Sunday, September 21, 2008

WP:OFFICE and the Nielsen fiasco

It's been a fair while since I made a post, so maybe it's time I dusted off my keyboard and stepped to it.

Apparently the Wikimedia Foundation was recently contacted by Nielsen Media Research, regarding some information, copyrighted by Nielsen, which was posted on Wikipedia. By my current understanding, Mike Godwin was consulted and the material in question was deleted by foundation staff. As unfortunately sometimes happens with office actions, a small but chaotic tempest has ensued.

In a posting to the administrators' noticeboard, one user pointed out several templates had been deleted, apparently citing a DMCA takedown notice sent to Wikipedia's private OTRS system. Some users suggested such a notice should be forwarded to Chilling Effects; others insisted the message was private. Some users furiously debated whether the material in question was, or even could be, copyrighted. Others seemed lost simply trying to figure out what parts of which material were under debate. One user was reminded of last year's Lava Lamp incident. The story quickly made its way to Slashdot. Given the original lack of clear communication, comments from across the board have been getting increasingly chaotic and heated.

This is one of those cases where Wikipedia's usual "deny everything" approach has hit its limit.

How many times has this happened, before? In an information vacuum, the most sensational stories can gain traction they otherwise never would have. Key posts from Wikimedia staffers Cary Bass and Mike Godwin clarifying important details have been pretty prompt, considering this storm unfortunately happened over the weekend, but I worry they'll be lost in the crowd -- the sort of problem for which Wikipedia has no perfect solution.

Certainly we as a community should respect the office staff's needs, when they believe urgent action is necessary. In some regards, it is our respect for those needs that causes some of these difficulties: when a mass of editors, administrators, and OTRS volunteers enters the fray, either taking action to help or vocally interpreting the actions of others trying to help, it becomes less clear who did what, and who is allowed to do what moving forward. Users trying to resolve this confusion aren't helped along by the fact they will inevitably hold incomplete information, or by the lack of any easy way to distinguish important, clueful communication from mere noise.

This is just an idea, but I think there are a few problems we might avoid if we logged all office actions, preferably at a central location edited only by office staff. First, and most important, it would be eminently clear which actions are performed under the auspices of WP:OFFICE -- any action not specifically listed at the log, wasn't. Secondly, it would allow administrators to more easily enforce office actions, should doing so become necessary. Thirdly, it would allow staffers to -- if they choose to -- make prominent statements regarding office actions in a place where all interested parties will quickly be able to find them.

Such a log might attract troublemakers; I think the added clarity and reduced confusion would be well worth the cost, given the troubles we already face each time this sort of thing blows up. I believe an office log would be better for our own in-house communications, and possibly for our public relations as well.

Monday, July 21, 2008

#wikipedia-en: forking to ##wikipedia-signal

"I'm as mad as hell, and I'm not going to take this anymore!"

I can't be the only person frustrated by the recent levels of off-topic and frankly spammy conversation in #wikipedia-en on freenode. I've tried agitating for change, and have gotten nowhere; the hands-off approach taken by current operators is hardly helpful, here. Things have only been getting worse, and I've come to the point where I'm willing to fork the channel, if there's interest.

Tonight's topics of conversation in #wikipedia-en have included: womens' breasts, mens' breasts, lactating breasts, and other things that decidedly are not Wikipedia. Channel operators are often unwilling or unable to stop these trends, recently, and I'm sorry to say that what once was a useful channel is now even worse than #wikipedia ever was.

It used to be that #wikipedia was (mostly) off-topic and #wikipedia-en was (mostly) on-topic. Currently that's still the official story. Lately they're both off-topic, although surprisingly #wikipedia actually has the slightly more "intellectual" sort, where #-en is a more "frat" atmosphere. Don't get me wrong, I love social channels and generally favor a hands-off approach to moderation... but one on-topic channel hardly seems too much to ask for, and current chanops just aren't making this happen, it seems.

Wondering why I care? Just look at how effective #wikimedia-tech is. Currently I'm not aware of any public channel which regularly discusses the English Wikipedia with any degree of seriousness. Private channels are smaller, and less representative of the wider community.

With that in mind, I'm starting a ##wikipedia-signal channel for public, on-topic discussion of the English Wikipedia and its sister projects. Realistic concessions to social conversation will be made, but overall I'd prefer to keep a high signal-to-noise ratio (hence the name), and plan on backing that up with moderation as needed.

There are, however, a few options to consider:
  • Option 1: Good old culture and self-moderation. Allow anyone to join and speak, and encourage them to stay on-topic. Proactively kick or ban repeat problem users as needed, as other channels do (or should).
  • Option 2: Set the channel +i (invite-only), but keep trivial standards for access. Essentially an opt-in, semi-private channel. Repeat problem users could have their invite access removed. Easy to keep out misfits and botnets, but risks creating a barrier to entry, if users with the ability to access user flags are difficult to find.
  • Option 3: Set the channel +m (moderated), so that anyone can join but only "voiced" users may speak. Probably would set most regular users to be voiced by ChanServ on joining (+V flag), and might allow a smaller group to voice other users by request (+v flag). Repeat problem users could have their voice access removed. Similar advantages to Option 2, but creates a smaller barrier to entry if more users can be trusted with the ability to give others voice.
Currently I'm favoring the third course. Thoughts?

If the atmosphere in #wikipedia-en improves, so be it. This isn't a power play; I just want a decent forum for on-topic talk.


Update, 2008-07-25: Looks like ##wikipedia-signal hasn't taken hold. Thanks to the users who gave it a shot, and to those who offered feedback both positive and negative. However, it's worth noting that several concerned users have intervened in the affairs of #wikipedia-en since this post was made, and the situation there has improved. For now. Forking hasn't proved viable, currently, but it seems to have successfully drawn attention to the issue.

Thursday, July 17, 2008

BADSITES marches on

The loudest battle no one cares about.

Despite the community's clear and longstanding rejection of BADSITES as policy, some editors continue wield it or its cousins in debates, often seeking to discredit others in ways which may or may not make any apparent sense to the casual observer. The continued promotion of a siege mentality is, in my opinion, disruptive to progress on more important issues.

Yes, we should take steps to protect users who are harassed and victimized. When I'm able to, I try to do so. No, that does not mean I or any other Wikipedian should be obliged to support a divisive and jingoistic political agenda apparently based in anger and bitterness.

When users fail to unconditionally support a BADSITES siege, for whatever reason, they're often labelled "trolls." If site management doesn't support the siege, it's toted as evidence the site has been "taken over" by "troll enablers." For those who missed it, the line is quickly drawn between "us" (fighting evil BADSITES) and "them" (trolls and their sympathizers), with no room left for the subtle distinctions inherent in human relations.

I can understand why some people have come to view the situation that way, but they should know that they're also marginalizing their own position by painting with such wide brushes so often. Sorry to say, but not all users are just going to flop over and agree with everything a given editor says, especially when much of it is based in vague or secret evidence apparently discussed only in echo chambers.

All too often, the "them" in an "us vs. them" scenario does everyone involved a great disservice by playing along with the siege. The resulting squabbles can be quite discrediting to all involved. If we just saw a bit less of that, things might actually calm down. Sometimes the best response is no response.

For most of us, the project continues on without missing a beat. Remember that encyclopedia we're supposed to be working on?

Abuse on-wiki is actionable on-wiki. In extreme cases, off-wiki abuse with on-wiki effects may be actionable on-wiki. Absent evidence of such abuse, I fail to see how merely having an account on a given website could ever be considered inherently abusive and actionable on-wiki.

In short: let's talk about abuse, not siege.

Sunday, July 13, 2008

Wikimedia Foundation removes image of its headquarters from Wikipedia article

Image uploaded and added
Image speedily deleted and removed
Deletion review opened
Deletion review forcibly closed, citing WP:OFFICE

It may be worth noting that at the time of the DRV's closure, comments up to that point unanimously supported overturning the deletion. (ed: maybe not; see comments)

If the Wikimedia Foundation won't allow a picture of its headquarters to appear in the relevant article, so be it. I just wonder what justification there is that wouldn't likewise apply to headquarters images found in other articles, including Microsoft, Apple, or Craiglist.

It's not entirely clear that this is or is not an official office action, at the moment, but absent a demonstration to the contrary, the spectre of OFFICE has been invoked and will probably have its usual effects.


Edit 1, 2008-07-14: Added in-line note, referring to comment from Matthew Brown regarding comments at the DRV. See this post's comments for details.

Saturday, June 28, 2008

ArbCom fumbles the ball

Something is rotten in the state of Denmark.

This has been an interesting day on Wikipedia, to say the least.

First, arbitrator FT2 made a post to the admin noticeboard, mentioning that "A large amount of work has been in progress by the Arbitration Committee, in the background, for a considerable time, to look at a number of systemic problems deemed of importance, and possible solutions," before unveiling a hidden arbitration case where Orangemarlin has been sanctioned in absentia. FT2 also created a June 2008 announcements page, covering proposals including a new ban review group delegated by arbcom, activation of a new view-deleted-pages userright, the potential appointment of new checkusers, clarifications of the committee's role, an unfinished work on the BLP enforcement guidelines, and pages on old problems regarding skilled edit warriors and the consensus process.

This was sure to make some waves, all by itself. Kirill Lokshin, another arbitrator, made a post to the admin noticeboard several hours later:
The announcements made today by FT2 (including both the Orangemarlin issue and the various other matters) were posted without the approval or prior knowledge of the Committee as a whole. Further, no formal proceeding, secret or otherwise, has taken place regarding Orangemarlin or any other editor named in that particular statement.

As far as I'm concerned, these announcements have no authority or binding weight whatsoever.
Obviously, at least one of these fine gentlemen is wrong. After a brief attempt to archive the thread, pending news from the committee, the frenzied discussion moved to Wikipedia:Administrators' noticeboard/Orangemarlin and other matters. Given the unfortunate lack of concrete information, many confused and upset users immediately leapt into the fray with supposition and guesswork. Popular theories have supposed that one or both arbitrators have gone rogue, or that some effort is underway to destroy FT2's credibility. My theory is that we have a serious failure of communication.

This would not be the first time arbcom's penchant for private discussion and decision has been problematic. Just recently, arbitrator emeritus Raul654 filed a request to give Kingturtle checkuser access, on behalf of the committee; the issue was challenged, and then up in the air until FT2 offered some blunt clarification. It's been my experience that it's regularly difficult to tell whether an action is being undertaken by a single arbitrator, or the committee as a whole -- they sometimes seem to take advantage of this confusion. Often the only clear issues are those in which the committee has taken a formal vote on the wiki, and those are unfortunately quite rare; in many cases, it's difficult to tell even what the committee has or hasn't considered. It's a black box model at its worst.

Here, today, we have FT2 saying that there was a clear consensus supporting his announcement, and Kirill suggesting in reply that there was little or no discussion. It's hard not to be reminded or the arguments following Durova's block of !!, based on another private mailing list. I expect the committee will be discussing this extensively, but I find it disturbing that none of the other arbs have yet seen fit to make any prominent announcement to the effect of "We're looking at it, please stand by." No doubt they are currently unavailable, simply haven't thought to, or wish to avoid the storm of insistent messages any such statement would surely bring; it would take a certain strength of character to ask an angry mob to wait. Outside of FT2 and Kirill, do we have arbs of such character?

FloNight and Morven have both posted something to this effect, but I fear they'll be lost in the crowd. Still, they have my sincere thanks for saying something.

Obviously this is going to need clarification. We need word from the committee; silence simply will not do. Between FT2 and Kirill, the situation is chaos. In the short term, are these announcements valid? Are the decisions binding? What happened, here? In the long term, how can we avoid these problems in the future? Is there some way to make agendas or votes public, without necessarily revealing entire discussions? Do we need new policy or cultural reform, to deal with this?

If nothing else, communications breakdowns like this one can strike serious and unfortunate blows to the credibility of the committee and its members, and to order within the community. This confusion is damaging to the project. Perhaps it's past time that we as a community insisted on clarity from this committee regarding its decisions and operations.

Thursday, June 19, 2008

BLP special enforcement

Note: this was originally posted to wikien-l in this thread, but as I'm having technical difficulties (post keeps getting truncated), I'm posting it here and linking to it. For reference, it discusses the BLP special enforcement provision of the Footnoted quotes arbitration.

On Tue, Jun 17, 2008 at 8:26 AM, Sam Blacketer wrote:
It is headed 'special enforcement' rather than 'special policy', and I think the distinction is more than merely terminological. The policy basis is WP:BLP which has been in place for some time and has wide acceptance; I agree it would be wrong for Arbcom to change that policy.
From our own article on the subject: "A policy is a deliberate plan of action to guide decisions and achieve rational outcome(s)."

The distinction you mention is semantic and academic at best. We have here a committee which is not only manifestly decreeing policy in a way that toys with the philosophical foundations of this project, but even preparing itself to make content decisions, both things the committee was and is (supposedly) not supposed to do. I welcome you to make your case, but to pretend this isn't a big deal is an insult to the intelligence of the community that you represent -- it is a very big deal.

Administrator status is "not a big deal" precisely because we can self-police amongst ourselves. Reprehensible actions are less problematic on a wiki precisely because they can, in most cases, be quickly reversed. This ability of anyone to edit, more than anything, is what makes a wiki a wiki. Obviously, with blocking and protection, we have made necessary concessions to the reality that too much chaos can be damaging. As a community, we expect that consensus will show us a proper path, further reducing chaos. But here and now, we're providing an avenue by which adminship is absolutely a "a big deal" because any admin will, for any reason, be able to take any action of any kind against any user or article, without any need for petty things like consensus or discussion, and without any easy way to reverse that action. It takes that pesky "wiki" process out of the picture; so much easier to run the project when people can't easily edit, no?

We're even threatening dire consequences for any user, admin or not, who dares to challenge one of these actions without filling out the proper paperwork to request permission to do so. "Be bold!" we say -- just not if you're the second person on the scene. Specifically, I see a lot of mention that we should trust the good judgement of administrators, and yet this policy decree seems to do the exact opposite.

Where are the consequences for abuse or misuse of this power?

We as a community should be very careful when erecting systems which will inevitably become barriers to the creation of content and the free editing of the wiki, or which will enable users acting unilaterally to create such barriers arbitrarily and without easy oversight or accountability. We as individuals should be very careful to scrutinize the Arbitration Committee when it seems to overstep its bounds, as many users seem to think it has here.

The remedy is, as of yet, untested, and I believe predictions of impending doom are a bit overblown, but the fact remains: this is not a minor issue and should not be dismissed as one.

Monday, June 16, 2008

Metapedian, 2008-06-16

Board elections are open for another few days, if you hadn't been aware. You'll have until the end of June 21 to review candidates and submit any votes. Votes this year are being counted using the Schulze method, which allows voters to rank candidates according to their preferences -- a touch more robust than the approval voting used last year.

At the close of the "Footnoted quotes" arbitration, the Arbitration Committee has authorized administrators to apply new "special enforcement" provisions with regard to biographies of living persons, including the use of "any means at their disposal," and requiring appeals of any such action be directed to the admin noticeboard or the committee directly. Associated policy proposal is here. Many users are concerned by what seems to be a concentration of power, either into the hands of administrators or the committee.

On a somewhat similar note, the committee is currently voting 6-3 not to propose a "sourcing adjudication board" in the Homeopathy arbitration.

After a recent poll, the requirements to become an autoconfirmed user have changed: users will now need an account that is 4 days old with 10 edits. A new poll aims to increase the requirements to 7 days and 20 edits.

A new draft of the Wikimedia privacy policy is in the works at meta. Trimmed drafts are available on the talk page; other discussion is available there or on foundation-l.

The global sysops proposal (formerly Anti-Vandal Fighter) is currently up for polling at the metapub. Extensive discussion here.

Several users have begun an in-depth review of the requests for adminship process, and are currently seeking users to fill out a questionnaire.

An Ethnic and cultural conflicts noticeboard has been opened.