Jump to content

Wikipedia:Wikipedia Signpost/Single/2012-10-15

From Wikipedia, the free encyclopedia
The Signpost
Single-page Edition
WP:POST/1
15 October 2012

 

2012-10-15

AdminCom: A proposal for changing the way we select admins

Looie496 is a neuroscientist specializing in learning and memory. He has been an English Wikipedia editor since 2008, where he writes articles about the nervous system and maintains WikiProject Neuroscience.
The views expressed in this op-ed are those of the author only; responses and critical commentary are invited in the comments section. The Signpost welcomes proposals for op-eds at our opinion desk.

There is wide agreement among English Wikipedians that the administrator system is in some ways broken—but no consensus on how to fix it. Most suggestions have been relatively small in scope, and could at best produce small improvements. I would like to make a proposal to fundamentally restructure the administrator system, in a way that I believe would make it more effective and responsive. The proposal is to create an elected Administration Committee ("AdminCom") which would select, oversee, and deselect administrators.

What's wrong with the system we have now?

Related articles
Reforming RfA

Will the new RfA reform come to the rescue of administrators?
16 May 2024

Jimbo's NFT, new arbs, fixing RfA, and financial statements
28 December 2021

Editors discuss Wikipedia's vetting process for administrators
26 September 2021

Administrator cadre continues to contract
31 July 2019

The Collective Consciousness of Admin Userpages
31 January 2019

The last leg of the Admin Ship's current cruise
31 July 2018

What do admins actually do?
29 June 2018

Has the wind gone out of the AdminShip's sails?
24 May 2018

Recent retirements typify problem of admin attrition
18 February 2015

Another admin reform attempt flops
15 April 2013

Requests for adminship reform moves forward
21 January 2013

Adminship from the German perspective
22 October 2012

AdminCom: A proposal for changing the way we select admins
15 October 2012

Is the requests for adminship process 'broken'?
18 June 2012

RFAs and active admins—concerns expressed over the continuing drought
14 February 2011

RfA drought worsens in 2010—wikigeneration gulf emerging
9 August 2010

Experimental request for adminship ends in failure
13 October 2008

Efforts to reform Requests for Adminship spark animated discussion
23 April 2007

News and notes: Arbitrators granted CheckUser rights, milestones
6 February 2006

Featured picture process tweaked, changes to adminship debated
27 June 2005


More articles

As I see it, there are two fundamental problems with the current system. The first is that it makes each admin into a sort of Lone Ranger. Each admin independently decides what to work on and how to handle it. Good admins seek advice and consent from the community, but nothing forces them to. Admins are selected using a chaotic process, and can only be removed for egregiously bad behavior—but even then the removal process is very arduous. There is no coordination or effective oversight. On television, the Lone Ranger is a hero; in real life, where people do not always share a common understanding of right and wrong, lone-rangering is an endless source of trouble.

The second problem is that the current system forces each admin to be a politician. The requests for adminship process (RfA), regardless of any propaganda to the contrary, is an election, and it is impossible to win an election without behaving like a politician. Furthermore, winning an election serves as a validation of the views that the candidate expresses, and therefore adminship, like political office, becomes a power base and a status symbol, rather than merely a way of serving the community. What we want are admins who behave like civil servants, such as a dogcatcher or police officer; what we get are admins who behave like politicians. We talk about our admins being merely editors with some extra buttons, but that is not what the RfA process gives us.

These problems amplify each other. Because bad admins are so difficult to remove, the community has become very strict in its examination of RfAs—but the more difficult it is to succeed with an RfA, the more candidates are forced to behave like politicians, carefully avoiding any action that might offend a significant group of voters and carefully hiding any views that might be controversial. And the more difficult it is to succeed with an RfA, the more status comes from success, making adminship a goal in its own right.

How can we fix it?

What we need is a system that accomplishes three goals: (1) the community is in charge; (2) adminship is easy to give and easy to take away; (3) admins do not need to be politicians.

Having an AdminCom would accomplish all of these goals. The community would ultimately be in charge, because the community would elect the members of AdminCom. We would simply replace direct democracy with a representative democracy. Handing out the buttons and taking them away would be as easy as AdminCom decides to make it. Admins would be directly answerable to AdminCom rather than to the editing public at large, and so would get clearer guidance than they currently do.

The primary power of AdminCom would be to give and remove adminship. Other powers, such as overseeing and advising admins, would naturally follow from this. I do not believe it would be a good idea to formalize internal procedures for AdminCom: the members should be able to decide for themselves how to make things work. Naturally they would express their beliefs about proper procedure when they run for office, and the Wikipedia community could make its choices accordingly.

Objections and answers to them

If having admins be politicians is bad, why is it good to have politicians choosing and overseeing them? The answer is that any effective organization requires politicians at the top. Consider, for example, a city government. The people who carry out primary duties are civil servants, and answer to their bosses. In order to have a responsive system, though, the bosses have to be answerable to the public. That way the very difficult task of making the public happy is left to people who are experts at it (the politicians), and is not directly imposed on people who would only be impeded by it. The advantage of an AdminCom is that it separates the political functions from the administrative functions. Doing away with the political functions entirely would only be possible in a dictatorship.

Another potential objection relates to bureaucrats: do they not already fulfill the AdminCom role? No. Bureaucrats, as they currently function, have far less decision-making power than admins. All they really do is count votes, with a strictly limited power to decide how votes should be weighted. Thus, turning our current bureaucrats into an AdminCom would give them powers and responsibilities that nobody had in mind when they were chosen.

In conclusion

I have been thinking about these issues for a long time, and this is meant as a serious proposal. Of course I am aware that many editors will react to the AdminCom idea with dismay, for a variety of reasons. What is not so clear is whether there is a substantial community who would be favorably disposed. If a large number of editors endorse the idea, it will encourage an effort to put the proposal into a specific form and then submit it to the community as a referendum.

Reader comments

2012-10-15

Wikipedia's language nerds hit the front page

Beatles debate on the front page

John F. Kennedy International Airport, 7 February 1964: The the? Beatles arrive to rapture
This week saw a front-page story in the Wall Street Journal (the largest newspaper in the U.S. by circulation) on editorial debates in Wikipedia. The story focused on the title-naming dispute surrounding the Beatles article, and specifically the RfC on whether the "the" in the band's name should be capitalized or not.

This isn't the first time The Beatles (or is it "the" Beatles?) has shown up in the press: a 2009 Telegraph story named it as the number-two read article on Wikipedia, behind the eponymous (and often accidentally reached) wiki.

As of this writing, there were 45 comments on the Wall Street Journal story, many of them debating the capitalization point itself, while others raised the question of whether consensus on Wikipedia was a viable decision-making model. The most recommended comment opined "They argue so loud because the stakes are so low...."

The results of the RfC will be announced next week.

Conflicts of interest

Since the last installment of In the media, a few stories reported on the controversy surrounding GibraltarpediA (see the Signpost's 24 September detailed report and the followup on 1 October). The Telegraph ran a story on 2 October, following the publication of two more CNET stories by Violet Blue, who first wrote about the story.

In brief

2012-10-15

Second star to the left

This edition covers content promoted between 7 and 13 October 2012
LH 95, a new featured picture
"We Can Do It!"
Profile of a Boulonnais horse
A Hercules beetle

Five featured articles were promoted this week:

  • "We Can Do It!" (nom) by Binksternet. "We Can Do It!", an American propaganda poster from the Second World War, was produced by J. Howard Miller in 1943 to boost worker morale; it was shown for only a month, in several Washington-area factories. Although rarely seen during the war, after the poster was rediscovered in the 1980s, it was seen as a feminist icon and became widespread in popular culture, although its origins are frequently misunderstood.
  • USS Lexington (CV-2) (nom) by Sturmvogel 66. Lexington was a ship built for the United States Navy that, although ordered in 1916 as a battlecruiser, was converted to an aircraft carrier and launched in 1928. After years running training and humanitarian exercises, during the Second World War Lexington participated in several battles in the Pacific theatre before being scuttled in the Battle of the Coral Sea in 1942.
  • Boulonnais horse (nom) by Dana boomer. The Boulonnais is a heavy draft horse breed known for its large but elegant appearance. The breed's origins trace to a period before the Crusades, although the current breed features Spanish Barb, Arabian and Andalusian blood. During the 19th century Boulonnais saw popularity in the US and France, but populations declined during the World Wars. Today the breed is making a come-back as a common source of horse meat.
  • OK Computer (nom) by Brandt Luke Zorn. OK Computer is the third studio album by the English alternative rock band Radiohead. Their first self-produced album, OK Computer and its highly experimental music was recorded in the St. Catherine's Court mansion. Before its release in 1997 the album was deemed "uncommercial", but it eventually sold more than 4.5 million copies, charting in several countries. It has been called one of the best albums of all time.
  • "The Truth" (The X-Files) (nom) by Gen. Quon. "The Truth" is the two-part 2002 series finale of the American series The X-Files. The episode follows several FBI agents uncovering a conspiracy including genetically engineered supersoldiers. The most watched installment of its season, "The Truth" received mixed reviews, with several citing the episode's lack of closure.

Ten featured lists were promoted this week:

Four featured pictures were promoted this week:

The USS Lexington in October 1941


Reader comments

2012-10-15

Chapters ask for big bucks

The FDC is the most significant reform of the WMF's finances, and will hand out up to more than $11M this financial year. The community can comment on the 11 applications of the first application round until next Monday.

The volunteer-led Wikimedia Funds Dissemination Committee (FDC) and interested community members are looking at Wikimedia organization applications worth about US$10.4 million out of the committee's first full year's operation, in just the inaugural round one of two that have been planned for the year with a planned budget of $11.4M.

The WMF introduced the new finance structure earlier this year, with the FDC at its core, to promote transparent and accountable spending among Wikimedia entities. Fifteen organizations of 19, including the foundation itself, have met the transparency criteria in general, although several chapters encountered problems in satisfying the new requirements for reporting on their past finances and activities—some documentation having been overdue since 2010. Eleven chapters plus the WMF submitted funding proposals.

The largest chapter, Wikimedia Germany, which employs more than 44 people and has some 500 active volunteers and more than 2500 members according to its submission, is requesting the largest amount of funds among the chapters. WMDE is requesting $1.82M (24.3%) to fund parts of its planned $7.475M expenditure over the next fiscal year—a considerable increase on the chapter's current spend of $4.316M. The chapter's funding and expenditure are more complicated than for most: for example, funding for the Toolserver in Amsterdam, which is increasingly important to the running of Wikimedia sites, has been the subject of intense debate. Other projects, such as Wikidata ($2.75M), have been funded by the chapter from third-party donors. Among WMDE's initiatives will be new efforts to make state-owned cultural works freely accessible on movement sites, a greater presence at EU level, and the testing and evaluation of two innovative support tools to help readers and editors to improve balance on Wikipedia articles.

Wikimedia UK has five permanent staff and one intern, 87 active volunteers and at least a further 100 who occasionally participate, and 330 registered members. WMUK is asking the FDC for $919K (67.4%) of its annual budget of $1.365M. The entity's activity plan for the upcoming year provides considerably more detail than most other chapter plans. Among the chapter's initiatives will be support for Europeana uploads, a "train the trainers" program, the development of modular online training, outreach to editors working in the Welsh language, and moves to increase female participation in the movement. The chapter will continue its support of Wikipedian-in-residence positions with partnered GLAM institutions, including digitization efforts and a focus on Scottish museums.

Wikimedia France, with four staff members (soon to be five) and 303 members (20–30 of them "very active"), is requesting $961K (68.8%) of a total budget of $1.397M. Its plans include reaching out to new editors in universities—including PhD students, other students, and academic researchers—with about 15 training workshops, particularly on ecological topics, and a new partnership with the Société Française d’Écologie. The chapter will spearhead the promotion of scientific knowledge about Wikimedia, including the running of a research project on the geolocalisation of Wikipedia articles, involving engineering students from the major engineering college École Centrale de Lille. There will be a number of new GLAM projects, and new French-language outreach efforts, particularly in Africa and in support of minority languages in France. Language projects will include the launch of a francophone newsletter about Wikimedia projects, inspired by the Signpost and the French Regards sur l'Actualité de Wikipédia.

The WMF's own application, for $4.46M (10.6% of its $42.070M budget), involves the organization's non-core spending. For the first time, then, this part of the foundation's spending will be subject to community comment and review. Affected are programs such as the global Education Program ($718K)—of which the US and Canadian components are currently the subject of an RfC—and editor engagement experiments (E3; $1.2M).

Other chapters applying are Wikimedia Argentina, Australia, Switzerland, Israel, Netherlands, Sweden, Austria, and Hungary. All applications can be reviewed by the community until 22 October. The FDC will consider both submissions and comments, and will submit its spending recommendations to the WMF board of trustees by 15 November 2012.

In brief

A 19th-century colour portrait of English mathematician and writer Ada Lovelace, after whom the US Ada Initiative is named.
  • Ada Lovelace Day. Wikimedia UK and the Royal Society have organized a women-in-science editing event for Ada Lovelace Day on Friday 19 October 2012 in London. To be held at the Library of the Royal Society, the event has the backing of notable Member of the Royal Society, developmental psychologist Uta Frith. Participation is free, although demand is such that all places at the afternoon editathon have been taken up; remote online participation is welcome. The evening panel discussion, which will include Professor Frith, is open for attendance to all who are interested in the topic. Wikimedia UK has a dedicated page for the event. See also the Twitter hashtag: #WomenSciWP.
  • English Wikipedia:
    • Arbitration report: there are no open or pending cases, nor any requests for clarifications and/or amendments. There is one motion open on the "net four votes" rule. The RfC concerning the December 2012 Arbcom election is in full swing. The issues concern the composition of the committee, the election itself, and the minimum support percentage required at the end of the election.
    • English Wikipedia main page redesign: On 11 October, a straw poll to determine community support among the 25 submissions has been started.
    • DYK eligibility criteria under review: On 7 October, Gilderien proposed a reform of "Did you know" to allow new good articles to go through the DYK process, as opposed to only new and 5x expanded articles.
  • Wikimedia USA bylaw-vote: Discussions of an organization supporting chapters and other volunteer groups in the US reached a new stage last week as Guerillero called for a vote to ratify the draft bylaws on Meta. Editors located in the United States who are interested in on-the-ground work are invited to make their views known in the proceedings until 22 October 2012.
  • Travel guide project named Wikivoyage: On 16 October, the WMF announced the RfC result, concluding that the travel guide will be called wikivoyage.org, reflecting the name of the largely German site that was set up in 2006 and will migrate to the foundation. Meta is now looking at potential logos for the new project.
  • Wikipedia Zero partners with Saudi Telecom: On 14 October, the WMF announced it will join forces with the Saudi Telecom Company to make Wikipedia freely available to the company's 25 million mobile customers, mostly in Saudi Arabia, Bahrain, and Kuwait. The initiative has been made possible through a collaboration between the foundation, STC, and Intigral, a company that specializes in providing digital media technology to telecom operators. This is the third major partnership with a mobile provider the WMF has finalized this year to promote free knowledge in emerging societies, as outlined in the foundation's strategic plan, with its priorities for supporting both mobile users and Arabic-speaking readers and editors.
  • Wikidata main page design: Wikidata, the new project aimed at a collaboratively editable database to support Wikipedia and other foundation sites, has published a call for proposals on how to design its main page.
  • Toolserver funding proposal: On 16 October, the key volunteer administrator of the Wikimedia Toolserver, DaB., announced his draft proposal for the WMDE general assembly to request sufficient funding for the Toolserver infrastructure for the next fiscal year. If the draft progresses as outlined in its introductory statement, the WMDE membership will vote on the issue on 24 November 2012 after the chapter's CEO, Pavel Richter, resisted concrete funding commitments in community deliberations.
  • Wikimedia Chapters Association resolution vote: Chapters taking part in the long-running effort to establish a chapters association (WCA) are voting on a resolution that a consultant should be hired to support a recruitment process for a CEO, to be called the Secretary General. All of the resolutions can be found on Meta.

    Reader comments

2012-10-15

Wikidata is go: well, almost

ContentHandler merge sets stage for Wikidata test

Wikidata developers and support staff celebrated the successful merger of over 10,000 lines of their code with the baking of a special cake.

A trial of the first phase of Wikimedia Deutschland's "Wikidata" project—implementing the first ever interwiki repository—may soon get underway following the successful passage of much of its code through MediaWiki's review processes this week.

At the heart of those developments of its "ContentHandler" branch, which comprised some 10,000 lines of code targeted at introducing alternative page formats to the vanilla "wikitext" variety (wikitech-l mailing list). This is required by Wikidata to allow it to serve editable pages that use its own structured data format rather than wikitext, but has potential applications for a number of projects such as that to introduce Lua code to MediaWiki. The merger of the branch was marked by a small number of (fortunately resolvable) bugs, although users are asked to be vigilant for more as the code hits larger wikis.

Also merged was the Wikidata-developed "Sites" facility. Initially running in parallel to existing processes, the "Sites" code takes the form of a not inconsiderable upgrade to MediaWiki's existing support for interwiki links. Pertinently for readers looking forward to the rollout of Phase I, review of both of these headline features (as well as a number of smaller patches also merged this week) had been the main items on the pre-trial to-do list for several weeks. That trial—scheduled for the Hungarian Wikipedia—is now expected to get underway early next month should no major bugs be found in the meantime.

"We're thrilled that this huge amount of work we've done over the last 6 months has finally made its way into MediaWiki core." Lydia Pintscher, Wikidata's communications chief, told the Signpost. "It's a huge step towards getting the first deployment done and at the same time allows a lot of great stuff unrelated to Wikidata to happen in the future."

The Wikidata project is currently looking for suggestions about how its main page should look and technical comment on its proposed update system. Discussions about how ContentHandler-aware extensions ought to be are also ongoing.

In brief

Not all fixes may have gone live to WMF sites at the time of writing; some may not be scheduled to go live for several weeks. Although there is no poll this week, you can still give your opinion on the topic "Which of the following best reflects your view about the desirable relationship between WMF staff and non-WMF-deployed extensions?"
  • MediaWiki 1.21wmf2 begins deployment cycle: 1.21wmf2—the second release to Wikimedia wikis of the 1.21 cycle—was deployed to its first wikis on October 15 and will be deployed to all wikis by October 24, taking in the English Wikipedia on October 22. Because it incorporates all the Wikidata-based work referred to above, some 700 changes to the MediaWiki software that powers Wikipedia are included in all, making it the largest fortnightly deployment to date. Among the changes not covered above is the introduction of "hi-res" images for those with screens and browsers that support it plus an overhaul of the CologneBlue skin to make it more future-proof.
  • MediaWiki 1.20 release and beyond: In related news, a first release candidate of MediaWiki 1.20 was also released to external sites this week, suggesting a final release is not far off. (Volunteer) release manager Mark Hershberger also gave his thoughts on a possible twice-yearly timetable for future releases if the WMF continued its policy of releasing to Wikimedia wikis fortnightly. The schedule suggested the introduction of "LTS" (Long Term Support) releases suitable for organisations that could not risk frequent updates; the number of versions of MediaWiki that developers should be expected to keep their extensions compatible with remains a contentious issue.
  • Gerrit 2.4.2 patched, but 2.5 not yet ready: Posting an update on the state of Wikimedia code review system Gerrit, developer Chad Horohoe described a new patch for the currently installed version (2.4.2) that improves Gerrit's support for "continuous integration" methods such as automated testing. 2.5 would not be deployed until a major authentication problem with it had been fixed by Gerrit developers, however (wikitech-l).

    Reader comments

2012-10-15

WikiProject Chemicals

WikiProject news
News in brief
Submit your project's news and announcements for next week's WikiProject Report at the Signpost's WikiProject Desk.
Various oxidation states of plutonium in solution
White and yellow aluminium chloride
Stick model of nicotinamide adenine dinucleotide
A brown, acidic solution of ferric chloride

This week, we experimented with WikiProject Chemicals. Started in August 2004, WikiProject Chemicals has grown to include over 10,000 articles about chemical compounds. The project has a unique assessment system that omits C-class, Good, and Featured Articles. As a result, the project's 11 GAs and 9 FAs are treated as A-class articles. WikiProject Chemicals is a child of WikiProject Chemistry (interviewed in 2009) and a parent of WikiProject Polymers. We interviewed Wim van Dorst (Wimvandorst) and Smokefoot.

What motivated you to join WikiProject Chemicals? Do you have some academic or professional experience in chemistry?

Wim van Dorst: I joined the WP:Chem at the early days of Wikipedia nearly 10 years ago, just to share opinions and proposals about new chemicals articles. There weren't many wikiprojects about, and the groundbreaking work was very interesting: everything was new. Chemically educated at the University of Technology Eindhoven, I have always been interested in information about chemicals. While commercially working in one of the leading chemicals companies worldwide, I have found the opportunity to share my wider interests with others in WP:Chem and Wikipedia.

The project maintains a unique assessment scale which divides articles into just four designations: A-class, B-class, starts, and stubs. Why was this system chosen over the typical Version 1.0 assessments? How does the project treat articles that have been recognized as Featured or Good Articles independently of the WikiProject? What have been the benefits and drawbacks of the system used at WikiProject Chemicals?

Wim van Dorst: Current day's typical assessment was actually a followup of our own proposal at the time. FA and GA came later and were of course taken up into the process. Yet, for historical reasons we retained the four classess. It has always been enough to serve the prime target of classes: to upgrade articles to better quality. More granularity would have lead to just more vagueness.
Smokefoot: My impression is that contemporary editors here pay little attention to article rankings. I view it as bureaucracy, ditto for FA and GA business.

A member of WikiProject Chemistry noted in a 2009 interview that WikiProject Chemicals is a more cohesive group than its parent project. Do you agree? What brings the editors of WikiProject Chemicals together? How much overlap exists in the scope, membership, and effort of WikiProject Chemicals and WikiProject Chemistry?

Wim van Dorst: WP:Chem came about out of the wider Chemistry group, to focus on a very specific object: to provide really good, structured, systematic information about important chemicals. This focus lead to close co-operation between the then handful of WP:Chem editors, first to determine which chemicals are important, and then to improve the articles themselves. Creating an assessment scheme was merely a tool to know which articles to work on, and measuring how far we were. There were specific articles that any editor preferred to work on for whatever reason: mine was Hydrochloric acid, because recently before that I had been a marketing manager for that product, and Walkerma worked on Gold(III) chloride, although we all dug in when an article was just before A-Class (our aim). That we achieved to get a handful of articles to Featured Article when the whole FA procedure was brand new was just more fun than the actual target. A great example of all of us joining forces was on Lead(II) nitrate, a historically very important chemical but nowadays rather obscure. It had been on our worklist for this good reason from the beginning, but given the lack of data it had not gotten much attention: a typical Stub-class article. Somewhere in 2006 or so, we all joined forces and made it a good A-Class article. And for the heck of it, we also ran for Featured Article. Yes, we were a cohesive group, apparently all in for some focussed edit work but some laughs about it as well. And nearly all of us are still around, although certainly not so focussed anymore.
Smokefoot: Unlike many topics in Wikipedia, our personal views are pretty orthogonal to our chemical inclinations so we don't bother each other. We are relatively collaborative, in part because we share interests and there are relatively few real controversies to divide us. As a practical matter, the scope is so large that many opportunities exist. The editing tone was set several years ago by a couple of editors who have maintained a very active participation and have a strong chemical knowledge.

Are there any significant holes in Wikipedia's coverage of chemicals? How successful has the project been in building articles for the red links included the project's lengthy lists of organic compounds, inorganic compounds, and biomolecules? What are some challenges projects face when cataloging vast collections of information like chemicals?

Wim van Dorst: Oh, yes, there are many chemicals not listed on Wikipedia. But then again, WP:Chem has never been focussed on ALL articles about Chemicals, only on the 382 designated core articles, selected by WP:Chem experts to well cover the 8 groups of chemicals compounds. And in the heydays of WP:Chem (viz 2004-2009), we achieved over 95% of all core articles at least Start-Class, and two thirds towards A-Class. So, the core of the WP:Chem did achieve a large part of the target what we are aiming for.
Smokefoot: These lists were constructed long ago, and one wonders if they were created by practicing chemists. The project is in good shape. There is some variability in the balance, clarity, and depth of articles, but factual inaccuracies are dealt with quickly.

WikiProject Chemicals ranks among the top ten most active WikiProjects by changes. What is the project's secret to remaining active? What are some areas in which the project could still use improvement? How can a new contributor help today?

Smokefoot: The chemical knowledge of the editors is high. There exist opportunities for both very advanced scientists and beginners to contribute.
Wim van Dorst: Wow, are we?? Awesome. I wasn't aware. Just a bunch of chemically interested people working on chemicals articles.

Anything else you'd like to add?

Smokefoot: There is some pull by nonchemists for greater advocacy about consumer issues and including generic safety information. I see this as an overall push to make Wikipedia more "politically correct". It is also a struggle to maintain an emphasis on more general sources vs specialized journal articles to support articles.

Next week, we'll take a break from our usual routine to shine a light on neglected parts of the world. Until then, explore the archive.

Reader comments

If articles have been updated, you may need to refresh the single-page edition.