IMHO, it should be easier to get adminship as a) it's No Big Deal(r) and b) all actions can now be reversed - and users can quickly be dysopped. The three month tenure, however, I find silly - if someone was good enough and behaved well for those three months then obviously they should be allowed to continue. Desysopping after three months would turn the whole RfA process into a lottery-like game - "if it's no big deal, why can't I continue as an admin?". I also disagree with not allowing users to object to more than one RfA every three months - Wikipedia is not a democracy. On 2/9/07, Timwi <timwi(a)gmx.net> wrote: -- Gary Kirk

On 2/10/07, MacGyverMagic/Mgm <macgyvermagic(a)gmail.com> wrote: Many newbie admins readily admit they don't understand policy or how to perform most of their duties. They start slowly, and watch other admins for guidance. How is this harmful? Steve

MacGyverMagic/Mgm wrote: Where's your evidence for that last sentence? As far as I'm aware, this theory has never been put to practice, so I don't see how you can make such a claim.

Thomas Dalton wrote: I didn't contest any of that. I contested the (fallacious) conclusion that "it doesn't work in practice". Everyone thought that about editing too, until Wikipedia proved everyone wrong. Timwi

On 2/12/07, Thomas Dalton <thomas.dalton(a)gmail.com> wrote: Actually, no. You need more positive contributors than negative contributors. Giving the positive contributors more power makes their work easier, but it's not necessary.

Thomas Dalton wrote: This is a false assumption. It may be true in some cases, and it may well be the case that the admin features in MediaWiki happen to be such cases. But it is not true in general, and MediaWiki is still in active development. Well, I don't know about other proposals, but mine doesn't. Quite to the contrary, I am proposing to have several levels of blocking. The first is that which is currently called "de-adminning". Timwi

Thomas Dalton wrote: Ah, but fortunately we have anti-vandal-bots, too, automating much of that work. If necessary we could run a lot more of them, tuned to catch a lot more vandalism, but it's not a big enough deal (yet, anyway). -mark

Thomas Dalton wrote: Yes, exactly. This is wrong. Wikipedia was working right from the start, even before adminship existed. You're also forgetting that Wikipedia is not the only wiki, and that MediaWiki is not the only wiki engine. Most other wiki engines do not have an "admin" user level at all, and last I checked the Pattern Repository was still working just fine. Timwi

On 2/12/07, Thomas Dalton <thomas.dalton(a)gmail.com> wrote: Don't need anything particularly drastic to prove it. Any mediawiki installation that isn't locked down or supported by a strong admin presence will be spammed into oblivion in short order. See: http://www.wikiindex.org/index.php?title=Category:Spammed -- geni

Thomas Dalton wrote: Is this true? I thought the primary way Wikipedia dealt with vandalism is by treating it as just another edit, the bad ones of which are more likely to be nixed by subsequent edits. My personal experience is that a tiny percentage of vandalism requires more power to solve. Is that not the case? Thanks, William

Thomas Dalton wrote: Extrapolation is indeed the logic of the simple. Had we depended entirely on extrapolating the visible world of Newtonian physics, Einstein's relativity might never have been discovered. Wow! Do you understand what you said? You have managed to undermine the entire 17th century debate between rationalism and empiricism. =-O Ec

Thomas Dalton wrote: This is just plain wrong. It is true that Wikipedia is not an experiment in a lot of things (incl. online democracy, anarchy, etc.), but Wikipedia *is* and has always been an experiment to create an encyclopedia using a wiki. A lot of things have changed in Wikipedia because people realised at some point that certain aspects didn't work very well (e.g. the CamcelCase convention was removed because people eventually realised that having articles titled "DemocracY" and "TelephonE" is stupid). There is no reason to believe that Wikipedia has now reached a perfect state and requires no further changes to the way it works. Timwi

Luna wrote: You are right; however, it *does* refute the previous argument that tried to justify that it "should not" be implemented. The reason it should be implemented is because nobody can currently know whether it will work better than the current situation or not because no-one's tried it before. If it doesn't work out, we can still go back to the old system, but then we'll at least *know* that it didn't work and can address whatever problems came up. I have given several ideas why I think it might work, and why the proposal addresses existing problems. Surely this is the same old argument that also says that wikis can't work because anyone could rack up 20-30 socks and keep reverting their favourite articles to their favourite biased version. If you let more people be admins, there *will be* enough admins around to block 20-30 misbehaving accounts. Notice that if one of them gets blocked (desysopped) for good reason, one of the socks resysopping him would also qualify as misbehaving; eventually all of the socks will be blocked (desysopped). Vandals will soon realise that it is not worth it, and after a short initial rage, people will stop even attempting to create 30 sockpuppet accounts. Timwi

geni wrote: Not sure what you're talking about there, "without admin backup"? Too late for what? Yes, precisely. Timwi

geni wrote: Why are you talking about a situation without admins? I don't know where you got this from. OK, maybe you don't know this yet, but all changes in MediaWiki can be undone. There is no such thing as permanent damage. The irony lies in the fact that your ironically-meant statement is actually true. (Well, almost. Of course there are negligible numbers of people who are truly persistent. But on the whole, very close to 100% of all vandals give up.) Timwi

Thomas Dalton wrote: So the wiki idea is effectively an idea to remove editors? Your creativity is taking wings! No-one so far has talked about bad press or hurt feelings. In particular, no-one so far has suggested that my proposal would increase the incidence of such. Much less has anyone provided any *actual evidence* that it would -- it might just as well reduce it (and I have already provided arguments why it might). But your innovative invention of specious arguments is truly remarkable, I must say! Timwi

Thomas Dalton wrote: That is correct. That is very different, however, from saying "There are [effectively] no admins". That would be like saying "On a wiki, there are [effectively] no editors". That's just absurd. Timwi

Thomas Dalton wrote: If you're so hostile to wikis why do you waste your time with them? Ec

On 2/16/07, Timwi <timwi(a)gmx.net> wrote: It's the closest analogue of an admin vs admin conflict available that you could really understand. There are attacks that would take a lot more effort to undo than to do. I could put the goatse on every page tubgirl on every watch list and lemon part on the main page. Throw in some stuff from the pain series on the block page as a little surprise for whoever tried to stop me. WOW? Communism Vandal? Mr treason? General Tojo? Squidwad? how many of these would you be happy to have their hands on admin powers? -- geni

Timwi wrote: That's why I suggested a random sample of 100 - subject only to some fairly low level conditions that indicate a willingness to stick around, and no obvious signs of dementia. Track that 100 for three months to see what happens. Also track the 100 most recent RfA appointments and compare the statistics for the two groups. Why should anybody object to controlled results? The idea that there could be 30 sockpuppets in a random sample of 100 users is out of touch with reality. Ec

Timwi wrote: Are you suggesting no conditions at all? What I'm suggesting is really low level stuff. If somebody stuck around for 2 or 3 daya a month ago, and maybe put together a good series of edits, but we haven't seen him since what's the point of making him an admin if we're not going to see him around again. I'd like to see the whole system liberalized so that it works. I'm sure willing to allow a few easy conditions for the sake of having it happen. That's a lot more important than making some ideological point. Ec

Observation is not the same as experimentation. Experimenting requires doing something and then observing the results. Observing what happens in the real world and using that as a basis for out decisions is always a good idea, I agree, but experimenting can do more harm than good. In science, experimenting is always the best way to reach a decision, since the purpose is simply to gain knowledge. We are not here to learn how to run Wikipedia, we're here to actually run it, which means we have to accept that we can't always learn everything we want to.

If Wikipedia were a type of science then experiments would be the only way to find out the best way of operating. However, we are not a type of science, our aim is to operate as well as possible, not just to find out what the best way is. That means doing the experiment could be extremely damaging - this isn't the kind of proposal that can be tested on a test wiki, it must be tested on the real thing, and it could be horribly wrong. In situations were it isn't practical to test a proposition, we have to use logical arguments to work out which is best without actually doing it. It's not ideal, but in some cases it's the only way of finding out without the risk of it going wrong - I feel this is one of those cases.

That's your hypothesis. As it hasn't been tested, it's not possible to say whether you are correct. On 2/9/07, MacGyverMagic/Mgm <macgyvermagic(a)gmail.com> wrote:

In message <C1F24F19.16BF%michaeldavid86(a)comcast.net>et>, Marc Riddell <michaeldavid86-Wuw85uim5zDR7s880joybQ(a)public.gmane.org> writes It's an encyclopaedia. Everything else is purely coincidental. -- Arwel Parry

Gary Kirk wrote: You may even be right about what happens after a three month tenure. In that case we would be able to drop the probationary provision. Choosing a first random batch of 100, and reporting after 3 months about what happened to this group would support the radical concept of basing later choices on facts. Thus: 1. How many refused the nomination outright? (This group could be immediately replaced by new nominations.) 2. How many went nuts? 3. How many stopped editing within 1 month and within 2 months? 4. How many continued editing without using admin powers? 5. Which admin powers did they use? 6. other questions? Ec

Thomas Dalton wrote: A question that cannot be answered now, or as a result of this test: How many vandals exist NOW because of the ludicrous way that admins are chosen, the fundamental discrepancy between an "anyone can edit" encyclopedia and a "only zealots need apply" adminship? Any suggestions on how to get a handle on that number? -Rich

Basically, you are suggest removing the concept of administrators entirely, giving everyone access to all the tools, and then introducing a new kind of block that take some tools away from users that do something wrong. The system we use for editing pages (the one you think should be extended to admin powers) results in an enormous amount of vandalism, which takes a lot of effort to clean up. It's considered worth it, so we just shut up and get on with it. Using the same system with admin powers would result in an equal amount of admin vandalism, which takes more effort to clean up, does more damage, and is not at all worth it. Also, if adminship is opened up that much, we would have to introduce a new level of admin anyway to handle blocks (which would actually be desysopings, as everyone would be able to unblock themselves), and then all the problems of RfA just get moved to Requests for Super-adminship, so we have absolutely no gain at all.

Luna wrote: How is (a) related to your concerns about "random vandalism"? -Rich

Thomas Dalton wrote: Essentially yes, except that I'm not entirely opposed to the idea that the admin tools would not be available immediately (just like the "autoconfirmed" state is not granted immediately). My only gripe is that the process is not automatic, i.e. adminship is decided by people, and as a result people are able to (and do) put ridiculous criteria in place, thereby actively stopping most people from ever getting a chance to play with the tools. I don't deny that some changes to the software would be nice that would make reverting a user's actions easier. Until we have those, I would agree that admin vandalism is difficult to clean up, but it's not an insurmountable problem. Remember we used to have huge problems with page-move vandalism because it was difficult to revert? So now we have new features to revert it more easily. The same will happen to reverting admin actions or mass actions if we need it badly enough. You have not provided any evidence for that. There are two flaws in your logic. First, if your adminship bit is removed, how do you "unblock" yourself? So no, you don't need an extra "super-admin" level to do de-sysoppings, you can just allow admins to do it. Second, your argument would equally apply to page deletion (you can publish any crap because if anyone deletes it you can just undelete it!), and yet admins can all delete *and* undelete and it still works (because they largely follow a certain policy). Timwi

Thomas Dalton wrote: See, this last conclusion of yours is a fallacy. All you (strictly) need to prevent this (very hypothetical) scenario you described is either enough admins so that there's always someone around to stop such a bot quickly enough, or a cap on the maximum number of blocks (i.e. desysoppings) that any one user can do in a fixed amount of time (let's say, one per hour). Timwi

Thomas Dalton wrote: If everyone can edit pages, then a vandal with a handful of sockpuppets can do constant vandalism, always reinstating any of his edits that are reverted. Nice, you're refuting yourself :-) It doesn't become any truer the more often you say it. Timwi

Thomas Dalton wrote: What's the harm in NOT allowing admins to unblock themselves? -Rich

I'm not sure really. I'm not sure there would be much gain, though - all it requires is a sockpuppet to do the unblocking. If it's easy for a vandal to get one admin account, it's easy enough for them to get multiple.

On 2/19/07, habj <sweetadelaide(a)gmail.com> wrote: Yes. Would take hours/days of solid work in the case of long histories. -- geni

Cheney Shill wrote: I'm sorry, I don't understand how. Well yes... who else do you think it is applied to? Clearly not the admins :-p I honestly don't see what NPOV has to do with anything in this thread. What you're saying is not making much sense to me I'm afraid. Timwi

On 2/10/07, Erik Moeller <erik(a)wikimedia.org> wrote: Agreed, but the regular voters at RFA are very insistent about their right to set their own parameters for what makes a good admin. It gives them a feeling of power, I guess. Steve