Wikipedia:Wikipedia Signpost/2005-01-10/SPV
Single-Page View Archives |
---|
| ||
Volume 1, Issue 1 | 10 January 2005 | About the Signpost |
| ||
(← Prev) | 2005 archives | (Next →) |
|
| |
Home | Archives | Newsroom | Tip Line | Shortcut : WP:POST/A |
|
From the editor
Welcome to the inaugural edition of The Wikipedia Signpost! I hope this will be a worthwhile source of news for people interested in what is happening around the Wikipedia community. I plan to publish it on a weekly basis, every Monday.
The name, The Wikipedia Signpost, was chosen to be like the name of a newspaper, since a newspaper is what I would call this project. Though it will almost certainly never appear on newsprint paper, it will nevertheless take on this role for our community. It should have some resemblance to the other newspapers you may happen to read in the course of your life (which I venture to guess many of you read online anyway, rather than the paper copy).
With no slight intended to other projects or languages, The Signpost will focus strongly on the English Wikipedia. The news coming from other language Wikipedias could undoubtedly fill its own newspaper. Correspondingly, rather than trying to translate The Signpost for the benefit of other segments of the community, I hope that as other projects develop a need for this kind of resource, they will adapt and develop this idea for their own uses.
The name also especially suits Wikipedia, because it alludes to a practice here and on other wikis, in that we communicate primarily through "signed posts," as on talk pages. While the wiki system may be used to develop and publish articles, because this is original reporting the reporters will use a byline to "sign" their posts. Since this is not in the article namespace, guidelines such as "no ownership of articles", and particularly "no original research", will not necessarily apply. However, The Signpost will strive to maintain its objectivity as would be appropriate for an independent media organization elsewhere.
The need for a community newspaper is tremendous. Already long ago the speed of recent changes on Wikipedia surpassed anyone's ability to follow edits thoroughly. By now, we are well past the stage where, even when considered in broader terms, anyone can singlehandedly stay on top of events here. To attempt this, your watchlist would be unmanageable, your inbox inundated with mailing list posts, your browser overwhelmed with open tabs, and your time spent flitting from reading about Wikipedia in the blogosphere to hanging out on IRC—even when asleep! I hope The Signpost can spare people the effort of trying to be everywhere and read every discussion.
The subjects covered here should be whatever community subjects interest the readers. Some people will be more interested in things happening with featured articles, others will want to follow Wikipedia's statistical trends. Not everyone will share all interests, but I hope to have something for everyone, and to hear from readers what else they want to know. And to those who might call this navel-gazing, I merely ask—so why are you reading then?
Finally, given the size of Wikipedia that makes it necessary, even a small community newspaper is a huge task. I don't plan to do it alone, and anyone interested in writing for The Signpost should get in touch with me so we can organize the work. I especially welcome anyone who's been dying to try their hand at original reporting, but isn't really sure whether they have material worth publishing on Wikinews (and no, this project is not meant to make navel-gazing publishable at Wikinews). And really, what more logical place is there to develop our skills with original reporting than here, where news is being made and people are interested in this news?
With that, I wish you all happy reading!
Wikipedia website changed to multilingual portal
After more than two years of sporadic discussion, the www.wikipedia.org domain changed Friday from a redirect to the English Wikipedia into a portal showing all Wikipedia languages having over 100 articles.
When it was created, the English Wikipedia was located at www.wikipedia.org (originally www.wikipedia.com). Other languages were assigned domains based on ISO 639 language codes. The English Wikipedia was eventually moved in October 2002 to the appropriate domain for its language code, en.wikipedia.org, while the previous domains were set to redirect people to the correct location.
The move to create a portal was endorsed by Jimmy Wales and acclaimed by Wikipedians from other languages as long overdue. However, the change surprised a number of people and upset some users of the English Wikipedia who disliked the new arrangement.
Reopening of discussion leads to action
Prompted by a question on the Wikitech mailing list, Member Representative Trustee Anthere took advantage of the opportunity to revisit the issue. Modelled on the practice applied to matters like the Swiss domain at www.wikipedia.ch, Anthere commented, "I think portal is the solution chosen for all our projects, so best to stick with it probably." She suggested that a design contest be held for the new portal.
Wales said he thought it was "a fantastic idea", but added, "It doesn't make sense for us to accept an ugly page." He encouraged people to use the wiki process to develop an attractive portal.
Acting on these suggestions, developer Tim Starling put up the new page and set up a www.wikipedia.org portal for people to edit on Meta. The content, originally based on the "Other languages" section of the English Wikipedia's Main Page, experienced a number of edits, although people were still struggling with the problem of using English words on a theoretically multilingual page.
English Wikipedia response
Reactions on the English Wikipedia were not quite as glowing. Ta bu shi da yu criticized the lack of interesting and eye-catching features on a page frequented by many new visitors to Wikipedia, saying that at the least, "I hope that someone will either fix the design or make it more interesting."
The problem was aggravated by the fact that discussion of the change took place entirely on the Wikitech and Wikipedia mailing lists, as nobody took the time to specifically alert the English Wikipedia about the matter, even though its users were the ones most likely to be affected (Wikipedians in other languages having already learned long ago to bookmark their main page correctly). Raul654 complained that the change was "totally unannounced", pointing out that the English Wikipedia wasn't notified either before or after the change. He added that he thought the conversion to a portal would cause the English Wikipedia to miss out on potential new contributors.
In response to the complaints, Anthere pointed out that links in the form of www.wikipedia.org/Foo—which are therefore intended for specific Wikipedia articles—still redirect to the corresponding English article. She added that in her experience, even many non-English journalists give Wikipedia's address as www.wikipedia.org, and that this has sometimes prevented speakers of other languages from discovering that Wikipedia existed in those languages as well.
Still, in spite of the complaints, development of the new portal was moving forward in typical wiki fashion. Thus far, the changes have amounted to tweaking the presentation of the languages, but a successful redesign has yet to be presented. Ropers did try to give the portal some more visually appealing content by adding flags, but this was rejected because of the problem that country flags do not correspond very exactly with languages. So the challenge of coming up with an aesthetically pleasing portal remains.
Broadsheets and blogs debate over Wikipedia
The fallout from Larry Sanger's New Year's Eve Kuro5hin article criticizing Wikipedia for its "anti-elitism" trickled into the media last week, which took up the ongoing debate over Wikipedia credibility. Simultaneously, discussions on the Many-to-Many blog over this issue stimulated internal debate there as well as informing the mainstream press.
Media topics changing
As the news cycle begins to shift away from coverage of the 2004 Indian Ocean earthquake (including widespread discussion of Wikipedia's article on the subject), media attention on Wikipedia has also shifted, and two publications carried extensive articles that return to the subject of weighing Wikipedia's advantages against credibility concerns. In the Financial Times, John Gapper wrote an article called "A new entrant to the knowledge market[permanent dead link ]", which, as suggested by its title, introduces Wikipedia as a new competitor to Encarta and the Encyclopædia Britannica. Meanwhile, The Observer published John Naughton's column entitled "Why encyclopaedic row speaks volumes about the old guard", dealing with the reactions of established reference works like Britannica to Wikipedia's success.
Although it provides an introduction to Wikipedia, Gapper's article is not merely a glowing review. Not surprisingly, given his audience, Gapper presents the competition between encyclopedias in economic terms. In his analysis, "The combination of the internet and Wiki software is a disruptive technology [disruptive to existing competitors, that is]: it allows Wikipedia's enthusiasts to produce something that is extremely cheap (free, in fact) and good enough to satisfy many people."
In his overview, Gapper presents a condensed version of the debate between Wikipedia's founders, Jimmy Wales and Sanger, over the weight, or lack thereof, given to academics editing in their field of expertise. (Kuro5hin founder Rusty Foster's observation on this debate: "Public bickering is the lifeblood of internet-based projects.")
In essence, Gapper assumes the argument of Sanger and others, that Wikipedia does in fact need to "bridge the gap" with the traditional world of academic expertise and peer review. At the same time, he comments on the trickiness of doing this without losing the productive nature of the Wikipedia community.
Gapper also notes that in spite of the Neutral point of view policy, Wikipedia does have biases in its material resulting from the views of its "core group". His example is the McDonald's article, which he says "devotes a lot of space to criticisms by environmental and social activists but says little about how the company originally succeeded." (While there has been renewed activity on the article since Gapper's comment was first published on the web, his point hasn't really been addressed yet.)
Still, Gapper concludes that Wikipedia is "surprisingly good", and says that if used carefully, "it provides a lot of information on a wide variety of topics." If the Wikipedia community is able to compromise with academia, he suggests, it could seriously threaten the future of existing published reference works.
Observer tracks recent media discussion
Naughton, while he does not mention Sanger or Wales directly (perhaps preferring not to undermine his portrayal of Wikipedia in terms focusing on group dynamics), does discuss the article by former Encyclopædia Britannica editor Robert McHenry that was probably the most extensive published critique of Wikipedia before Sanger's article. Naughton also quotes directly from comments by Clay Shirky (see below) that were written specifically in response to the Sanger article.
In addition to analyzing the overall debate over the merits of Wikipedia, Naughton can be seen tracing some of Wikipedia's recent press coverage elsewhere. Besides the McHenry article and other previously published quotes from Britannica editors, he mentions the debates over and protection of the George W. Bush article, discussed in several media outlets, and also recaps his own initial article about Wikipedia. He also mentions "a well-known crackpot" writing an article about himself, apparently referring to Sollog.
Naughton consistently evaluates Wikipedia's product as being of good quality, calling even these disputed articles "informative" and "admirably detached". Dismissing the critiques of Wikipedia as having a "whiff of hysteria", he sums up Wikipedia as "an exceedingly useful online reference work often consulted by this columnist and countless others."
Going one-on-one at Many-to-Many
On the Many-to-Many blog Archived 2005-12-20 at the Wayback Machine, NYU professor Clay Shirky gave an initial response last Monday to the Sanger article after the news reached Slashdot. While conceding that Sanger was partially right in his criticisms, Shirky said Wikipedia was being given "too little credit" for its efforts in dealing with the problems Sanger complained about. He also argued that while Wikipedia resembles a traditional encyclopedia, its critics fail to recognize that ultimately it has a different nature, much like an automobile differs from a horseless carriage.
The next day his colleague on that blog, danah boyd, took the other side of the debate. Echoing McHenry's observations about Alexander Hamilton being "edited into mediocrity", boyd suggested the quality of collaborative work was overrated: "Wikipedia appears to be a legitimate authority on a vast array of topics for which only one individual has contributed material."
The debate continued as Shirky and boyd exchanged further responses. Later on, Shirky posted a Wikipedia screenshot[1] containing the output of a script he had written, which showed below the title of the article how many times it had been edited, by how many users, and the dates of the first and most recent edits. He suggested that this would give at least some idea of how trustworthy any given article is, and hoped somebody would implement it as a setting on Wikipedia. Whether that happens, we shall have to wait and see.
Articles on major disasters fall short of featured status
September 11 article shot down over stylistic issues
September 11, 2001 attacks, an article with an important position in the history of Wikipedia itself, was nominated for featured article status last week. However, the nomination appeared doomed to fail, as discussion of the article's quality degenerated into an argument over whether the title required another comma.
The "missing" comma after "2001"—already debated extensively on the talk page, with an entire talk page archive dedicated entirely to the issue—continues to plague the September 11 article even after a truce on more emotional issues, like the use of the word "terrorist". Key figures in this debate include Maurreen and jguk.
Maurreen has consistently supported adding the comma, citing a variety of professional style guides for the rule that the year should be followed by a comma or other punctuation mark when a date is written out as day, month, and year. Jguk and others argued that this custom does not apply in this instance, based on widespread usage with the comma missing, particularly in international English usage. They also pointed out that the guidelines cited by Maurreen do not address usage when a date is used as an adjective, as it is in this particular situation.
The discussion and two polls on the talk page in October failed to reach consensus. Others suggested avoiding the issue entirely by moving the article to "Attacks of September 11, 2001" or a similar title.
How far apart do the waves need to be?
Meanwhile, another article that has gotten Wikipedia a lot of press based on similar circumstances, 2004 Indian Ocean earthquake, had its featured article nomination overwhelmingly rejected. It seems the article (which was still getting 68 edits in a day as of last Sunday) is not stable enough for everyone's tastes. Some people also didn't want it to end up as the featured article on the Main Page while it was still being showcased as part of the news section.
However, brilliant prose was being recognized in areas besides "major disasters of special significance to Wikipedia". Articles that did receive acclamation as featured articles this past week included Witold Pilecki, Liberal Party (Utah), Chemical warfare, Ryanair, and Craters of the Moon National Monument and Preserve.
In addition to the featured articles, here's last week's newly-made featured pictures:
Blog sets off edit war, gets itself protected
Law professor Glenn Reynolds, who writes the Instapundit blog, triggered an edit war last week over the article about his blog. After he noted the article's inaccuracies in his blog, a surge of traffic and vandalism led to the article being protected from editing.
Writing in his blog last Thursday, Reynolds commented on two matters in the article, one of them a picture showing him wearing a t-shirt that reads "I had an abortion", and the other a reference to a story about him putting puppies into a blender to create an energy drink. He also noted that contrary to the information given, his blog had never been hosted by servers at the University of Tennessee (where he teaches). Reynolds' assessment was that this "does little to inspire confidence" in Wikipedia's trustworthiness.
Puppies, lies, and photoshop
The picture, which is apparently a photoshopped fake, had been in the article since November when it was added by Urbansombrero (a user whose only contributions were to upload the picture and add it to the Instapundit article). Until Reynolds commented on it, nobody had bothered to do anything about the picture, although the fakery was noted on the talk page.
The "puppy-blender" story has a longer history, as it was originally promoted in April 2003 by Frank J. on the IMAO blog as a "filthy lie" for people to tell about Reynolds. Since then, the joke/lie has had a fair amount of circulation in the blogosphere, and its over-the-top nature makes it fairly easy to detect as falsehood, even if the teller doesn't actually point this out.
As the Instapundit article is regularly visited by vandals, this story naturally makes popular fodder for their material. The story has often been added as if it were fact, and on other occasions simply mentioned as being a story told about Reynolds. Either way it has regularly been removed again by people monitoring the article.
Traffic moves in both directions
By blogging about these issues, Reynolds managed to drive some additional traffic to the Instapundit article, some of which continued the vandalism. Finally, administrator Violetriga protected the article for about 24 hours. Since the unprotection of the page, things have calmed down, but a picture to illustrate the article is still lacking.
Apparently Reynolds also got some feedback from people who didn't take kindly to his comments, much like some other media figures have noted receiving flames when they publish material that is critical of Wikipedia. In any case, Reynolds later updated his blog to add, "I didn't mean for the original post to be a big slam on Wikipedia, just a comment on a not-very-reliable post."
Foundation fundraising drive postponed
The Wikimedia Foundation's next fundraising drive, originally planned for this month, has apparently been postponed until sometime in February.
The Wikimedia Foundation has been conducting these fundraisers roughly on a quarterly basis recently, primarily by posting a notice on all Wikimedia project pages inviting people to donate to support the Foundation's efforts. The most recent fundraising drive, which coincided with the milestone of reaching 1 million Wikipedia articles, raised over US $50,000 to support site operations.
The planned fundraiser was initially scheduled to coincide with Wikipedia's anniversary this month, as announced last Tuesday by Volunteer Representative Trustee Angela Beesley. She suggested that in preparation, the sitewide notice for the fundraiser could benefit from improvements, and mentioned the possibility of sending out a press release for the fundraiser and anniversary.
Beesley added that developer Tim Starling is working on a database that will allow the Foundation to start accepting memberships. Meanwhile, a quarterly budget for the use of these funds is also in the works.
Responses to the announcement indicated that the timing of the fundraiser was a bad idea. In addition to the fact that people may be strapped financially after the holiday season, Dori pointed out that this would be very close to the 2004 Indian Ocean Earthquake, which is commanding much of the world's attention in terms of charitable giving. Wikipedia itself has been contributing to the situation by having a prominent link on the Main Page to information about donations to the relief effort.
With these considerations in mind, Member Representative Trustee Anthere indicated that the Foundation could afford to wait a little bit. With Chief Financial Officer Daniel Mayer in agreement, it appears that the fundraiser will be held in early February instead.
The Report On Lengthy Litigation
This will be a regular column dealing with cases before the Arbitration Committee and other matters going through the dispute resolution system. Hopefully, it will not get too long to write. (Yes, the acronym is intentional, with tongue planted firmly in cheek.)
New Arbitrators finish off more cases
The recently elected crew of new Arbitrators continued to make progress toward reaching decisions on their existing cases. Within a couple of days of beginning their terms, they had already closed a couple of obviously stale cases, and concluded a one-year ban on Lir.
Last week, the Arbitration Committee closed two additional cases, prohibiting original research and edit warring over the Phil Gingrey article, and banning Alberuni for over a year. The Arbitrators also moved toward issuing rulings in most of the other cases currently on the docket.
In their oldest case, however, involving 172 and dating back to last August, things actually moved backwards when former Arbitrator Martin Harper withdrew his votes because he is no longer participating in the process. This left it unclear how long it would take to issue a ruling.
Cases with fresher and well-organized evidence seemed easier to deal with, as reflected in the Alberuni case. Alberuni was brought to Arbitration after a long period of stirring up matters on the Arab-Israeli conflict pages and exacerbating the tension among editors there. In his activity, Alberuni engaged in numerous personal attacks and created a number of sockpuppets for this purpose (sockpuppets that could be identified have already been banned).
In Alberuni's case, the Arbitrators voted strongly in favor of a one-year ban for personal attacks, plus an additional month for violations of the three-revert rule. After cleaning up loose ends in the ruling, the Arbitration Committee closed the case officially on Monday and the ban was promptly imposed.
By the end of the week, the Arbitrators had also managed to begin making proposals and voting on them for all existing cases, even on their newest case involving Rienzo, which was only just accepted on 1 January. One additional new request was accepted and moved into the evidence presentation stage (see below).
More POV templates spring up
Quite possibly the most active Arbitration case last week did not involve deliberation and voting by the Arbitrators, however. That would be the case of CheeseDreams, which prompted additional controversy when she returned to active editing.
CheeseDreams, who has a history of getting into conflict over articles related to religion, particularly Christianity, had been given a temporary injunction against editing articles related to Christianity. In the course of these conflicts, CheeseDreams created several sockpuppets, generally taking the same username with a slight spelling variation (such as Cheesedreams). No evidence has been raised that the sockpuppets were misused, however.
Whatever the account involved, until last Wednesday CheeseDreams hadn't edited since before Christmas, so the issue seemed to be dormant. But then she returned and created two new templates, dispute tags indicating that the article was skewed towards "Bible POV," which she then proceeded to add to a number of articles related to the Old Testament or Tanakh.
IZAK promptly seized upon these edits as violations of the preliminary injunction and listed the templates at Templates for deletion, also requesting that a block be implemented. IZAK also removed the tags wherever they had been posted.
Nobody had given CheeseDreams notice of the injunction, however, until after these edits (of course, her use of sockpuppets compounds this problem by making it hard to know the right place to give such a notice). The notice was finally given by Arbitrator mav Thursday on CheeseDreams' talk page after the controversy over the templates was raised.
However, Ta bu shi da yu, another participant in this Arbitration case, acted on IZAK's complaint and imposed a 24-hour block. This had to be reversed because the edits preceded the notice of the injunction, and Ta bu shi da yu posted an apology on CheeseDreams' talk page.
The fate of the templates remains to be decided, but they too had a dizzying career. After first being listed on Templates for deletion by IZAK, they were speedily deleted based on the apparent violation of the injunction. This led to their being nominated on votes for undeletion, where they were promptly undeleted because of the irregularities in the speedy deletion. At last check, the discussion had returned to templates for deletion again.
Speed does not equal unanimity
The Arbitrators also got an opportunity to demonstrate how quickly they would respond to new requests when dab made a Request for Arbitration involving Antifinnugor. The response was in fact quite swift, as seven Arbitrators responded to the request within 24 hours.
But even this was not enough to get a decision on whether or not to accept the case. While three of the necessary four Arbitrators had voted to accept, three others abstained and one voted to reject it, and their comments indicated that they still thought formal mediation could help the situation. Dab's response was that attempts at informal mediation had been ignored, and the Arbitrators did eventually decide to accept the case.
Dab's was the only new request accepted last week, so the net effect of this activity is that the Arbitrators have considerably lessened their caseload.
Wikipedians fight possible earthquake aid scam
Radman1 (also known as RaD Man or by his real name, Christian Wirth) last week led a campaign to remove information from Wikipedia about a shadowy organization exploiting the 2004 Indian Ocean earthquake to solicit donations.
Who is QuakeAID?
While busy arguing for the inclusion of a donation banner to help victims of the earthquake and tsunami, Wirth came across a newly created article for QuakeAID, the work of Baoutrust. This user was apparently an employee of the Baou Trust, QuakeAID's parent organization. His suspicions aroused, Wirth soon listed QuakeAID, along with other articles by Baoutrust, on Votes for deletion.
As to QuakeAID itself, its internet presence before Wikipedia consisted of little more than a skeletal website with links to buy memberships as a means of giving. Memberships would be set up as a recurring charge handled through a third-party credit card processing service. Some contact information in New York was provided, but efforts to identify QuakeAID as a charity registered in New York or anywhere else have been unsuccessful so far. In addition to the donation link, an online form to apply for relief was subsequently added (keeping in mind that internet access may not be readily available to many of the victims, no information was available as to how many applications had been submitted). Still, the operation showed many signs of being a scam set up to prey on the goodwill of people stirred by the crisis.
The initial response on VfD was overwhelmingly to delete the articles. Baoutrust was eventually blocked for violating the three revert rule over this and removing the VfD notices from the articles.
Baou Trust gets personal
The Baou Trust then used its press release service, known as OfficialWire, to carry an editorial attributed to one Jennifer Monroe, entitled "One man's personal quest against earthquake charity". The editorial savaged Wirth for his "personal war against QuakeAID". It did, however, become one of the first outside "media" publications to spell Wikimedia correctly (a feat also accomplished last week by John Gapper in the Financial Times).
Interestingly, Wirth is known for being an extreme inclusionist in terms of voting to keep articles listed on VfD. This comes partly from having twice defended the article about himself from deletion (his article reports that he is a figure in the computer art scene).
From his background in art, Wirth was able to recognize two more questionable articles by Baoutrust, about the artists Kaith (or Katherina Theohari) and George Dracos, and nominated them for deletion as well. As Dpbsmith put it, "A VfD nomination by him carries extra weight and deserves careful consideration".
Investigations continue
In the meantime, several Wikipedians were busily investigating QuakeAID to determine if it was in fact a fraud. As a result of these efforts, considerable new information was added to the article, converting what was originally a promotional piece substantially into an exposé on the organization.
As the situation developed, upon further reflection Wirth reconsidered and changed his position on the QuakeAID article, saying it had "significantly developed since its inception". A number of other people also changed their votes, and additional comments were running strongly in favor of keeping this article.
OfficialWire had an article created as part of Baoutrust's endeavors as well, which somebody else nominated for deletion. At last check, it appeared that except for the QuakeAID article all of the deletions would proceed, for the fairly common reason that Wikipedia is not a vehicle for advertising and self-promotion.
Note: As a general policy, The Wikipedia Signpost takes no official position on the question of whether particular articles should be deleted. However, and not just out of modesty, we do suggest that just because we mention a subject here, that fact alone would not make the subject newsworthy enough to warrant an encyclopedia article. Otherwise, the meta-implications of having a news publication like this would be enough to make your navel spin.
Work continues to fix MediaWiki upgrade
The upgrade to MediaWiki 1.4 continues to struggle, as implementation of new features is aborted and bugs in previously working features remain to be fixed. A number of fixes were in the works, as developer Brion Vibber announced the experimental release of a fourth beta version early Saturday prior to the final public release, but some problems still persist.
Abortive RC patrol experiment
The ability to mark edits on Recentchanges as being checked by a reader, a long-awaited feature that would reduce the redundancy involved in RC patrol, was shut off again Thursday after developer Tim Starling enabled it for a trial run. In the few days that it was available, users checking the diff of a change could mark it as "patrolled" by clicking a link on the diff page. Unpatrolled changes showed up with bright exclamation marks on Recentchanges and were highlighted in yellow on Newpages.
As it turned out, a number of people struggled to figure out the new feature and wondered whether it would really work out. Quite a few observed that hardly anybody seemed to be actually using it.
Additionally, the way the system was set up, a change could only be patrolled once, and users could even patrol their own edits. In fact, there was no way to know who had checked an edit or evaluate whether the patrollers were themselves trustworthy.
Starling criticized it as a flawed feature and said, "I'd rather write my own from scratch." Pcb21 was more hopeful but commented, "I think the little exclamation marks are going to have to appear on watchlists too before this will really take off." But in spite of the potential to help improve Wikipedia's review processes, the feature was deactivated again, and will presumably get an overhaul before we see it in action again.
Running some statistics after the experiment shut down, developer Jamesday determined that fewer than one percent of all edits had been marked as patrolled. Newpages, which contains a substantial amount of Wikipedia's garden-variety vandalism, was patrolled somewhat more often at eight percent, but this reporter's experience suggests that the real amount of patrolling done on Newpages is significantly higher. As might be expected, anon edits were checked with greater frequency than those of logged-in users.
Problems with blocks
Meanwhile, automated logs of activity such as uploads and deletions have undergone a conversion to act more like page and contribution histories, although some functionality was missing. More seriously, however, the software change has broken the blocking feature—or, as it might be more appropriate to say, the unblocking feature.
This bug or collection of bugs has several manifestations, the most serious of which is that when an admin sets a block to expire at a given time, the block nevertheless does not expire. Instead, the user account or IP address stays blocked until it is unblocked manually.
The problem is compounded by the confusing situation on the block log and list of active blocks, neither of which is displaying properly. The block log fails to indicate when any of the blocks are supposed to expire, making it of little use in determining which blocks are ready to be undone.
Expiration times are available by checking the list of currently active blocks, but due to an additional malfunction, the list has to be used carefully so as not to make mistakes. A number of blocks are deliberately set indefinitely, with no expiration time, but the list, which used to indicate these properly no longer does so. Instead, at various times it has shown indefinite blocks as expiring either at the precise time the list is being viewed by the reader, or at the time the block was initially implemented. Thus, an admin going through the list to reverse expired blocks must sort through this incorrect information to avoid undoing a valid block.
The developers are aware of these problems, but due to other issues connected with the software upgrade, as well as the basic work needed to simply keep the site operational, haven't been able to fix them as yet.
Wikipedia moves into top 100 websites
Just before the New Year, Wikipedia cracked Alexa's list of the top 100 English-language websites for the first time, and its traffic rankings continue to rise.
Wikipedia first entered the top 100 on 29 December, as the result of steady growth since September in particular. While the initial surge was part of an Internet-wide increase in traffic to reference sites, attributed primarily to start-of-school traffic from students, Wikipedia appears to be holding its gains and building on them.
When last checked, Wikipedia was sitting at #92 for English-language websites. It had also moved into the top 200 in global terms (using Alexa's three-month average), reaching #196 in that list.
In the near future, Wikipedia might become the second most popular "dot-org" in the world, based on Alexa statistics, ranking behind only craigslist. At the moment, Wikipedia traffic is very close with that of miniforum.org, a Chinese language site, for second place.
But future growth also depends on having the technical capacity to grow. In addition to the challenge of keeping enough servers to handle the load, the Wikimedia Foundation has new limitations to overcome.
Lately, Wikimedia has been running up against the limits of its internet connection speed, which had a maximum of 100 megabits per second. According to developer Jamesday, a switch to a one gigabit per second connection was planned for last Wednesday, but the colocation facility was missing a part needed to make the change. At last report, the part had not yet arrived and the schedule for the changeover was not determined.
AMA begins plans for new election
The Association of Members' Advocates, an independent association involved in the dispute resolution process, has started discussions on having a new election for the position of AMA Coordinator.
Since the initial election of Alex756 to this position last April was for a term of six months, it appears the new election is already several months overdue. A poll on the question of having new elections was running strongly in favor, but general discussion among the members was a little more uncertain how to proceed.
Voicing an outside opinion on the AMA, Ambi said she thought it was "completely dead" and hadn't seen it in action anywhere. Some advocates responded that their work hadn't been done in the later stages of the dispute resolution system, noting that they seemed mostly to be helping newcomers who weren't familiar with Wikipedia procedures. Since this was rather different from the role the AMA originally conceived for itself of working primarily in mediation and arbitration, it opened the question of rethinking the AMA's basic focus entirely.
To figure these issues out, Alex756 proposed that the AMA have a meeting for all advocates. The meeting, tentatively to be held via IRC, would address matters such as choosing officers, drafting a charter, and how to get the organization more active and effective in Wikipedia.
Given the uncertain state of affairs, several advocates expressed interest in improving publicity for the AMA as an organization. Alex756 suggested conducting a study to analyze the operations of the Mediation and Arbitration Committees and issuing a report. Advocate KeithTyler responded that he thought this might exceed the role and function of the association.
It appears that further discussion of these issues will have to wait for the AMA's meeting, but it is not yet certain when that will be held.
Expansion of speedy deletion up for vote
Voting continues this week in a proposed expansion of the speedy deletion criteria. With a full week of voting already concluded, strong trends in the voting were already apparent.
Creation of proposal
Originally developed by blankfaze, the proposal aims to expedite the deletion process for some obvious cases that do not need fuller consideration at Votes for deletion. Blankfaze originally suggested five types of cases that should be covered, but several have been added since discussions began. The sprawling full proposal actually contains eleven different proposals, each being voted on separately.
Started on January 2, this survey will continue until January 16. According to its instructions, any specific proposal would go into effect on "candidates for speedy deletion" if it receives support from at least 70% of voters.
What's passing, what's not
By the 70% standard, the trend of votes being cast indicates that most of the proposals will not pass. In fact, several do not even have majorities in support.
As of the first week's voting, the leading proposal that would succeed is speedy deletion of "articles" that are nothing but external links, tags, or similar trappings. Also apparently headed for passage is the option of speedily deleting articles when requested by their author—if that author is the only person to have edited the article.
Falling a little bit short of having enough votes to pass was a proposal to speedily delete "Extremely short articles which add no information beyond what is obvious from the title." Most of the other proposals were being rejected overwhelmingly.
One particularly odd case was included as a proposal: "Any article which consists only of attempts to correspond with the person or group named by its title." However, unusual as it may seem, this proposal definitely seemed headed for passage. As Humblefool observed, "There aren't many of these anyway, are there?"
It is not clear how much the results of this survey will actually affect speedy deletion practice, or whether any of these measures will reduce the workload headed to Votes for deletion.