Sunday, July 29, 2007

Reflecting on VT

I recently got a note from somebody researching the history of Wikipedia's article on the Virginia Tech massacre. In part:
You were among the top 5% of contributors (either editing or on the talk pages) for that article, and I was wondering if you’d be willing to answer a few questions by email.
Hadn't been aware of that, at the time. I do remember putting in quite a few reverts, though. User:Geraldckane's questions and my responses follow:

1) What is your Wikipedia screen name? Note: optional, if you reply by email I will not be able to connect your email to your screen name.

I am a volunteer administrator, username Luna Santin.

2) On average, how many hours per week do you spend editing articles on Wikipedia?

That's hard for me to say. I tend to alternate between studying, coding (I'm a computer science major), work around the house, and general web surfing. Javascript tells me I've been making about 50-60 edits per day, this last month, but many of those are single-click reverts.

3) Why do you contribute your time and energy to developing Wikipedia articles?

Used appropriately, I think Wikipedia is an excellent resource for the world community. Helping out gives me a chance to do something I enjoy, and while it is productive, it also gives me significant opportunities to apply abstract lessons in building community, working collaboratively, and other, more technical issues related to web-based projects in general.

4) What types of articles to which do you typically contribute?

I tend to focus on internal affairs -- catching and dealing with abusive editors, including vandals, trolls, and sockpuppets, resolving or calming down disputes when possible, and the like. I also moderate one Wikipedia mailing list, and several IRC channels. When I do contribute to articles, I tend to focus on either local affairs in my area, or on subjects I've recently covered in school or otherwise have easy access to -- took a history class, this Summer, and the textbook proved valuable.

I also try to spend some time copyediting, cleaning up new articles, welcoming new users, and helping people who have questions.

5) Why did you choose to become involved in the Wikipedia article on the Virginia Tech Massacre?

Whenever we have news events this major, especially something like this that'll really get the emotions going, there quickly forms a veritable flood of people, and Wikipedia needs a quick response to that. Some people work to build up an article, other people work to clean it up. I guess I tend to be in the latter group.

It really is amazing to me how quickly these things can happen. As one example, several Wikipedia administrators and "recent changes patrollers" were aware of Steve Irwin's unfortunate death even before it hit the news tickers on BBC's and CNN's websites. We have to be quick.

6) What was your primary role in the process of creating the article on the Virginia Tech Massacre (e.g. copy editing, fighting vandalism, contributing news, managing a particular section, etc?)

Definitely cleaning up after vandals and other misguided people. Some people are angry and don't understand that they aren't helping anything, other people really are just looking to cause trouble. Primarily, I protected the article from these disruptive influences so that it could be allowed to grow more appropriately.

7) How was your experience with this article similar to or different than other Wikipedia articles to which you have contributed?

It wasn't all that different from other prominent articles -- Wikipedia features one article per day on its front page, for example -- except that the number of people involved made a lot of things happen *very* quickly.

8) What were some of the most challenging issues facing the successful development of this particular article on the Virginia Tech massacre?

Several people were, I think understandably, feeling very angry, and some of them felt a need to express that anger in the article. Not the best choice, but I don't know if I can blame them. We also had a number of people adding unverified (or conflicting) information, but that's more easily sorted out among the experienced users -- we were fortunate to have so many veterans watching over things.

Internally, I recall some lengthy debates over whether to include Cho as one of the "casualties" of the shooting, or whether to use the Westernized Given name-Surname or Eastern Surname-Given name constructions. There were also some arguments over what exactly to call the article -- "Virginia Tech shooting" or "massacre," for example? I believe we eventually decided to go with whatever the media predominantly used to refer to the incident.

Beyond that, a sudden and massive influx of new users, many of whom were actually quite helpful, brings in a lot of people who aren't too familiar with wiki syntax, policies, and practices, but it's also an excellent opportunity for us to *teach* them these things. I'd like to think we managed pretty well, given the circumstances.

9) What do you think were some of the primary reasons that this article was successful (i.e. cited in the press, rated as a “good article” by Wikipedia standards.)

Wikipedia's ability to respond to changes in real time, and to amalgate content from diverse sources and viewpoints. By having more eyes on the article, we'll hopefully get a more complete picture of what happened more quickly. The trick is avoiding confusion, and figuring out at what point being open to modification is helpful or hurtful.

Administrators can "protect" pages from editing, locking out all new users or all non-admin users, but the majority of admins involved with the article agreed that doing so (except for short periods when absolutely necessary) was the way to go. By carefully managing the situation, we were able to maximize the benefits of the wiki system, while hopefully mitigating its vulnerabilities.

Any time you have a dedicated core of people diligently checking every change to a page, the situation is intensely unique.

10) Is there anything else I should know about the Wikipedia article on the VT massacre?

Hm, tricky one. I can't immediately think of anything I haven't already said. While I did feel, and still do feel, that the incident itself was incredibly unfortunate, it was of some help to see the outpouring of heartfelt sympathy for the victims and students of Virginia Tech, and to see that most people are truly good at heart.

Thursday, July 26, 2007

Guerrilla marketing

Between classes, lately, I've been hopping over to the school computer labs to take care of quick business on Wikipedia -- clearing AIV, skimming the noticeboards, answering a few unblock requests, and so on. Welcomed a number of people who'd edited the introduction page, today.

Getting on to the point, I've made it a habit to clear all my cookies and leave the browser window open on Main Page, when I leave for my next class. Nothing major, but I figure it's a harmless way to let people know we're out there. Been doing it for a few weeks, now.

So, today, I wandered over to the library after getting out of comsci early. Wound up getting the same machine. As I was logging in, I noticed the college's IP had some new messages, turned out it had been blocked after a little vandalism spree. Browser history was revealing.


Perhaps it's time for me to put more work into those talking points I've been musing over. It's amazing, the things some friends have told me when I casually drop that I'm active on the project.

I've done some tabling, before, usually for speech classes; it'd be interesting to really get out there to promote something I'm deeply committed to, in time and philosophy. Shame I'll be leaving the SF area, soon, or I might put more effort into getting involved with the fine group of Wikipedians out there. Time will tell what I find up north.

Tuesday, July 24, 2007

Notes on the 2007 board election

After the close of the 2007 elections for the Wikimedia Foundatin's Board of Trustees, it's time to congratulate Erik Moeller and Kat Walsh on their retained seats, and to welcome Frieda Brioschi to the board. See the results page for vote tallies and all that goodness.

My own tendency is to look back over the election, and try to evaluate what I know of it -- the lessons learned can and should be put toward improving future elections. Meta has an open request for comment on the matter, aside from the thread or two that've cropped up on foundation-l. There's a few issues I imagine should be taking center stage.

Gmaxwell's mass mailings to eligible voters at en.wikipedia proved to be controversial, but they also highlighted the inadequacy of previous, centralized efforts to get out the vote. Small groups of users from other projects were quickly able to adapt Greg's methods to send out mailings of their own. Just a few users, with access to the right resources and information, were able to have quite a hefty impact on voter turnout -- clearly, our goal should be to enable individual users to be bold.

To that end, and following a few discussions on IRC, I'd recommend a bottom-up effort over what currently seems to be top-down organization. Greg's effort demonstrated the effectiveness of small, tight-knit groups, the sorts of groups which already exist in ready supply on all of the WMF wikis. Meta could easily be used to provide a space where interested groups and individuals could make themselves known, to allow easy communication and coordination with their home projects. This "project contact" status needn't be an official position with any particular authority, but even knowing which people are interested in being directly contacted, how they can be contacted, who shares which languages, who will coordinate efforts, and the like, would seem to be pretty useful.

Set up each group with guidelines on neutrality, make sure they avoid endorsing particular candidate(s) over others, maybe even give them some basic source material, but on the whole, I imagine they'll be more productive if we just let them get to it.

I'm looking mainly at efforts to keep communities aware of elections and developments within them, there, but translation efforts are another concern. I wish it weren't so, but I don't have nearly the vocabulary I'd like, for German or Spanish, so it'd be difficult for me to help out. Some people have suggested hiring out some work to professionals, at least for the elections; I sincerely appreciate and encourage current efforts at Meta in this area, but I can't help but wonder what we might do to be more timely and complete. I can only imagine how left out I'd feel, if the foundation didn't do the majority of its business in English; my sympathies to those who have difficulty participating, due to language barriers.

As Dmcdevit pointed out, in #wikimedia, there's no reason to wait on some things. Sitenotices, emails, and all the like should really be produced ahead of time. The majority of these don't need to be produced in real-time, and doing so distracts from efforts to produce the things that do need to be produced on the fly. We can recycle prior notices, we can leave dates to be filled in later, but advance preparation is surely a must.

On a larger point, the timeframe of the election should be generally considered. Do we want more time between candidate selection and voting? How far in advance should the preparations and campaigns begin?

The use of approval voting seems to be an issue. It's been rightly pointed out that the race was incredibly close -- Oscar van Dillen lost his seat on the board by 20 votes, and was ahead of Michael Snow by only 5. Once a voter selects their "serious" candidates, they're faced with the option of supporting everyone they like equally (essentially, a null vote) or not supporting candidates they otherwise prefer. Some degree of preference voting might be interesting, but I do worry that some of the rather elaborate systems I've seen proposed would deter community participation. Even allowing people to allocate single-, double-, or triple-votes to candidates would allow for a great deal more flexibility, and allow for a more accurate measure of the community's will. I don't think I have the best answer, here, but it's definitely an area that should be discussed.

No doubt, there will be discussion of the "400 edits, 3 months" threshold, or the various other requirements for participation, but I think I'm content to let that one go, for now.

Beyond that, it might be great if people got past the trap of only caring about foundation issues at election time. But that'd be asking for quite the miracle, history tells me.

Wednesday, July 18, 2007

CharlotteWebb arbitration closed

A particular checkuser happens to run across a user while checking for Tor nodes. Not a problem. Checkuser reveals this private information on the wiki while user is running for adminship, user reacts poorly, candidacy is effectively torpedoed. Problem.

A lot of discussion followed in a lot of places. Personally, I do agree that there was no strict breach of the Wikimedia Foundation's privacy policy, nor the letter of the checkuser policy... but I'm also quite concerned that Jayjg seems to have refused to acknowledge that a large group of people have basis to be concerned or disappointed in his disclosure. A highly trusted user, recently involved a number of arguments regarding Tor blocking, revealed priviledged information at a very sensitive time, in a very sensitive place, and proceeded, essentially, to mock those who even suggested there might be a problem with this.

The lengthiest replies from Jayjg, at least at first, seemed to be on the wikien-l thread: "Jayjg: Abusing checkuser for political ends?" Apparently ignoring or rebuffing the users concerned by this, SlimVirgin and Jayjg preferred to respond with vagueness:
"What's to stop people from creating undetectable sockpuppet accounts using various proxies and anonymizers?" --Jayjg (wikien-l)

"We have issues with users running more than one admin account, and one of the ways they're allegedly doing this is by using open proxies. Being able to log an admin's real IP address is the only tiny bit of accountability the Foundation has regarding admins." --SlimVirgin (wikien-l)
Note the lack of any specific scenarios of that nature, or any direct response to the question, "Was it okay to reveal this information?" It seems they're assuming it was okay, and ignoring anyone who might think otherwise. Not that I can't sympathize with some of these concerns, or the desire to make decisions with the most information possible, but at the same time, I have to wonder if someone with such a cavalier attitude to releasing private, sensitive information should continue to have that sort of access.

Even on top of that, the association of "anonymizing proxies == suspicious and bad" has been questioned, and repeatedly, including this example:
"I think it would be fair to be particularly suspicious, not to give the benefit of the doubt, if the account smelled funny and it was also using anonymous proxies to edit. But using them in itself isn't an indication that someone wants to do harm, nor is someone intending to do harm much limited by the restriction." --Mindspillage (wikien-l)
As the Arbitration Committee became aware of the controversy, the eventually voted 7 to 0 to accept an arbitration case. I can only hope I'm not the only community member who hoped this might bring some resolution to the issues at hand. There's a huge number of statements from prominent Wikipedians, I'm not going to cover them in-depth at this time, but I do recommend taking a look at the comments from Geogre, JzG, John254, Chacor, Navou, ChrisO, and Rory096.

Arbitrator Paul August said, at the opening of the case, "There are many important issues here which might profit from an examination by this committee."

Nearly a full month later, the results are in. Looks like nothing is going to be done, after all -- at least not in the public eye. Let's have a look at the two remedies, both of which passed 7 to 0:
1) The Committee notes that CharlotteWebb remains a user in good standing, and is welcome to return to editing at any time.
Not even so much as a pat on the back. "Yep, you got screwed. You can come back, y'know, if you feel like it. But we don't really care, either way. Not our problem." I don't know if it's fair to expect a flowery message from something written by a committee, but I personally think we owe a little more than that, for a user who was recently forced out of the project when an unknown checkuser systematically blocked all of their IP addresses, proxy and non-proxy alike. Interesting this other checkuser isn't even mentioned in the results.

Even if I do feel this is a bit inadequate, I'm pleased they made sure to include something to this effect. The rest will have to be filled in by the community, I suppose.

On to the second remedy:
2) Jayjg is reminded to avoid generating drama by making public proclamations of misbehavior before attempting private discussion and resolution of the issue.
I guess that... sort of hints that he might have sorta kinda maybe done something just a little questionable and in bad taste. But there's nothing actionable here, nothing about what may happen the next time this comes up, and it's not even all that firmly worded. It's hard to even call this a slap on the wrist.

That all assumes there isn't a more serious exchange behind closed doors, and I could be mistaken on that count. I certainly hope I am.

Monday, July 9, 2007

San Francisco meetup

Funny thing, just a week or so ago I checked on the chances for any wikimeetup in the San Francisco Bay Area, and didn't see an impressive showing on the wiki -- looked like I'd have to push for it, myself, if I wanted anything to happen. But then today, on wikien-l, Eloquence mentioned he'll be stopping by this weekend, and suddenly things are falling into place.

Notice how things magically come together when a trustee is in town.

Yes, this is short notice, but if you're near SF and free this Saturday, you should consider making an appearance. See wikien-l for more coordination details.

Sunday, July 8, 2007

Drama lamp

With thanks to SPUI for the post title.

The Register ran a rather quaint piece about OTRS, the other day. En.wikipedia's article on the lava lamp was stubbed for about two weeks, pending legal concerns. According to writer Cade Metz and interviewee John Barberio, there was no explanation.

It's rather apparent Metz didn't put much effort into researching the issue, and I'm a little perplexed why Barberio, described as a seasoned editor of Wikipedia, wasn't able to point him in the right direction. With just one click, we can take ourselves to the article's talk page, where we see numerous references to trademark concerns, all made shortly after the OTRS action. Inside another minute, we can use the whatlinkshere function to find this thread on the admin noticeboards, which in turn links to numerous user talk page discussions, including this one on the acting volunteer's talk page. This has been discussed a few times on IRC, including on public channels, and most or all of those discussions quickly reached the conclusion that legal concerns were afoot.

Now, I don't expect the vast majority of web denizens to immediately find and process this information. But I would expect that a Wikipedian with two years of experience, or a reporter hoping to investigate their story, would be more than able. That neither of them bothered to notice they were swimming in a sea of explanations seems to indicate they couldn't be bothered to check.

We can debate all day whether the aforementioned legal concerns should have led to the blanking of the article. As I write this, the article has been restored, and the legal complaint has evaporated (I'm told they stopped responding, when it was pointed out that a search of the US trademark database yielded no results, and they were asked to provide specific evidence of their claim). OTRS actions can always be discussed, and other OTRS members can (and did) participate in these discussions, to great effect. At the time of the incident, the Wikimedia Foundation was lacking a designated legal counsel -- with no one to fall back on, for legal matters, can we really fault one volunteer for playing it safe?

Hopefully, with Mike Godwin hired on as counsel, this sort of thing will be less problematic, now.

I'm still at a loss to figure how this ever became an issue where "censorship" was a rallying cry, or why one possible overreaction, from one volunteer, at a bad time, in a project which has been running for several years, somehow constitutes doomsday.

In closing, I do think this has demonstrated a need for improved communication, within the community, and is one more downfall of the foundation's recent staffing issues -- both of which are quite serious issues -- but I don't know if we should take it as any more than that.

Thursday, July 5, 2007

Noticeboard bloat

Noticeboards are multiplying on en.wikipedia faster than I can count. There's the admin noticeboard, the incidents subpage, the 3RR subpage, the ArbCom enforcement noticeboard, the community sanction noticeboard, six sections of the Village Pump, a noticeboard or two for many WikiProjects, in addition to the WikiProject's talk page, active talk pages for several key policies and processes, including articles for deletion, images and media for deletion, miscellany for deletion, redirects for deletion, stub types for deletion, categories for deletion, and now even user categories for deletion -- all of these filled with active discussions and, more often than not, subpages.

This doesn't yet factor in the various abuse boards, including admin intervention against vandalism, ISP abuse reporting, long-term abuse profile pages, suspected sockpuppet cases, requests for checkuser, open proxy checking, and now a seperate usernames for attention page.

Don't forget dispute resolution! We have wikiquette alerts, requests for third opinions, requests for comment, straw polls, the mediation cabal, the mediation committee, the arbitration committee, and of course countless user and article talk pages.

This is before we even consider addressing the issue of mailing lists. wikiEN-l, foundation-l, wikipedia-l, announce-l, unblock-en-l, wikitech-l, and who knows how many others. Let's not even talk about IRC, for the moment.

This is all from memory, by the way. I'm probably forgetting a few.

If memory serves, the admin noticeboard's header template linked to about ten pages, when I first came to en.wikipedia. Since then, the conflict of interest noticeboard, the biographies of living persons noticeboard, the usernames for attention noticeboard, the community sanction noticeboard, the fair use galleries noticeboard, and now today the fringe theories noticeboard have all been added to that list. Only one has been closed, that I can recall, and there's sixteen linked, currently. That's a 60% increase in "central" noticeboards in about a year.

I get the feeling I'm really beginning to see what people mean, when they say it's difficult to get things to scale, at this level.