over a course of several weeks if at least one member of the community is interested. And before you judge the story as being non-notable, it has been on the frontpage of CNN.com and other major news outlets several times. The story also features prominently in our video broadcast pilot (see below).
This is an example of an original Wikinews investigation. Essentially, this article disputes the official reading of a report on the safety of the Central Artery Tunnel in Boston, Massachusetts, USA. Most news media blindly rely on press releases about reports and studies, and the authors of these press releases are fully aware of that and provide journalists with quotable soundbites. Wikinews doesn't face the same time and budget constraints and can dig a little deeper.
An especially nice recent addition by Pingswept, with lots of original photos. Ping followed it up with a funny and detailed interview with one of the runners. Combined, these articles serve as an excellent example of the quality of local coverage we can provide. The interview also shows that we are experimenting with different writing styles relatively freely.
Original reporting is not necessarily field reporting. In this case, a Wikinews reporter watched a webcast of a rocket launch and described the events as they unfolded. The resulting story, according to the author, was up before other news services carried it.
This is perhaps the archetype of the self-referential story, with photos and reporting by Pingswept. I think the story shows a reasonable level of neutral distance, though at the time of writing it lacks the Template:WikimediaMention which we normally use to indicate that a story refers to one of Wikimeda's projects. While Pingswept was taking notes, the talk was also transcribed live to IRC.
Spontaneous eyewitness reports are also a type of original reporting and will likely become more common as our community grows. The fact that the reporter in such cases will often not be a Wikinews regular opens up interesting questions of credibility and verification.
An example of a story that has transcended from the blogosphere (notably BoingBoing) to Wikinews, where much more detail and background is provided. While we have discussed the plausibility, we haven't actually verified whether the story is true.
This story is notable because of the level of detail it provides about the internationalization of the OpenOffice.org office suite into Swahili. As Pingswept observed, 55 million speakers just got the first office suite in their language, and almost no major media outlet reported about it. We did.
An interesting example of a Wikinewsie getting a quote from an expert, in this case, from an Assistant Professor of Economics at University of California, to add to a report. It will be interesting to watch whether Wikinews will follow the shared mythology of the established media about who is considered an expert and who is not.
Notable because we were the first to report that Encarta's new editing interface allowed free access to articles which were meant to be subscriber-only. We documented this using screenshots, but because Wikinews does not presently allow fair use images (more on this below), these had to be deleted.
As you can see on any of these articles, we encourage our reporters to document what parts of a story are based on original reporting, and what their sources are, on the discussion page or on a special "/Notes" subpage.
Looking again over all the examples of original reporting on the English Wikinews, I don't find a single story that caused us substantial problems. This may change, especially as we further explore the field of investigative journalism. But we are certainly not deluged with activists trying to insert propaganda.
Indeed, much of the reporting above demonstrates that Wikinews is something truly different from existing outlets of citizen publishing. It is certainly very different from the blogosphere: Wikinews articles try to eliminate subjectivity, blogs thrive on it. But the list of stories is also relatively free from the topics you would find at Indymedia: protests, arrests, and activism.
Currently, we do not require reporters to be accredited through some formal process (we do have the Wikinews:Accreditation requests page, but it is intended mostly for outside representation). This may change as the number of such stories grows.
There is a category of articles in German which are called "Originalbericht" ("original report"). What this means in practice differs from the English edition. People who want to submit Originalberichte are asked to request community accreditation at Wikinews:Akkreditierung first. This page defines such reports as articles where an individual is responsible for at least some of the information in an article, instead of a cited source.
This definition is not without its problems. I have found quite a few articles where the Originalbericht tag seems to mean nothing but "an article written for Wikinews without sources." Reporters are not required to document the work they have done, so the reader is often left in the dark as to what parts of the article are actually original in any way.
As of this writing, there are 40 articles in that category. Some examples:
- Solaris: Science-Fiction auf die Bühne gebracht - a report from a theater performance of Stanislaw Lem's Solaris
- 15. Deutsche Cheerleading Meisterschaft in Riesa - report from a cheerleading tournament, with original photos. There are lots of such sports-related articles.
- Dresdener ProfessorInnen reagieren aufgeschlossen auf Gasthörer und motivieren zu weiterführenden Studien - a brief report to the effect that professors at a Dresden university are open about allowing non-students to join their lectures. While original reporting, it comes across as somewhat overenthusiastic.
- Ayrton Senna würde 45! - a rather questionable article that is nothing but a tribute to racing driver Ayrton Senna, who would have turned 45 at the time the story was written (that is the only reason given for the story). This would be filed under Op-Ed in a regular newspaper.
- ThyssenKrupp verkauft EWK - article about the sale of Edelstahl Witten Krefeld by its mother company Thyssen Krupp. It is not clear to me what part of it is original reporting.
There are quite a few local reports, especially from Nuremberg and Dresden, but, as noted above, it is often difficult to see what the original reporting actually consists of. It's good to see that the German edition is experimenting with alternative models, and the effects of the accreditation process in particular deserve our attention. However, I believe that better documentation of original reporting would be desirable.
A category for original reports was recently created in the Polish edition, and it already contains more than 50 articles. It is difficult for me to make any comments on reports in a language I don't speak, so I have asked Datrio for his comments. He responded:
- I'll mention two of my stories - the first one is a sport report from the fight of a Polish boxer, Dariusz Michalczewski, with French Fabrice Tiozzo. Sadly, I couldn't participate in this event, but I was watching it on one of the Polish news stations, and reported every round of it.
- The second one, I wrote just a few hours ago - there was a Catholic mass for the memory of John Paul II - I participated in it, so I could write up some information from it - sadly, as I noted earlier, I didn't take any photos, as I don't own a camera. Many news sources mainly copied their stories from the Polish Press Agency (they have a contract with them), but we wrote our own report.
The two sories are: pl:Narodowa msza święta za Jana Pawła II and pl:Relacja z pojedynku Michalczewski - Tiozzo. Since I interviewed Datrio, the number of original reports has increased drastically. As I understand it, this is in large part due to contributions from Ariok, a sociology student from Lublin who also works for a radio station called Radio Centrum which has allowed him to post his stories to Wikinews.
Amgine interviewed Ariok on IRC, and Ariok described his motivation as follows: "I like to improve my journalism. In radio you need to have [a] special style of editing news, and pl.wikinews is the next skill: news editing on the Internet." As an example of cross-edition collaboration, Ariok has provided us with advice on our own radio efforts (see below).
Sblive also told me that the Polish community is "working on 'Serwisy regionalne' ('regional services') of our Wikinews, which will be based on original reporting." He also explained that Wikinews articles with speculative elements about future events (e.g. match-ups in a sports event) are considered original reporting in the Polish edition.
Portuguese has a policy and a category for original reporting. The category currently includes 2 articles, one of them about Wikinews, the other about an art exposition (with photos): Exposição de Luís Pinto no Instituto Português da Juventude em Faro.
Most or all other editions do not appear to have original reporting policies in place yet.
Copyright and translations
Copyright violations have so far not been much of an issue on the English edition. Wikinews:Submit a story often catches questionable articles, and the community of regulars is so small that we have a good idea who is doing what. As of this writing, only one English article is flagged with the copyright violation template.
My survey of the other editions has yielded largely similar results, but some editions had to deal with more frequent copyright violations. Ascánder of the Spanish edition wrote: "We have received a lot of copies of news taken from newspapers. There are also news copied from state owned biased reports. We mark them as copyvio and delete them after a couple of days."
An editor of the Polish edition told me about an eleven-year-old Wikireporter who had "copied some publications from services like Onet, Gazeta.pl etc.", and who did not stop until faced with the threat of a ban. Datrio, another Polish editor, explained that such stories are generally removed from the Main Page, and that an attempt is made to rewrite them to remove copyrighted information. "We deleted a few copied stories, though, as no one wanted to rewrite them."
Neutrality and balance
So far, the number of neutral point of view (NPOV) disputes on the English edition has remained at a reasonably low level. The overall articles in dispute (all types of objections, not just NPOV), at the time of writing, is 9. This includes a few tricky cases such as Argentinian workers preparing to defend control of factory, a very detailed article with many sources, where the objection is that it relies too much on left-wing sources and paints a one-sided picture.
Independent of this article, the issue of balance is one we have to carefully work out over the coming months. My personal conviction is that it is unfair to halt the publication of an article purely because of a perceived lack of balance, since such a lack can almost always be construed (some argue the New York Times is left-wing, others see it as part of the corporate media establishment). Claims have to be properly sourced and attributed, and we must take care not to make other people's opinions our own. If it can be demonstrated that specific information is missing, that is fine, but general complaints of one-sidedness are not helpful.
Generally, I think the right reaction with issues of balance is to fix them yourself instead of putting an article into dispute. Perhaps I will be able to articulate this view into a policy proposal. It is very important to me that dispute tags cannot be used as a sort of filibuster until an article is no longer relevant: We must be able to report on controversial issues that push people's buttons, whether it's left wing, right wing, or postmodernism.
Other editions report small POV problems here and there. Carlosar from the Portuguese edition writes: "Just a minor problem with Fim da pena de morte nos Estados Unidos, para delinquentes menores (U.S. Supreme Court: Death penalty for juveniles is unconstitutional). We solved the matter adding another point of view, no deletions occurred. The problem was easilly solved, everybody agreed, no complaints."
Romihaitza from the Romanian language said that the only problem so far was with the daily editorial, which some people opposed.
The Polish edition had some NPOV debates after the Pope's passing. As Datrio explains, "Christianity is the main religion in Poland, and we were calling the pope a saint man, the biggest person in Poland, etc." Some even wanted to change the colors of the site as a sign of mourning. Sblive also reported that there was a dispute about whether the Polish Wikinews should cooperate with Indymedia, and that such a cooperation would probably not be possible because of Indymedia's strong POV.
I also asked editors from the different editions about the importance of translations. The answer was unanimous: translations are very important, especially from the English version. This is different from Wikipedia, where people often prefer to write their own articles from scratch. Evidently, many Wikinews writers feel that the effort that has gone into a story should not be needlessly duplicated.
There is, of course, a specially strong relationship between the Spanish and the Portuguese editions. Julián Ortega Martínez of the Spanish edition writes about the sources used: "Mostly English and Portuguese Wikinews. Lately, with the Pope's death, Italian Wikinews is used too. Currently, Carlosar is an active Portuguese-Spanish translator, and even started a campaign so Spanish articles can be translated to Poruguese."
Ascánder, also from the Spanish edition, concurred and estimated the number of translated articles to be about 60% (this is from early April). Osvaldo Gago from the Portuguese edition also states that translations are "very important, most articles of the Portuguese edition are translated." Carlosar concluded: "Other editions save a lot of work, we can write great articles by extending other editions' articles."
Licensing of images and text
Wikinews content is still in the public domain. I have already summed up the licensing discussions in my last report. One important thing that has changed since then: Several smart people have told us that it is very difficult to legally put works in the public domain. The Creative Commons public domain deed is a rather complex procedure for that reason, and CC head honcho Larry Lessig argues that click-through agreements are not legally sufficient to achieve the same effect.
My own stance is that, while we should reach a decision soon, in practical terms, I doubt that our current practice will lead to difficulties anytime soon. Someone who unknowingly put their submission in the PD and wants to revoke it would find themselves in a similar situation to someone unknowingly putting it under CC-BY, FDL, or any other license; normally we would try to accommodate such people rather than going to court.
Instead of quickly switching to another license, we should carefully evaluate what options we have, or we might wind up in an FDL-like situation (though, truthfully, due to the nature of chronological publishing, it is easier for us to switch licensing than it is for Wikipedia). The Creative Commons Attribution (CC-BY) license is most frequently mentioned as an alternative. However, if you actually read the license you will find that attribution is to be given to the "Original Author", which contradicts the wiki way of doing things (egoless writing). Creative Commons to the rescue: there is actually a draft wiki license that makes it possible to attribute edits to the wiki as a whole. This approach would probably be ideal, but the license currently only exists in a copyleft variant, and is still in beta.
Perhaps we can work with the Creative Commons people to find a licensing option specifically for Wikinews, or maybe we can license our content in such a way that the Wikimedia Foundation is allowed to change to another free content license. In any case, this is not something I'm losing a lot of sleep over.
More urgent, in my opinion, is a resolution to the fair use issue, and I will give this another push soon if necessary. Simply put, none of the Wikinews editions currently allows fair use of images and other content, for the reason that we believe that we have to be very cautious about other people's copyrights, especially photos from other news sources. But we do need exceptions for screenshots, logos and perhaps publicity photos explicitly meant for press use -- the Encarta case in particular, where we had to remove useful screenshots, demonstrates this.
We will continue to use the Wikimedia Commons as our main image repository. This will always be a bit cumbersome until Wikimedia gets single login, but Ilya has given us a new application for Windows called Wikimedia Commonplace that makes it possible to just drag and drop the files you want to upload. I hope someone will write a Linux port or work-alike soon.
On March 21, I started the first Wikinews writing contest here on the English edition. The rules are simple: You have to write one story a day, or you drop out. (You can join in later, but you have to work off an accumulated article debt.) Quite a few Wikinewsies have contributed prizes, and the contest, epsecially initially, has been very successful at boosting the number of articles. The harsh drop-out rules of course meant that the flow of articles generated by the contest ebbed relatively quickly (3 contestants are left at the present time). By my count, 218 stories have been written as part of the contest so far, so I think it was very much worth it.
Most people from the other editions I talked to feel that their communities are still too small for a similar contest. I believe, however, that a fully international version of the same contest idea could work quite well. Every edition would have to nominate at least one judge, and it wouldn't matter if there's only one other active user -- they would compete with Wikinews writers from all other languages. I'd like to tweak the ruleset a bit before doing this, though, to keep the activity level high throughout the full duration of the contest.
Meanwhile, the Wikinews design contest is still running, and has seen a few submissions, notably a complete skin re-design by Noclip. Nobody has made any attempt to redesign the portal at http://www.wikinews.org/ yet, though -- I think an attractive and dynamic entry page would be a good idea, but my own design skills are too limited for this.
While the main writing contest is winding down, Ilya has prepared his original reporting contest. It is important to note that this is not just a general contest about original reporting, but specifically about getting quotes -- a very important way to enrich an article without doing a lot of additional work. I'm very excited about this and hope that I will find the time to make a contribution or two myself.
Building a free archive
"Decoded at last: the 'classical holy grail' that may rewrite the history of the world." This is a headline from the Independent, dated April 17, 2005. It is interesting for one reason: If you want to get more than the headline and a brief abstract, you have to pay one British pound.
I stumbled across this article as I wanted to read some more background on our brief story Famed classical works rediscovered. It struck me then that this was an excellent example of a point I tried to make earlier on the mailing list: Even if a Wikinews edition is inactive for a while, it continues to have value as a free archive of news content.
When is such an archive useful? When you want to look something up, when you want to view multiple sources on a topic, when you are conducting historical research, or when you want to do any kind of quantiative or qualitative analysis of news content. In short, for any use that goes beyond reading a story the moment it appears, an archive is essential. £1 per story may not sound much, but it can quickly grow into £10 or £100 if you have to download multiple stories to find the one you are looking for, or if your goal is to do larger comparisons and searches. As the English edition is increasingly likely to have at least all major global events covered, I believe it will quickly grow into one of the most valuable free news archives on the web.
It turned out that 3 days after our original story, someone started a second, much more detailed article on the same topic: Infrared technology enables recovery of lost classical writings. Wikinews is not bound by the same cycle of publication as traditional news media. No editor will shout at you if you cover a topic we've already written about a few days ago from a different angle. We don't just work for free, we're free to work on whatever we want, and we can afford to spend our resources staying on stories which are already yesterday's news for our competitors.
Predicting the news
A recently introduced process on the English edition is the story preparation page. As you might guess, we use it to work on stories before the events they describe have actually come to pass. The article German Cardinal Ratzinger elected Pope Benedict XVI was prepared in this fashion, and the Results of 2005 United Kingdom General Election are currently being worked on. In the case of the papal election, this has allowed us to have a story ready in a very timely fashion, even before quite a few mainstream news outlets: not bad for one of the most significant events of the year.
The strategy of preparing stories -- days or even years in advance -- is of course common fare in all news organizations. What could eventually be different in Wikinews is the scale of this effort. Imagine a Wikinews that is 100 times the size it is today: Wikinews writers would compete to come up with the best predictions of what might happen. Instead of going to the Foresight Exchange, you could take a look at what the Wikinews community is preparing for. It is no coincidence that the German equivalent of Wikinews:Story preparation is called Glaskugel -- crystal ball.
I believe that content partnerships with other news sites and blogs are essential for our growth. Thanks to our free content model, any single Wikinews member can go out and try to get others to use our stories, without asking for permission. If other people use our stories, this helps us by driving people to our site through backlinks, by spreading the Wikinews name, and by improving our Google rank. I've started a page called Wikinews:Story propagation to assist with the effort, and I encourage other Wikinews editions to do the same.
I've initiated two partnerships so far. The first is with SciScoop, a science news site, and has so far mostly consisted of us submitting stories there. However, the new owner of SciScoop has stated that he wants to explore the use of a free content license for SciScoop's own stories.
The second partnership is with The World Forum, a general news and discussion website. The World Forum has used many of our stories, and unlike Wikinews itself, is indexed by Google News (which has led to Wikinews content being on the frontpage of Google News several times). Drog, founder of TWF, has even made some contributions here on the site. For a time, he prominently featured the Wikinews logo on his website, but after some brouhaha over permissions, this had to be removed, at least temporarily. Hopefully, the Wikimedia Board will officially allow Drog to use the logo to link to us.
I'm also trying here and there to convince authors of quality content to put their material under free licenses. Steven Aftergood, author of the Federation of American Scientists' Secrecy News newsletter, has privately agreed to put his articles under CC-BY-SA, though for technical reasons hasn't actually added the licensing information to the newsletter yet, so I'm somewhat reluctant to use his articles (which are good, but subjectively colored). When you see an interesting blog or newsletter, please do try to get the author to switch to a free content license, preferably CC-BY.
At some point, I'd like to push a major effort to actually catalog high quality news sources which aren't widely known. My problem with blogs is that the content that rises to the top is not necessarily good, it is emotional -- personality tests, Flash animations, funny stories, outrageous ones, etc. There's nothing wrong with that, but as a news source, our responsibility is to report a story if it is significant, even if it doesn't immediately get your blood pumping. I have a feeling that there are lots of people quietly working away on good stories we don't know about. Eventually both the people and their stories will find their way to us, but we should try to find them first.
Wikinews radio and TV
The transition from text to other media is natural for news. Nevertheless, I find it quite amazing that real work on this is actively being done by our volunteers. Certainly the most impressive development is our very own Wikinews video broadcast, an effort led by David Vasquez. DV is a video specialist, and he has single-handedly created a virtual studio, a theme song, several scripts, various animations, and a few on-the-scene reports (on the chili finger case described above). That he is even able to do all this on a shoestring budget should make TV executives really, really worried.
You will find links to the first minute of the Wikinews video pilot on Wikinews:Broadcast. A lot of discussions have surrounded the question what politically correct (i.e. not patent-encumbered) video format to use. Ogg Theora is available, but not very wide-spread, so we also offer MPEG4 and Windows Media as alternatives. Currently, the video files are hosted externally, but I would very much like us to use the Wikimedia Commons for this.
Based on his work so far, I have no doubt that DV will succeed in creating a first impressive pilot episode. The challenge for us is to then mobilize a larger community to participate, as the Wikinews broadcast cannot succeed if it is not a collaborative effort (even if it's just collaboration on the scripts).
I'm sure there are a lot of video journalism grants that are just waiting to be picked up by the Wikimedia Foundation -- if we manage to get one of those, we could afford to send our most active volunteers video cameras, who would use them to produce reports and send them to David (or to a yet to be created virtual clearinghouse) for processing. The most interesting question is whether we can get citizen journalists into areas with little Internet coverage, and into crisis regions (the two will often be identical). If we can, will we also get their reports in time?
I am cautiously optimistic. The phase after the release of the pilot will be the most critical. Let us know if you have ideas how to get more people involved in this project.
Our radio efforts are also making progress. There is already a Wikinews online radio broadcast in Polish, twice a week, on a station called Radio In-Nocte. In the meantime, Amgine and Ilya have led the effort to produce an online radio broadcast of the English Wikinews edition. The project is coordinated on Wikinews:Audio Wikinews, and Amgine has created a few example broadcasts in the Ogg Vorbis audio format. These can be found on Commons:Category:Wikinews.
If you're looking for a fun MediaWiki task, I have submitted bug 1954, a proposal for a <playlist> tag that could be used to embed audio playlists into wiki pages. This, I think, would help a great deal in applying the wiki model to the audio file compilation process.
Dynamic lists and RSS
While our Main Page has become more dynamic, with three lead stories chosen by the community, it still essentially consists of a long list of uncategorized articles. In my last report, I mentioned a feature for building automatic index pages which has come to be referred to as "DPL" (DynamicPageList). This feature has now been taken live, in spite of scalability concerns, on Wikinews only. It is currently in use on index pages like South America. You specify the categories you want, the number of stories, and it returns the latest headlines from those categories. An example:
<DynamicPageList> category=South America category=Politics and conflicts count=5 </DynamicPageList>
This will generate:
- Major Colombian newspaper reveals Bogotá mayor does not hold claimed PhD degree
- Former President of Brazil, Luiz Inácio Lula da Silva, detained as part of corruption investigations
- Brazilian President arrives in Israel to start Middle East tour
- Police fight police in Brazil protest
- France's Sarkozy to meet with Brazilian president da Silva
The list is sorted by the date the stories have last been modified. This is suboptimal, as changes to the metadata (categories, interlanguage links) mean that a story is pushed to the top of the list again. The DPL also does not interact well with the cache -- pages using it have to be manually purged (using URLs like this) within regular intervals. Yet, it is clearly an improvement over manually maintaining hundreds of index pages, which is simply not feasible and the main reason most of the action currently happens on the Main Page.
In addition to automation, RSS feeds (for offline readers, syndication and aggregation) continue to be one of our most needed features. Dan100 is still maintaining his manually generated feed of headlines. Several people on the English edition have experimented with alternative approaches, but none of them has reached maturity yet. There is also a German and a Polish feed. The Polish one is generated using a script and is likely to break from time to time. I'm not aware of any other feeds that are currently in use.
But every cloud has its silver lining. One of our most talented MediaWiki hackers, Gabriel Wicke, has been contracted by Kennisnet, a Dutch educational non-profit organization, to implement RSS support for categories. This feature is currently in beta-testing (example feed). The time the category was added to the article is used for sorting the stories.
Gabriel has also written an extension to embed an arbitrary RSS feed into a page. Combined, these features could not only replace DPL, but also take Wikinews to the next level by providing users with the ability to subscribe to any news category they are interested in (combination of different categories would be a big boon).
The way this functionality has come into being, through funding from a third party, underscores the point I tried to make in my last report: We (as in, the Wikimedia Foundation) need to prioritize development tasks according to the individual project needs, and hire people to get things done. We shouldn't have to depend on an outside party without direct involvement in Wikinews to fix our software problems for us.
When I started the first Wikimedia Internet Relay Chat (IRC) channel, #wikipedia, in December 2001, you could be happy to find more than 2 or 3 people there in the first few months -- and usually those people weren't talking. The channel really took off, ironically, as Wikipedia went down, as the various "Wikipedia is down" pages pointed to the IRC channel. People met there to talk about their withdrawal symptoms, or to nag the developers.
Today, IRC is one of Wikimedia's most important communication media. All members of the Board are on IRC almost daily, primarily on #wikimedia. There are currently more than 60 IRC channels related to Wikimedia projects, including several channels that provide real-time feeds of the recent changes on a wiki.
It doesn't take much to figure out that a real-time medium like IRC can be especially useful for working together quickly and efficiently to report the events happening around us. Hence, one of my first priorities when setting up the first Wikinews pages was to promote the Wikinews IRC channel, #wikinews (see the page on Meta about the channel for a list of regulars).
We used the IRC channel very early, on February 5, to organize a meeting between bloggers and Wikinewsies, during which it became clear that there is a need for a global channel to coordinate the various citizen journalism efforts. Failing that, it might be good to try to do the best job possible to coordinate our own global citizen journalism effort. So, if you haven't already, please do pay us a visit soon and get to know the community that is there.
One problem with IRC is that it usually requires client-side software like XChat to be installed. Many people on the Windows platform have been burned by spyware and are very wary of using any service that requires installing new programs. Ilya Haykinson has set up a web interface to the IRC channel which should help to address this. Indeed, since its creation, we have seen a large influx of new IRC users.
Another change that has been made is the split of the #wikinews channel into the international channel (#wikinews) and the English language channel (#wikinews-en). If your topic only concerns the English edition, then you should try to discuss it on #wikinews-en. Hopefully, this change will stick. It is almost impossible to organize channel splits like this when the community has grown beyond a certain point: people won't move. It is very important to me that people from other editions feel at home in the international channel.
In practical terms, English is still used as a bridging language on #wikinews, but we have seen much more activity from other editions since the split, particularly from Polish. No matter what edition you edit on, it would be nice if you could join us at least once a day. And if you are one of the people welcoming new users on your wiki, please point them to m:Wikinews/IRC. If your edition doesn't have an IRC channel yet, make one! Feel free to ask on #wikinews, and someone will amost certainly take you through the process.
So, what can we do with this tool in practical terms? An example is the story MSN Encarta introduces wiki-like enhancements. I tried to get an interview with someone working on the project. While I never got a response to my inquiry, IRC was very useful in finding people and phone numbers. It turned out that one particular Wikinewsie even had the number of Microsoft's internal phone system.
This is just a vague hint at what has been described as the power of many. Imagine a global network of hundreds of people available for your questions at any given time. It is almost certain that you will find someone with the qualifications to help you in such a network. IRC enables this communication and knowledge exchange network to grow. It may in fact be a good idea to limit this network to Wikinews in order to keep the signal to noise ratio high.
Wikinews in the media
Wikinews articles are increasingly being cited in the blogosphere -- Technorati gives some current results. That is one of the beautiful thing about blogs: it's very easy for any provider of news to grow popular very quickly by utilizing the power of many, regardless of what recipe is used to create the news. Ideally, in the case of Wikinews, increased popularity should also directly lead to more editors.
Wikinews and blogs are not mutually exclusive -- many Wikinewsies are bloggers:
Perhaps, as our software improves, we will be able to use our user pages for subjective blogging about Wikinews, but for the time being, all these blogs are hosted on external servers.
Simon Waldman of Guardian Newspapers, whom you met in my last report, has found some nice words for our story about the UK budget recently. He begins his entry with "While many of my fundamental reservations about Wikinews remain", which is a phrase I expect to hear a lot more often soon. His original post promises a response to all the rebuttals he received "in a week - or so". Don't hold your breath: that was two months ago.
Meanwhile, Joanna Glasner of Wired News has written a longer article about Wikinews, published on April 22. Curiously, the one paragraph that I've seen quoted most frequently is this one:
- Nearly six months into an experiment to apply the collaborative, information-gathering model known as a Wiki to the deadline-driven field of breaking news, operators of Wikinews are finding their mission rife with frustrations and challenges.
Personally, I am actually surprised how well things are going. As you may recall, I was initially in favor of a much more rigid review process than the informal one we are using at the moment, for fear of libelous stories, copyright violations, and POV. It turns out that these problems have been relatively infrequent, and our main challenge has been to produce a steady flow of news stories and to get enough people motivated to participate in original reporting. Nevertheless, the article makes some good points. It quotes Wikimedia president Jimmy Wales as saying:
- "We've got five (stories), and two of them are about Romania. Presumably that's not the norm. It still is following the interests of the participants pretty strongly."
I always predicted that this would happen in the first months of the project. To quote something I said in the very early discussions about Wikinews:
- You are correct that there will always be systemic bias -- not covering certain topics or giving too much information about others -- but my attitude here is somewhat more relaxed. We have the tool of visibility to reduce the impact of this bias. Even if we have 90% Linux stories and 10% politics stories, we can cover the Linux stories on the "back pages" of Wikinews, and give the politics their fair share of exposure.
This is exactly what we intend to do with our topical index pages once our automation tools to build these pages are sufficiently mature. Of course, at the moment, even on the English edition, we can still afford to put all our stories right on the frontpage. When the number of articles per day reaches 30 or so, that will change.
Another instrument against bias are collaboration initiatives like CGorman's Wikinews:Country of the Week. They can help to focus the energy of the community on worthy topics, similar to the Collaborations of the Week on the English Wikipedia.
Alex says ..
A clever person quoted in the Wired story is Alex Halavais, director for the informatics school at the University of New York at Buffalo. He has elaborated in a blog entry on where he sees problems with the Wikinews model. While Alex' post is very favorable to our project, I must take issue with his comparison of Wikinews to Google News and clipping services:
- It’s a bit like a (relatively) clear news filter—the human version of Google News. Many of the stories are derived directly from the mainstream media. It is certainly not plagiarized—all of the news is clearly sourced—but it does have the feeling of a clipping service.
The first and most important mistake with that comparison is that everything we are creating here is original, free content. You can take Wikinews stories and do with them whatever you wish, and we are building a unique long-term free archive of news stories. That is not true for Google News or clipping services: the information remains proprietary, the source archives get locked down after a grace period; if you copy the content without permission, you can end up in jail.
The second mistake is that we are providing a service beyond collecting information content from various sources and putting it into our own words. We are synthesizing information content to give the reader a bigger picture: more information, less bias. Just one example: Chinese rioters storm Japanese embassy in Beijing (April 9). This article cites 11 sources, including BBC, Reuters, Xinhua, AP, Bloomberg, The Japan Times, Kyodo News, and The New York Times. Based on these sources, the authors have created a coherent single article (which, I must stress again, is original and free content and not copied and pasted).
To compare this creative act of synthesizing information content from multiple sources into an original expression with a clipping service or with Google's newsbot is very unfair to the hard-working Wikinews community. Ask yourself, as a reader, what is more valuable to you: a list of 500 search results on the Chinese riots, or a single story that combines the key facts from all of them?
While I agree with Alex' argument (which he also made to Wired) that Wikinews should add as much original reporting as reasonably possible, I am very much convinced that, even if the majority of Wikinews stories continues to rely exclusively on outside sources, what we are doing is extremely valuable and important, and something which in no way can be accomplished by a computer program that isn't sentient.
On Wikipedia and Wikinews, Alex writes:
- First, notice how many people turned to Wikipedia for news about the new pope. There is some space there between the depth of Wikipedia and the place of Wikinews.
The depth of Wikipedia is relative. As an example, take a look at Bush EPA nominee abandons insecticide-on-children study after Senate hearing. This is a very detailed Wikinews article about what some critics see as a study that would have subjected children to insecticide exposure. At the time we wrote the story, Wikipedia knew nothing about the controversy. About 8 days later, there's now a Wikipedia article called Children's Environmental Exposure Research Study, which, as of now, is 4 paragraphs long and lacks sources.
Another example is our article Japanese commuter train derails, apartment building smashed. The Wikipedia article, as it was when we published the story, had two paragraphs and 3 sources. The Wikinews story has more sources, is more detailed and includes information about the region where the accident happened.
That is of course not a criticism of Wikipedia at all -- in fact, I doubt any other encyclopedia will ever have an article about either event, so the fact that Wikipedia has independently created these pages underscores its unique nature. But it also shows that for events which are not of global significance, for everyday news, Wikinews is the place to go. When it comes to global events, I think Wikipedia will always excel at giving a comprehensive overview in a short amount of time, while Wikinews should be better at reporting the latest developments as they come in, no matter how small. The space which Alex seeks to define therefore seems very obvious to me: news.
Alex' other points are well taken: a collaboration between Flickr and the Wikimedia Commons is long overdue, and the idea of using cameraphones for reporting purposes is certainly one we should explore. I also strongly agree with Alex that we should work with journalism schools, but also with local media that are already doing original reporting on a daily absis. There's a lot more potential for collaboration there.
The Wikinews community
The English version of Wikinews has now been around for almost half a year, and social patterns are beginning to emerge. Let me point out the obvious once more: Wikinews is different from Wikipedia because you cannot just write a "news stub" and hope someone else will turn it into a decent article in time; you have to do a large part of the work to get the story up to snuff if you want it to end up on the frontpage.
This practical requirement acts as a social filter. The pressure of having to have a presentable frontpage with fresh news at any given time has the same effect. Wikinews requires motivation, dedication and passion more than perhaps any other Wikimedia project. It is doubtful whether it could have taken off so quickly without the massive media exposure and community attention it has received by being a Wikimedia sister project, since you need a large pool of people to get that special kind of person who is willing to put in an afternoon, or a whole day, to go out and write their own report, or to even write summaries of stories on the wires.
I would therefore contend that the personal investment into the project of the average Wikinews regular tends to be higher than on other projects. Yet, while the project is in some ways less collaborative than, say, Wikipedia, it is also more collaborative, insofar as the entire community needs to work together to decide what the most important stories of the day are, which stories are ready to be published and which ones aren't, and so forth. There is not (yet) the equivalent of the Wikipedia article which goes unnoticed by anyone for months: You can't publish a Wikinews story without getting in touch with the larger Wikinews community.
We are working hard to develop a truly global network instead of keeping the language editions isolated. This, too, should help to keep a check on our neutrality, and to make sure that bad policies are quickly reformed, and that good ones are quickly copied. We share a keen awareness of the fact that Wikinews is still young and fragile, and that we can't afford to get bogged down in petty fights. We are united by a code of ethics, a belief that as citizen journalists, we have as much a responsibility to the public as paid professionals do.
For all these combined reasons, I believe we have done quite well so far to avoid conflicts and edit wars. For these reasons, I am optimistic that we can build a community unlike any other.
Sure, there will be fights: that is inevitable. When they happen, I appeal to you to try to resolve conflicts privately, and to forgive and forget. If you're not involved, try to mediate by privately introducing yourself and offering your help. Don't take things public because of ego. If there's one Wikipedia feature we don't want to copy, it's pages full of public disputes and personal attacks that stick out like sore thumbs. Things that happen between two people are easy to forget. Things that happen in front of a perceived audience of hundreds will stay with you. If we follow this principle, it is also easy to recognize the kind of person who wants to disrupt the community, because they will never accept it.
I believe in this community. It has proven time and again that it has the ability to reinvent itself and its processes when necessary. This doesn't mean that we're safely on the road to success. But if Wikinews does succeed, it will be because of the community, and if it fails, it will almost certainly not be the community's fault.
Stay with us, or, if you aren't with us yet, join us. Interesting times are ahead.