Because Wikipedia is set up to prevent ordinary users from banding together (on a Wikipedia area anyway)
to protest abusive actions by a given admin, its governance is opaque.
Since there's no easy way for an angry mob to form via posts on WP servers, the manipulators rule.

For an example of what happens on another website when an abusive admin goes berserk and starts banning/censoring,
consider the recent case of Reddit's Starcraft subsection.

Its administrators apparently all quit recently, leaving one guy -- Shade00a00 -- completely in charge.
A user objected to something Shade was doing. And Shade banned the guy, deleted his posts, and
went elsewhere on Reddit trying to silence the criticism.

All hell broke loose, and thousands of people started posting objections on various Reddit areas.
Shade couldn't delete them all, the criticism became overwhelming, and he was finally forced to quit.

Note: if you go thru some of those threads involving Shade's censorship, you'll find remarks indicating
that Shade's real name, address, and other personal information were posted on Reddit and on other
Starcraft user forums.

Where did they get his real name? Easy. He had a Wikipedia account using the same ID, Shade00a00.
With his real name posted. Tristan Dumas Bonnier, of Montreal, Canada.
What a dummy.

An IP address (in Canada) attempted to blank it, but it's still there. Strange, yes?

Also: a Flickr account.

I wish more user shitstorms, like this Reddit thing, happened on Wikimedia servers.
The "encyclopedia" might even be better. At least it would be more transparent.

His LinkedIn lists him as a "Subject Matter Expert". I wonder if his area of expertise is "fucking around on Reddit" or "StarCraft".

Either way, I think this is a good reason why using your real name on the Internet is a Bad Idea if you plan to go crazy.

There's little chance of this happening on Wikipedia--except for maybe a small, neglected wiki, in which case there'd be no point in taking over if no one cares. If one steward goes crazy, they'll just have a dev lock everything down or remove all the userright-manipulating powers until everything is sorted out. If a dev goes crazy, then Wikipedia's screwed. But that's unlikely. I wonder how much it would cost to bribe Brion Vibber into messing with one of the servers and breaking all the sites for a few days.

Sad that the only way to get rid of an abusive authority in an Internet community is to destroy his/her mind and break him/her down emotionally.

But if forming a lynch mob is the only way to get any sort of "justice," then that's what people are going to do.

These sorts of incidents should encourage communities to put term-limits on administrators and create processes to deal with misconduct. It certainly would cut down on administrator burnouts and freakouts. Most importantly, clear rules of what administrators/mods should or should not do. You know, a Magna Carta of sorts detailing the rights of the lowly editors and the administrators. It's a balance between maintaining order and peace on a site and catering to the community. I'm not excusing Shade's behavior, but it is a difficult job and no one is infallible.

Does Reddit have set rules for administrators or anything like that?

As far as I can see, there is nothing. They serve at the pleasure of the founders, and can only be removed by Ohanian or Huffman, or one of their designated sysops. There is no policy or system set up for controlling people like Shade. The only way to stop them is to raise hell, and embarrass them into quitting.

As for lowly users, Reddit has an utterly charming system set up. There's a long list of "guidelines", that are "optional". However......If you post anything a moderator or sysop doesn't like, even an innocent link to a personal blog, they assume you're spamming, and you become an invisible "nonperson". And if you want to appeal the "nonpersonhood", you can't.....they completely ignore you. The account still appears to work normally, but anything you attempt to post will be invisible to all except yourself. You can't even ask for help. Sounds like Kafka? Yep.

This will tell you more.

I didn't say anything about the actual community there, and how it "functions". Just be aware that any link submitted--no matter how good or useful--gets at least 25% downvotes, automatically. Because Reddit is a haven for trolls. Instead of running around deleting content and punishing people, as a Wikipedia admin would, a Reddit troll can only downvote things. And apparently a large percentage of them do it, automatically, just to be assholes......

Reddit couldn't be more different from Wikipedia, in most administrative aspects. Yet the end result is the same--good content, riding on a river of crap, with considerable non-transparency of process.....the one thing Reddit has, is more-or-less mob rule. Sometimes it works better than "proper governance", sometimes not. Wikipedia is carefully screwed down so that mobs are impossible, on top of obscured governance.

One of the oddities of WMF is that many of their sites manage happily with reconfirmation votes for admins, but it is deemed impossible on the flagship. I have never fathomed why, unless it is because some sites just have a more intelligent, mature, sophisticated "community".

I'm in two minds about whether admins should be time-limited so nobody can be an admin for more than say five years. It would get rid of a lot of bad admins, but also a few good ones.

Another solution is unbundling the administrator tools and dispensing them to editors based on certain benchmarks like seniority and writing FAs.

Then again, requiring administrators to write FAs would practically turn FAC into RFA and FAR into Admin Recall. FAR would become a revenge platform!

Featured Article Review:
FA Critic: "Delist. Misuse of en- and em- dashes. I see one comma splice. No alt-text in images."
Admin: "But, I'll lose my tools if this is delisted. Aren't dashes just fancy-words for hyphens?"
FA Critic: "If you do not know the difference between em- and en- dashes, YOU DO NOT NEED TO BE AN ADMINISTRATOR!"

All you would have to do to get rid of an administrator would be to demolish his FAs, accuse him of WP:OWN, provoke him into using harsh and threatening language, drag his FAs to FAR, ram through a RFC on him when he's really angry and unable to respond without cursing, and then off to ArbCom to destroy him mentally and emotionally and making him a shell of the man he once was.

Oy vey! No matter what solutions I think of to solve Wikipedia's problems, I can't think of one that doesn't present its own problems. Someone's mind and emotional state is always being destroyed!

Wrong. The Wikipedians believe that dashes are evil due to obsolete Apple browsers and keyboards. They were pushing for hyphens-only:

http://en.wikipedia.org/wiki/Wikipedia:Adm...en-dash_dispute

Kww: This leads us to an obvious conclusion: if a Wikipedia editor is expected to be able to type it, a hyphen should be used.

I wholeheartedly agree. Admins who aren't detail-oriented tend to have no idea what they're doing when they block someone after only skimming a situation.

I don't see why a straight up simple, no nonsense, 2 year time limit wouldn't work. No need for all this mumbo jumbo and linking admin tools to FAs or whatever.

The usual counterarguments are "but we have to few admins already" (really?) and "RFA is already brutal". The obvious point, that if there was a built in break on potential abuse with the term limits, then RFA wouldn't be such a nasty experience, and hence more people would go up for it (and probably get passed) and the number of admins would be fine (if not go up) is completely missed.

Term limits are a naive solution, that results in the loss of the most experienced. Combined with supermajority election, this would destroy the base of actually neutral and intelligent administrators, who are going to make enemies.

It would make the situation worse, not better.

What's needed is sane deliberative process. (Sanity, among other things, requires that it be efficient.)

I agree that sanity is a good thing. Who doesn't like sanity? Certainly not me. But I don't see what's so naive about term limits. It's pretty straight forward. If you like we can make the threshold for getting reconfirmed for incumbents lower, say 1/2 rather than 2/3.

I expect some "neutral and intelligent" administrators would get screwed, particularly if they made enemies. But a lot more "non-neutral and/or non-intelligent" would be gotten rid off too and I think the benefits would very much outweigh the costs. It's the basic reason why most elective systems do have them. In the few special cases where they do not exist or the terms are very long (SCOTUS, the Fed) a special case is made for their absence - and it is actually made, unlike Wikipedia, where the absence of term limits is the tyrannical "status quo" so that even bringing up the issue is controversial (it should be the other way around, if we had this "sanity" thing and all)

I think the major roadblock here is people trying to make the perfect the enemy of the good. The basic reason seems to be that if term limits cannot solve all the problems, or if they have some drawbacks, then somehow the present really crappy and insane status quo is automatically to be preferred. That's not how it should work.

(As an aside, given the usual length of your posts Abd, I can TOTALLY see you making the perfect the enemy of the good )

The problem is that people will oppose any administrator who has made a decision with which they disagree, while others would support that admin for making such a decision. It's like Democrats vs. Republicans in the U.S.--regardless of how politically brilliant someone is, if they cast one vote you don't like, they're out of the question.

An electoral college model, maybe? The community elects an electoral board to choose and de-sysop administrators?

The U.S. Electoral College was a brilliant invention that was undermined as soon as possible by a structural failure that the Founders probably did not anticipate, the rise of the party system. It was understood that the manner in which electors would be chosen was a problem, and they had no clue how to solve it, partly because of the slavery issue, perhaps, so they punted, the method of selecting electors was up to each state legislature. Thus the majority party in each state could set up a selection process that favored that party, and other states, if they were inclined to divide up electors more fairly, were effectively punished, by comparison.

Asset does, in fact, set up a kind of electoral college, but members consist of all those who have received votes in the election. In writing about asset, I've often referred to the body of electors created as an "electoral college." But it isn't a formal, decision-making body. In public Asset elections, however, the electors, collectively, represent *all* the voters (or almost all, there might need to be some tweaks for practicality, but this is substantially true, with a decent implementation). It would take a tome to detail all the considerations.

The body created would, I'd assume, because it could be quite large as to full membership, itself elect an Assembly, which would handle the election of officers serving the community, and which could, of course, remove those officers. The Assembly would be, for the first time, a body which would actually represent all the community, not just some faction. I would allow that Assembly to determine its own process and its own size, because, by the conditions of the problem, the "community" itself cannot do that.

Asset produces this effect, I expect: you would know who *your* representative was in the Assembly. That representative was elected with your vote. In an open system, this knowledge is shared in both directions. In a closed system (secret ballot input at the first stage, which I don't recommend for Wikipedia, but which would be necessary for Asset in public elections, I think), you still know whom you elected. They just can't verify that it was you.

ArbComm has a need to pretend unity. The Assembly would not. Because of the mission, I'd assume that a sane community, and the Assembly would tend to be saner than the community, by design, would want -- strongly -- to maximize consensus, but it would make actual decisions by majority vote, basic democratic process. (Which, by the way, is far more supermajority oriented, as to the standard processes, than most who don't understand it think.)

My opinion: they should change the MediaWiki software, to force users to vote in RFAs when they log in.
It would not be difficult to implement.

But of course, it would "mess up" their twisted little "community",
and take some power away from existing crackpot admins.

And worse, it would discourage people from generating all that free content they want soooo badly.
They must not do anything to discourage nutcases like Dr. Blofeld from making 500 useless stubs
per day.....that would be worse than keeping RFAs exclusive, and electing an occasional Ryulong.....

Wikipedia is a content monster, it lives on data (useful or useless). No one can check all the content
for quality, because the Monster must be fed.

It's a sausage factory, in more ways than one.

All of ya'll, Eric, AbD, are getting to complicated and unrealistic with this. Do you think that if just simple term limits have a next-to-zero chance of being implemented then requiring users to vote in RfA's on logging has a higher chance of being implemented?

Same goes for this idea of an electoral college. I mean yeah, at some point in the future and all but let's even get this basic aspect of how real life decision making is done through.

One of the major things that screws up Wikipedia is how much people will object to reasonable but not perfect improvements in its functioning simply because the kind of marginal improvement proposed doesn't solve all the problems at a stroke. So we wind up with extremely idiotic rules because the slightly-less-idiotic rules are not perfect.

Well, this might be reasonable: to be an autoconfirmed editor, you'd have to name a validated editor (i.e., already autoconfirmed) as your "proxy." That simply means someone you choose as being the other editor you most trust,who is also willing to accept the designation. You can change it at any time, so a new editor could simply pick an editor from a list of those willing to serve.

(Yes, to be effective, you'd have to be accepted by the editor. "Autoconfirmed" status depending on this is just an idea, don't take it as the core of the concept!)

And serving would, indeed, be a task, you would *not* want to "represent" a few thousand editors directly. Remember, many of them will be asking you questions or for assistance! Once you have a certain number, once the traffic generated is as much as you want to handle (or more!) you would decline most new requests, instead referring inquirers to people who are your "clients" already.

To gain real voting power, you'd need to represent many editors who themselves represent many editors....

This is delegable proxy theory, which I worked out in detail before I ever became a Wikipedia editor. DP, in theory, will create a bottom-up hierarchical structure, a communications network capable of filtering information, and thus of acting intelligently. Each node in the network is a filter, deciding what information to pass on and what information to ignore or only handle directly. Twere it up to me, I'd set up off-wiki communication between proxies and clients, but that's not an obligatory part of the system, it will merely make it bulletproof, highly resistant to on-wiki control.

(This is classic, it happens in nonprofits, frequently: A faction gains control of the central organization and then controls all the communication with members, if precautions have not been taken or don't accidentally exist.)

The objection can be made to every single possible idea, and that people believe these objections, without deep consideration, is why it is, given that limitation, impossible.

Instead of outright rejecting the idea, how about considering it as input to brainstorming, where there might be some aspect of it that has value?
What I've proposed could be implemented today, by those who cared to participate in it.

It's circular. It's impossible because nobody does it, and nobody does it because it's impossible.

A proposal was made in early 2008: WP:PRX. You can see what happened. Now, the secret: the system didn't happen, not because of the Rejected proposal tag, not because of the MfD -- which failed, there were actually two failed MfDs --, but simply because nobody but the proposer and me did it. And he was quickly banned, and was a marked editor from then on, he jaywalked, he was indeffed.

I proposed, on the Talk page, a simpler file format that was more flexible and less vulnerable to disruption. Very simple, you just create a file in your user space with a particular form. And then any "committee" can form with a list of members, and a proxy table can run off of that list. No changes in policy are required, at all. You just do it. The information generated is purely advisory. But I can guarantee that if the proxy assignments represent a real agreement to communicate, real power will be generated and will be visible.

That's feared, for sure, which is why I expected attempts to stop it. In my experience, *two users* who agree to communicate have enhanced their power, it's well worth the trouble. That happens without the delegable proxy system. The system would be on-wiki documentation of the connection, allowing certain benefits that would take a tome to describe. (Nothing would or could prevent secret communication outside the declared system.)
Yes. And there is no decision-making system. There are systems, multiple ones, competing and conflicting, which create the appearance of decisions, with an appearance of consensus, all of which is not reliable.

It's reasonable in every way except for it actually having a chance in hell of happening. What's the point of discussing of what a Utopia will look like when you can't even make small changes to improve actual practice?

Actually, scratch that, it's not reasonable in every other way. It'd be like requiring people to vote for representatives when they show up to work in the morning. Sometimes I want to exercise my right not to vote, particularly in cases where I'm not familiar with the candidates. All you'd get in that case is a bunch of people voting randomly just so they can get on with the task of doing what it is they want to do when they log into Wikipedia (write articles, etc.) At best you'd wind up with random decisions but more likely you'd get the Reddit problem Eric's complaining about where most people just default to "No"

You guys are still talking about how to polish a turd. The problem, which you don't seem to accept, is that the inmates run the asylum. Talking about various ways of making the trustees more accountable and represent larger constituencies, or of making the inmates vote more regularly ignores the basic fact that 99% of Wikipedians are pure idiots whose primary motivation is not "a better encyclopedia".

If, on the other hand, an online encyclopedia was structured that could tolerate a certain amount of "user-generated" content and user feedback on articles, but that also had real editorial power vested in (paid) subject-matter experts, you would generate better encyclopedic content with little or no drama or overhead.

Of course, no one wants this, because it blows up the "community", which is vastly more important to Wikipidiots than the random articles over which they so vigorously tussle. What makes the post-apocalyptic warlord society of Wikipedia work, with its arbitrary admins, its willing servants, its show trials, its banned enemies, and everything else, is the fact that everyone who spends more than a little time there does so out of vanity and the quest for power within that society. If you made it work simply, effectively, and efficiently, you'd take the fun out of it for most of those morons.

Not really. The idea of admin term limits is a lot more practical/feasible than the idea of an electoral college on wikipedia. Some ideas ARE intrinsically more plausible than others.



Sure, as brainstorming it's fine. But we're nowhere near where it's a possibility and that needs to be recognized.



Again, not really. It's impossible because we haven't gotten to the point where the functioning of the social system is sophisticated/robust enough to even consider such a possibility. A simpler solution is much more likely.



I think that actually supports my contention that a lot of this is just daydreaming. But the ability to daydream should not be an obstacle to actually make improvements. That's usually when you got to slap somebody, at least in the movies, to get them back to focus on the here and now.




Sure. But what does this have to do with the specific issues at hand?

I think both of you and Eric are committing a fallacy, common to technologically minded folks, that just because something is technologically/technically feasible, it IS in fact feasible. But that's not the case. Social/behavioral feasibility matters just as much if not more (humanity could have had democracy, private property rights, pursuit of science, equality among men and women etc. back in... well, back when we were a bunch of Post-monkeys running around the savannah. But it didn't, because social constraints prevented those things, even though they were technologically possible (all it would have taken is for the Post-monkeys to agree on such things)). It is possible for us to redistribute income so that everyone in the world gets about 7k $ per year and we have perfect equality (about half the US poverty line incidentally, for an individual). It ain't gonna happen. It is technologically possible for us to terra form Mars and start some limited human colonies there. But it ain't gonna happen.

This is the difference between terra-forming Mars (great, but it won't happen) and sending a unmanned probe to Mars (not so great but still good and feasible), or the difference between completely redistributing world income to make it perfectly equal or to just alleviating present inequalities to the extent it can be done. Social and behavioral constraints do matter.

You couldn't make a "small change" by naming a proxy with a proxy file? Well, if you are blocked, perhaps. But otherwise?

There is a point to discussing improved systems; what I've described is not a Utopia, it is what I see as a natural consequence of a fairly simple structural change -- the initial changes are extremely simple, and then facilitate the next steps as possible and natural. Obviously, I cannot predict beyond the first step or two, because I'm talking about designing an intelligent, self-correcting system, which will sooner or later deviate from my concepts.

My theory indicates that, if it does so, it is likely to deviate in the direction of improvement over my ideas.
You missed it entirely, Radek.

The original proposal, yes, was preposterous as stated, but the *concept* wasn't preposterous. What I did was to set up a reward (autoconfirmed status) as a consequence of taking a one-time action (naming a proxy, which you might as well call a mentor, the function could be very similar, if you'll exercise that brain a bit.)
I'd call this a poor imagination. While early decisions might be more or less random, they would not stay that way, not once the communication was functioning.

This is a bit like complaining about the possibility of human intelligence, because babies don't know much.

This is a TANSTAAFL system: if you get lousy service from your proxy, you'll pick someone else, and if your client is a pain in the rump, you'll drop your consent. The whole point is to pressurize toward functional communication. Exactly how to do that would involve implementation details. Suffice it to consider, for the moment, that it's possible. One hint: this is how functional societies already work, it just isn't documented. Documenting it allows the overall process to become more reliable, that's all.

But just for starters, such a documentation structure makes, as I've mentioned, a crackerjack proportional representation election system. It could be done with anonymous voting, as a start. So, say, ArbComm would be elected by all those who have received votes in the election system that already exists. If you wanted to elect N arbitrators, and if people vote for just one (!), and there are V voters, you might decide to consider an arb elected if they get V/N votes. To be safe, you might make N a bit higher than the number you actually want to elect. (This risks electing the higher number. Big deal, it would represent amazingly complete cooperation!)

Some arbs might be elected directly, but other candidates would be holding votes. They can "spend" those votes to create seats. If they don't want to compromise and they don't get any votes from anyone else, tough. TANSTAAFL. If that was my vote, I'd not be thrilled! Next time, I'd vote for someone else. That's why this is better if it's done by a standing election, a file placement..... I could, mid-process, shift my vote if I didn't agree with what my candidate was doing....

(Votes used for election, perhaps, would not be revocable. There are practical details I won't describe. A mature system, I believe, would *continuously maintain representation*, it's not difficult to do that, but it does require more steps, more understanding, and, probably, more experience with the process and how it *actually* works. The trick in the mature system is that voting power of representatives would be "diluted" in some cases. That allows direct voting by any "elector," i.e., someone who voted publicly. It's hybrid direct/representative democracy, where deliberation rights are confined to designated representatives, but voting rights remain with the sovereign, the people.)

Wikipedia has already used some advanced election methods, so it's not impossible, in principle. In practice, what I'd say is impossible is predicting the future with certainty, yet some of us do that routinely.

By the way, I'm assuming some familiarity with STV methods. To make this clear: suppose the quota for election is Q, and you get N votes, where N is greater than Q, and you decide to accept the seat. You would have N-Q votes left to distribute as you choose.

At an extreme, suppose everyone has a single person as their proxy. That person would be able to unilaterally choose the winners. Not likely to happen, eh!, but this is part of the reason why I'd rather see standing designations, in public files, with accepting a proxy being a consent to direct communication. It would slightly pressurize toward distribution of the assignments, with a very popular person being at the top of a hierarchy with independent persons, who themselves are chosen by others and who develop relationships with them, being likely to be chosen. The very popular person still gets elected, but depends for his or her power on a penumbra of maintained support. The "faction"-- we call this a "natural caucus" in FA/DP theory -- gets representation on the Assembly or whatever is being elected, according to its actual trust level in the overall voting community.

Utopian? No. Possible. It's *not* perfect, merely a hell of a lot better -- in theory -- than standard techniques. In practice? How would we know until it is tried?

I'm only aware of one trial, where the Election Science Foundation elected a 3-person steering committee from 5 candidates, with 17 voters. The result? Quite unexpected, actually. The candidate in second position as to popularity decided to withdraw and assign his votes to create the third seat, electing a candidate who only had a couple of direct votes. In the end, every voter's vote ended up being either for a winner (most votes) or for the choice of the person the voter had voted for.

In other words, the committee was perfectly and completely representative. I'm not aware of any other election method that could have done that, from a single ballot and a few days of discussion. The trick is that it isn't a simple election method, it's quasi-deliberative, it's not deterministic from the ballots alone. Arrow's theorem wouldn't even consider this a voting system.

I think that the fact that I'm having trouble even reading all of this, much less comprehending it, is a pretty good indicator that it's too messy to be even considered. This just seems to much like some kid playing at "if I ruled the world I would..." Rules and procedures need to be is simple. Maybe we have a different definition of "simple" here but I'm pretty sure my definition is closer to what can be expected from the average person than yours.

Sure. Now, is that inevitable? The present system does make it so, I'd claim. However, I've actually been in an "asylum," Gomi. And inmates can be saner than you might think, given functional social structure.
99% of "Wikipedians"? Who are they? Do you mean that 99% of edits are made by "idiots"? Frankly, the quality is better than that. Yeah, it's unreliable. And the structure is horribly inefficient such that to remain, one has to be insane in some way. But ....
Maybe. So, Gomi, why don't you do it?

Look, this is the reality. There is already a "certain amount" of paid content. Wikipedia could move toward the system you describe by actually hiring "subject-matter experts." Why not? Here is the problem, Gomi: how would this -- or any other structural change -- be decided? Who would choose the experts?

What I'm claiming is that the core problem is one of how to make decisions in a large organization that depends on volunteers. Those volunteers have, say, some access to funding, or, in the present model, they can recommend board members, who are then, perhaps, named by the existing board, it's a common model. The board could decide to hire experts to review articles. Why not?

The present problem, I'd suggest, would be the reaction of the supposed "core community" to this! That core does not represent the overall community and, for sure, does not represent the *readers.*
Right. It's a standard problem of nonprofits with a large purpose that depends on volunteers. The volunteers end up running it, which has good features and not-good features. How to keep the good and drop what isn't functional? That's the problem.
Yup. Now what?

It's easy to blame it on Those Morons. But, this question exercises me: what stops non-morons from being more effective? Maybe there aren't any "non-morons"? Or is it something else?

I think the reason is that the people who can see something better give up, don't believe it's possible, are afflicted by despair, and don't support each other, thus guaranteeing failure.

When I had the active support of other editors, even though they were in the minority, I was quite successful. When I was perceived as having become, effectively, a single-purpose account dedicated to a fringe idea, when I was involved, I was dead meat.

Poor process, in fact. Had there been other editors protecting me as I had protected so many, it would not have turned out this way. I had some strong supporters. They burned out on Wikipedia, entirely.

Whatchu said.

That's what Wikipedia is -- a "fun" place to act like an asshole.
It's essentially a drug, a big syringe full of digital heroin for OCD control junkies.

So I might as well demand unreasonable reforms--there will never be ANY reforms, to be bloody
minded about the whole thing. It will go down before it changes. Junkies trying to kick often die,
from suicide or simply because they give up on life.

Now, if you cleaned house at the WMF office, and installed honest people who actually want to make an
educational resource, it might have a chance. Care to take bets on the likelihood of that happening?

I do worry that with confirmation votes the good admins might actually fare worse than the bad ones because the latter may have more friends prepared to turn out and vote.


We already have essentially that for Checkusers and oversighters, namely ArbCom. Would we trust them to handla all admins?

Sums it up well. How many editors care about the encyclopaedia vs their pet articles/POV, however obscure? Think FAC.

The bad ones would also have more (well deserved) enemies I expect. The only danger I see, though this is pretty much unavoidable and it plagues Wikipedia across topics, is that admins will show up to support reconfirmation of fellow admins out of class-consciousness (just because they're admins) rather than on the merits.

Problem #1 with the electoral college idea is that there's only a dozen or so checkusers and oversighters, but a lot more admins. Not every position has to be elected/filled in the same way (in fact they shouldn't be).

It's impossible to care about the entire encyclopedia--there's simply too much crap.

The problem with the current Wikipedia is the reaction of the "core community" to anything. No meaningful reform of Biography articles has made it past the "community". The "Pending Changes" half-measure didn't make it past the "community". The "community" bans members that ArbCom won't -- when they threaten the stability of the "community". The "community" will oppose anything other than the status quo on Wikipedia, and in doing so the ones who stick it out will continue to become more powerful. Those who don't wish to play the game will leave, and the current smell of a stagnant pond will become more evident to casual observers.

If you want a more plausible plan to fix Wikipedia from within (although still abso-fucking-lutely guaranteed not to happen), here it is:

1. Empower the 'Wiki-projects' - If you actually believe in the Wiki-way, you need groups of fewer than 100 participants -- and ideally 30-50 active ones -- to focus on a content domain of tractable size. (Communities without hierarchical structure that grow larger than 100 to 150 members tend to fail in the way Wikipedia has). The existing "Wiki Projects" have membership, and some of them have something resembling leadership. Give them real control over whatever pages they believe are in their scope.
2. Revise 'Pending Changes' - Revise the Pending Changes tool so that every page within the scope of more than one Wiki-Project is unreleased until all Wiki-Projects approve it.
3. Limit admins to a single Wiki-project - Rejigger admin rights so that admins can prohibit edits by certain users from their Wiki-project articles, but not from everywhere.
4. Create a content dispute "court" - For the inevitable disputes between Wiki-Projects, deputize senior members of unrelated Wiki-Projects to sit as a jury on content disputes, the results of which are binding.

This scheme would have the effect of decentralizing Wikipedia and its community. Sure, there would still be battlegrounds, but they would be more localized, and the damage wouldn't be reflected out into the real world so much. In many ways, it would create several hundred little Wikipedae that look more like the Wikipedia of 2004 or so. Again, because this dilutes the power of the entrenched Wiki-potentates, this will never happen.

There is a ton of quite good research about the growth, success, and failure of online communities (and communities in general). Wikipedia has no interest in it at all. It does not contain the seeds of change, and so, like the dinosaurs and other creatures with small brains and little ability to adapt, it will die. The only question is what form the meteor that changes the environment will be.

I actually quite like all of these ideas. The empower Wiki projects one in particular though here the implementation can be messy. Quite likely you might see turf wars develop. If there's trouble - as there's likely to be - on anything with both a WikiProject:Poland and WikiProject:Germany tags on it - how do you reconcile the competing claims? You need some kind of a meta system for interaction between projects.

Another, more sanguine problem, is that a lot of these projects have very few active participants (in fact, often it's just one or two people doing stuff). In some cases that makes perfect sense - for example WikiProject Bund set out to achieve a particular goal, it achieved it (more or less) and has been quiet since then. On the other hand nothing ever happens at WikiProject Mexico even though there's a TON of work to do.

I've lost track of how many times I've argued for this. I even once suggested that the existing bureaucrats should simply arrogate this authority to themselves and made promotion decisions on the merits of the arguments made at RFA instead of the raw numerical vote. A few of the b'crats seemed to think that this would be a good idea but that they couldn't actually do that because it would have violated the community's trust in them, or some other such nonsense.

Whats the point of fixing Wikipeidia. To fix Wikipedia is like trying polish a long greasy turd. Wikipedia can't be reformed, the powers that control it don't want it and don't care.

The only way to reform Wikipedia is from the outside when a BIG FAT LAWSUIT lands at the door step of Wikimedia Foundation, Inc. I have, as a public service provide said address for the convenience of those who wish to serve said lawsuit to the Wikimedia foundation.

Having been on the Review now for about five years, I have heard this statement from some blowhard or another -- usually a blowhard without the financial means to pursue a lawsuit or an understanding of the complexity -- every few weeks, perhaps 100 times. It's uninspired and boring.

And WikiProject:SPAM shall have rule over all of Wikipedia. Based on the sorts of knowledge that they're calling "spam" now, that should be the equivalent of a "content ArbCom". It would be the most powerful WikiProject, and they even have a battleship.