In fact, most of the Congress edits were good edits, and only a small proportion were spin-doctored, vandalism, pranks, etc — not really more than in the general pool of edits — and there's no reason to cut those off unless unavoidable. Most people really do edit Wikipedia in good faith, not bad.
The George W. Bush article is probably a bad example. It's the most-edited article on Wikipedia by about 5:1 over the next-most-edited article. The current version is generally not awful, but not that great. It's an example of too many cooks spoiling the broth.
Fortunately, there's only about 200 articles (out of 900k+) that have more than 100 editors ever. Most articles are not in fact controversial.
The reason Wikipedia is so popular (today: Alexa no. 17 website, 30k/million reach, i.e. 3% of ALL web page hits are wikipedia.org) is that its competition isn't the Encyclopaedia Britannica — it's other websites. Wikipedia will tend to (a) summarise those sites (b) act as a guide to them. If you're looking up something on the web, you can scour Google for something that isn't an ad-festooned shopping portal, or you can look on Wikipedia to see if there's an article yet.
We have a guideline on living people's articles. Basically, (1) every statement has to be utterly verifiable (2) every statement has to be relevant to why they have an article. This is followed very imperfectly, but when followed it saves greatly on crap.
Not sure I'd call that 'content control', more editorial functions rather than writing. Some people really enjoy copyediting!
I don't have a link to the numbers on hand... the study was done by Kim Bruning and Gregory Maxwell and I posted about it to wikien-l, but I can't find which email it was. In the last month or so, though.
AIUI, the hard part is distributing the databases. There's a worldwide network of Squids which help a lot with the basic page hit rate from anonymous readers.
The performance limitations are not the PHP — they're the database, the web servers and the worldwide network of Squid proxies used to cache the web server output. These get the utmost of minute-by-minute attention from the sysadmins and DBAs.
Actually, the biggest performance limitation is Wikipedia's ridiculous popularity. Page hits went up 50% in January alone. A top-20 website is run by a nonprofit with one paid technical employee. Good fucking God.
No planned release date. I'm one of the people desperately pushing for this to happen for the English Wikipedia, and I'm extremely envious of the German Wikipedia's success in the matter. A fairly good and large encyclopedia on DVD for 10 Euros!
Wikipediareview is a board where their idea of incisive criticism of Wikipedia is to talk about an admin's teeth.
I probably have opposite politics to Jimmy Wales, but we get along very well working on the encyclopedia. You see this a lot in contentious Net forums; e.g. on Usenet group aus.politics in the mid-1990s, the discussion tended to separate not into left and right, but into lucid (people who could actually discuss things sensibly) and batshit (people who couldn't). The same tends to emerge on contentious articles on Wikipedia.
I wouldn't say anything to do with politics or history. But Israel-Palestine is an area with so many editors trying to push a point of view that the articles end up pathological. Linux is almost as bad — I've found it nearly impossible to get open source advocates to SWITCH OFF THE DAMN ADVOCACY and write something neutral, something that someone who disagreed with them couldn't reasonably argue against.
There are various proposals along these lines. I'm desperate to see at least one implemented on the English Wikipedia.
Note, however, that the German Wikipedia has just put out its third DVD edition without such a mechanism — just a lot of editors going over articles to format, fact-check and stabilise them for a fixed-media edition. So it can be done nevertheless.
Wikipedia is a project to write an encyclopedia first, and a community second. If its status as an experiment in Internet democracy affects the encyclopedia work, then the community structure needs work; and if it fails to be a suitable example of Internet democracy, too bad, because there's an actual job it's all supposed to be for.
This is actually a big problem at the moment because of the flood of recent attention — page hits went up 50% in January alone, which means 50% more readers, which means 50% more editors, which means lots of unacculturated newbies. It's A Tricky One.
There's various languages used, not just PHP — if you have a useful extension you want to write in Ruby on Rails, there's nothing stopping you. Mediawiki is written in PHP because the previous version of the software was. But what it really is is database programming, i.e. MySQL. That's why Mediawiki hacking is harder than the fact of it being written in PHP might seem to indicate.
Many people don't like the articles about them and many just don't like being written about. In cases like this, the living biographies guideline is useful: use only the highest quality references you can find, reference every statement you can and do your best to make the article right.
Probably most, actually. You just don't hear about it. Someone ran the numbers on this recently. Most articles are uncontroversial; it's only a very few that attract attention for their disputes, and maybe a couple of hundred (out of 900k+) that are so contested as to be pathological (the George W. Bush article, the most edited article on the entire wiki by a factor of five or so, is a good example).
The simple answer is "don't." Wikipedia is ideally an encyclopedia giving an overview of a subject and shouldn't be used as a direct reference above high school level. At college level, you should check the references given in the article yourself. If there aren't any, it makes it dubious already.
There's no easy technical solution to social problems. In practice, what we have is people who enjoy copyediting, fact-checking dubious claims and so forth. We particularly need to push this for living biographies, of course, after the Siegenthaler incident (which was a storm in a teacup IMO, but focused attention wonderfully). There's no substitute for clueful human review.
That's usually dealt with not by having article forks, but by having a main article and subarticles. This can be a tricky one to keep neutral — e.g. making an article about criticism of or controversy over something as a subarticle of the main article rather than a point-of-view fork — but it works well enough with reasonably sane editors of good will.
There are wikis that try this, e.g. WikInfo, which is a friendly fork of Wikipedia. The trouble is that it makes point-of-view-pushing editors happy at the expense of the article quality and the readers. Neutral point of view means that an article should contain all significant points of view on a subject, relative to their significance. (Thus, Wikipedia is a secondary or tertiary source.) This is a high ideal and difficult to reach in practice, but it does give us a reliable compass to work to. There's a tutorial as well.
The trouble is that a lot of experts can write text, but can't work computers in general. Wikipedia is amazingly usable by technophobes as well as geeks, and that would be a really bad thing to lose. What tends to happen is that you can write an article in plain text with references, and someone else will see it and fix it up with the fancy stuff. It works so far.
Note that Tycho's rant was provoked by Wikipedia having the temerity to delete an article he wrote about his own made-up fantasy world. Original research, no third-party references even possible... delete.
Slashdot Mirror
Please, please post this to UnNews.
Hence the Biographies of living persons guideline.
And TOR proxies are blocked at Wikipedia on sight. Because most of the edits from them are vandals and trolls.
In fact, most of the Congress edits were good edits, and only a small proportion were spin-doctored, vandalism, pranks, etc — not really more than in the general pool of edits — and there's no reason to cut those off unless unavoidable. Most people really do edit Wikipedia in good faith, not bad.
Fortunately, there's only about 200 articles (out of 900k+) that have more than 100 editors ever. Most articles are not in fact controversial.
The reason Wikipedia is so popular (today: Alexa no. 17 website, 30k/million reach, i.e. 3% of ALL web page hits are wikipedia.org) is that its competition isn't the Encyclopaedia Britannica — it's other websites. Wikipedia will tend to (a) summarise those sites (b) act as a guide to them. If you're looking up something on the web, you can scour Google for something that isn't an ad-festooned shopping portal, or you can look on Wikipedia to see if there's an article yet.
We have a guideline on living people's articles. Basically, (1) every statement has to be utterly verifiable (2) every statement has to be relevant to why they have an article. This is followed very imperfectly, but when followed it saves greatly on crap.
I don't have a link to the numbers on hand ... the study was done by Kim Bruning and Gregory Maxwell and I posted about it to wikien-l, but I can't find which email it was. In the last month or so, though.
AIUI, the hard part is distributing the databases. There's a worldwide network of Squids which help a lot with the basic page hit rate from anonymous readers.
Actually, the biggest performance limitation is Wikipedia's ridiculous popularity. Page hits went up 50% in January alone. A top-20 website is run by a nonprofit with one paid technical employee. Good fucking God.
No planned release date. I'm one of the people desperately pushing for this to happen for the English Wikipedia, and I'm extremely envious of the German Wikipedia's success in the matter. A fairly good and large encyclopedia on DVD for 10 Euros!
What Wikipedia is not.
I probably have opposite politics to Jimmy Wales, but we get along very well working on the encyclopedia. You see this a lot in contentious Net forums; e.g. on Usenet group aus.politics in the mid-1990s, the discussion tended to separate not into left and right, but into lucid (people who could actually discuss things sensibly) and batshit (people who couldn't). The same tends to emerge on contentious articles on Wikipedia.
I wouldn't say anything to do with politics or history. But Israel-Palestine is an area with so many editors trying to push a point of view that the articles end up pathological. Linux is almost as bad — I've found it nearly impossible to get open source advocates to SWITCH OFF THE DAMN ADVOCACY and write something neutral, something that someone who disagreed with them couldn't reasonably argue against.
Note, however, that the German Wikipedia has just put out its third DVD edition without such a mechanism — just a lot of editors going over articles to format, fact-check and stabilise them for a fixed-media edition. So it can be done nevertheless.
This is actually a big problem at the moment because of the flood of recent attention — page hits went up 50% in January alone, which means 50% more readers, which means 50% more editors, which means lots of unacculturated newbies. It's A Tricky One.
There's various languages used, not just PHP — if you have a useful extension you want to write in Ruby on Rails, there's nothing stopping you. Mediawiki is written in PHP because the previous version of the software was. But what it really is is database programming, i.e. MySQL. That's why Mediawiki hacking is harder than the fact of it being written in PHP might seem to indicate.
Many people don't like the articles about them and many just don't like being written about. In cases like this, the living biographies guideline is useful: use only the highest quality references you can find, reference every statement you can and do your best to make the article right.
Probably most, actually. You just don't hear about it. Someone ran the numbers on this recently. Most articles are uncontroversial; it's only a very few that attract attention for their disputes, and maybe a couple of hundred (out of 900k+) that are so contested as to be pathological (the George W. Bush article, the most edited article on the entire wiki by a factor of five or so, is a good example).
The simple answer is "don't." Wikipedia is ideally an encyclopedia giving an overview of a subject and shouldn't be used as a direct reference above high school level. At college level, you should check the references given in the article yourself. If there aren't any, it makes it dubious already.
There's no easy technical solution to social problems. In practice, what we have is people who enjoy copyediting, fact-checking dubious claims and so forth. We particularly need to push this for living biographies, of course, after the Siegenthaler incident (which was a storm in a teacup IMO, but focused attention wonderfully). There's no substitute for clueful human review.
If someone does an edit or some work someone else particularly likes, they'll often leave a note thanking them on their user talk page.
That's usually dealt with not by having article forks, but by having a main article and subarticles. This can be a tricky one to keep neutral — e.g. making an article about criticism of or controversy over something as a subarticle of the main article rather than a point-of-view fork — but it works well enough with reasonably sane editors of good will.
There are wikis that try this, e.g. WikInfo, which is a friendly fork of Wikipedia. The trouble is that it makes point-of-view-pushing editors happy at the expense of the article quality and the readers. Neutral point of view means that an article should contain all significant points of view on a subject, relative to their significance. (Thus, Wikipedia is a secondary or tertiary source.) This is a high ideal and difficult to reach in practice, but it does give us a reliable compass to work to. There's a tutorial as well.
The trouble is that a lot of experts can write text, but can't work computers in general. Wikipedia is amazingly usable by technophobes as well as geeks, and that would be a really bad thing to lose. What tends to happen is that you can write an article in plain text with references, and someone else will see it and fix it up with the fancy stuff. It works so far.
Note that Tycho's rant was provoked by Wikipedia having the temerity to delete an article he wrote about his own made-up fantasy world. Original research, no third-party references even possible ... delete.