On 12/5/05, Walter van Kalken <walter(a)vankalken.net> wrote: Since a few hours ago. It is more of a test than anything else. Brion implemented it today on Jimbo's request a few hours ago, so that some kind of results can be obtained. There is certainly not consensus support for it in the community, though a large faction certainly do agree. I think this is "wait and see" time. -- Sam

On Mon, Dec 05, 2005 at 02:27:21PM -0800, Daniel Mayer wrote: Correlation!=Causation. Because this concentrates on a particular symptom of vandalism -as opposed to the actual cause- I'm expecting new pages patrol to become more difficult as the easy telltales are now hidden. Let's see if that prediction comes out. Anyone care to place bets? :-) read you soon, Kim Bruning -- [Non-pgp mail clients may show pgp-signature as attachment] gpg (www.gnupg.org) Fingerprint for key FEF9DD72 5ED6 E215 73EE AD84 E03A 01C5 94AC 7B0E FEF9 DD72

On Mon, Dec 05, 2005 at 02:27:21PM -0800, Daniel Mayer wrote: ... ... A. Creating new crap article now consists of typing the name of the article, typing the article + clicking one button. B. Creating new article with compulsory registration adds aditional requirement of creating one crap account. Necessary effort consists of typing a random user name, 2x random password + clicking one button. C. Creating new article with compulsory registration with unique email adds requirement to crate one crap email. Necessary effort consists of typing a normal email into jetable.org, and clicking one button. and so on.... In long term perspective we'll achieve the same ammount of harder to spot vandalism. Effort required from vandals will increase e.g. twice, from one click to two clicks. Look from http://www.usemod.com/cgi-bin/mb.pl?SoftSecurity viewpoint. This is interesting - is there some real statistic on how watched articles are? I'm affraid there may be tens of thousand virtually unwatched articles no active wikipedians care about. IMO Wikipedia should less try to mimic what works for traditional encyclopedias (like the CITE madness) and concentrate more on inner proceses which works here - and improve software to aid them / identify where they fail. For example, locating various more likely dangerous articles, such as *unwatched articles *articles with high (external readers/wikipedian readers) ratio I'm affraid it only pushes us from SoftSecurity which works to HardSecurity which can not work for us. Jan Kulveit [[User:Wikimol]]

Jan Kulveit wrote: I think it would be excellent for us to study this. What I have long dreamed of is a regularly published list of "the most read / least watched articles". Take the number of pageviews divided by the number of watchers (plus one to prevent division by zero I guess). Or some more sophisticated variant of that to take into account minimum thresholds of concern perhaps. I think there are many cool tools yet to be developed, indeed yet to be dreamed of. --Jimbo

Jan Kulveit wrote: Actually, this *is* an example of SoftSecurity in action. SoftSecurity is about using minor speedbumps to discourage bad action in combination with making it very easy to make a good action to undo a bad action. -- brion vibber (brion @ pobox.com)

Jimmy Wales wrote: Unfortunately the pageview statistics are turned off :( Waerth/Walter

Jimmy Wales wrote: You forgot: 3. more crappy edits to existing articles (anecdotally). I agree we should collect more data, but we can't forget collecting data on #3, or else it will be pretty useless. -Mark

Daniel Mayer wrote: Well on the Dutch wikipedia the crap range for anon new pages is way over 50% so I guess on the english wiki you guys are really lucky ;) Waerth/Walter

On 12/6/05, Lars Aronsson <lars(a)aronsson.se> wrote: If by proven you mean works 100% of the time, of course not. If by proven you mean works 99% of the time, well, Wikipedia is the proof. Anthony

On 12/7/05, Jimmy Wales <jwales(a)wikia.com> wrote: "Encouraging new users to register"? When did Jimbo start using doublespeak? Try "forcing" or "compelling". Honesty and precision in language are our weapons in the fight against unmeaning. One of the potential negative consequences of creating a barrier to entry is that fewer good people will become editors because they will have been discouraged from creating a new article.

On Wed, 7 Dec 2005, The Cunctator wrote: New, usable articles created by anons accounted for around 40% of all new articles. This ignores speedied articles, etc which might inflate the figure in favor of anon-creation. Lots of them need wikification and start life as stubs; I wouldn't call that "crap". SJ

It is high time that we encourage the improvement of existing articles over adding new ones that will need to be maintained. This effort helps to push us in that direction. I hope we will change policy again if and when the negatives outweigh the benefits of letting new users create new pages. New users and anons can’t upload since undoing that is something that only admins can do. Creating new pages is also something that only admins can undo. Ergo… -- mav __________________________________________ Yahoo! DSL – Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com

--- Delirium <delirium(a)hackish.org> wrote: And that is two actions instead of one to remove patent nonsense (at least double the work). Thus it makes sense to limit the amount of patent nonsense we get (and thus that effort could be used somewhere else). Not allowing anons and newbies to upload images limits the amount of crap images we have to deal with. If *too many* anons that would have created crap articles as IPs start to create user accounts to post new pages, then it would be logical to extend the new page creation ban to new user accounts just as we already don’t allow new users to upload images. Slippery slope? Not really since there is already a ledge to land on that was created with the upload policy. It is a matter of weighing costs vs benefits as they pertain to our goal of creating the best and largest free encyclopedia in the world. We can and should remain as open as possible for as long as possible. But we need to constantly monitor our processes to see if they are, on average, bringing us closer or farther from our goal. Once a problem area is identified we then need to act but try to do so in a deliberate and thoughtful way. -- mav __________________________________________ Yahoo! DSL – Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com

--- Ilya Haykinson <haykinson(a)gmail.com> wrote: Recruitment like that is certainly a benefit and should be considered (the smaller the wiki, the more important recruitment is). But we also have to consider the bad effects as well. At least on the larger wikis the bad effects on anon article creation seem to greater than the positive. The larger wikis also have a much higher reader to editor ratio, so concentrating more effect on improving existing articles vs creating new ones that themselves need to be maintained may in fact be the better course of action. If setting aside 10 to 20 seconds to create an account is too big of a step, then that person really should not be increasing the maintenance burden on the community. Smaller wikis excepted since the likelihood of any one reader becoming an editor is much greater than on larger wikis. Having a user account does not a community member make. We all have all sorts of user accounts on the Internet. That does not mean we belong to communities associated with those accounts. -- mav __________________________________________ Yahoo! DSL – Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com

On 12/7/05, Daniel Mayer <maveric149(a)yahoo.com> wrote: New users can't upload images? I must have missed that one. Is this just on en, everywhere, in selected places? What is considered a new user?

--- Angela <beesley(a)gmail.com> wrote: That's it. My point is that we have already moved in that direction and the ability is already in the software to ID new users. -- mav __________________________________________ Yahoo! DSL – Something to write home about. Just $16.99/mo. or less. dsl.yahoo.com

On 12/7/05, Delirium <delirium(a)hackish.org> wrote: Maybe new page patrol needs to be more about researching and improving new articles and less about "deleting crap". The article that started this whole discussion wasn't "crap". It wasn't "nn cruft" or "vanity". But it did have some serious problems that could have been resolved if more time was spent researching it. I'm not saying this to lay blame, after all new page patrol is a volunteer effort so any help is better than nothing. But if you're ever "bored" I can think of a few million things to do, things that are more important than "deleting crap" anyway. Anthony

Anthony DiPierro wrote: Well, they aren't really comparable activities. New-page patrol for crap is a mindless activity that can be done to procrastinate and/or relieve stress. Researching someone I've never heard of, don't care about, and don't have the resources available to research in the first place isn't. On the occasions I *do* want to do serious research for Wikipedia, I prefer to do it on areas in which I either have expertise or at least some amount of interest (I have a long list of articles I care about that need improving or writing in the first place). Perhaps some volunteers can be persuaded to do "random researching", but I suspect it will be relatively few. -Mark

Daniel Mayer wrote: The way I read your answer, you assume that all improvement is coming from our existing editors. I do not know what you base this on. Do you have some sources for that that prove this assertion ?? It is not the time that it takes that IS the big step. You misrepresent the problem with vandalism on smaller projects. It takes quite some time and doing before there is a vibrant community. This is the time when a starting project is most vulnerable to vandals. The vandals drive people away and this often results in abandoned projects. The problem that exists is with a perceived problem, a problem that we have had from the beginning. We have always replied that we have a particular percentage of stubs and this has proved to be constant. We have always had a percentage of quality articles and this proved to be constant. We have always had a percentage of problematic articles and that proved to be constant too. Now the problem is that this cannot be helped. We can work hard to improve our quality and we do. But thinking that we can get absolute good quality is an achievable goal is like believing in Santaclaus. When people say that doomsday will come if we do not achieve this, they are spreading FUD. It is not helpful, if anything I fear the measures needed to get this unattainable goal may fracture and ultimately destroy our community. In several answers I have read it is put forward that people may be willing to provide sources but only for the stuff that they ARE working on. Many also state that they are not willing to find sources for stuff that is already there. Consequently there are 600.000 articles in the English language wikipedia that will probably not be annotated. We may not get new articles from anonymous users but it is much harder to check anonymous changes. So yes, we have an experiment going. That is all the good that can be said about it. I have read people say that when on patrol they get bored because not enough is happening... If anything the one thing that makes patrolling for problems a hardship, is the performance of our system. Having made the step to have a user account like described is a big step towards becoming a community member. That is the point that you try to deny. Thanks, GerardM

Daniel Mayer wrote: Over time more people sign on to our project. Given our importance the number of these people in a similar time frame increase. It is a similar situation as with stubs, numerically the numbers grow and as a percentage they stay the same. The point of this exercise is to decrease the existence of flawed articles. The blunt mechanism that is chosen for now will not help the existing flawed articles. It only leads on an ever increasing path of repressive measures. When you say that we need to improve our methodology I agree. I agree that it requires tooling. When you single the bigger wikis out to be correct, I think the tooling you are thinking of is probably flawed. The small wikis are probably more in need of improvement than the bigger ones. The chance of lawsuits is not less but more complicated. The consequences of not addressing issues may lead to a total ban of our resources in a given country. I rate the current block of wikipedia from China as more damaging for the relevance of our data. My motivation is that we know that there are issues with content and we do address these issues as well as we possibly can. By loosing our public we lose our relevance and we lose their point of view that helps our point of view to be neutral. Again, the number of users needs to be seen in relation to a time frame and in relation to the size of our popularity. I would not be surprised that the amount of new users has been fairly constant. I would be the last to deny our goal. I am all for better tooling and for being more discriminatory. I am all for stimulating annotations. I am however also afraid that the current train of thought is well under way, we are its passenger and we do not know where it will lead us and how far it lead us away from the very ideals that so many hold dear. Thanks, GerardM