6 ms·

Wikipedia is a wild success and an utter failure

62 points by stdbrouw 15y ago

jonnathanson 15y ago
The problem with Wikipedia -- one that stems from, and compounds, all of the problems mentioned in this piece -- is that the wilderness has been cleared. The fun, groundbreaking contribution work is done. Much of -- if not most of -- the big, foundational, crystallized knowledge base on the site has been written. All that remains now are small(er) updates to limited subfields, rote listmaking, current events updates (never Wikipedia's strong suit in the first place), and editing. Plenty of editing. None of those things is inherently fun for most people. Your average internet user -- someone who would have loved contributing a fresh article ten years ago -- probably has little desire to devote an appreciable amount of time to curating and categorizing other people's work. But yet, there's a vocal minority of users who do enjoy such things, and those are the "procedural whackjobs" of which you speak. Call this the rise of the Editorial Class. Wikipedia is now dominated by a small percentage of editors, who revert new work by contributors, and whose arcane rules, lists, procedures, politics, and supercillious disdain make contributing fresh material more a labor than a labor of love. And, lest we think this is noble work put forth by self-sacrificing custodians of knowledge, let me say that my encounters with such people have rarely left an impression of maturity, or even competence. No, for a lot of them, it's a petty power trip. (To wit: one of my articles -- thoroughly researched and painstakingly written -- was eventually lobbied for deletion, successfully, by a little clique of editors calling itself the "Counter Vandalism Unit." As far as I could tell, this is a group of grown-ass adults playing "24" dress-up on Wikipedia, complete with their own badge-like logo, who sadly wield a decent amount of editorial control over the site). On the flipside, as you point out, some of this works in Wikipedia's favor. More timely sources of knowledge, like Twitter, Tumblr, Facebook, search engines, etc., can beat out Wikipedia's relatively antiquated categorization systems and contribution mechanics fairly handily. But, by and large, most people contributing "content" or "information" through such channels are not really cognizant that that's what they're doing (witness the big uproar over Facebook's data collection; people are inherently surprised that they've been contributing data to any sort of repository of such). Rarely does someone tweet, or post an FB status update, thinking he's contributing to the grand collection of all human knowledge. So the sum total of all the knowledge-data generated outside of Wikipedia, more current and possibly even more interesting though it may be, is still a lot more flab than muscle. That will change eventually, as data mining and contextualization technologies improve. For the time being, however, the lack of an organizing principle similar to Wikipedia's, outside of Wikipedia, will keep Wikipedia relevant. Not indefinitely, but for quite a few years to come.
- stdbrouw 15y ago
  Here's where I think Wikipedia could benefit from a good look at how other social systems handle their editorial and decision making processes — for example, how something as simple as +1/±0/-1 makes resolving tough problems so much easier for open-source project, or just the fact that we bundle problems into issue reports. Wikipedia, on the other hand, has just one endless thread of discussion comments, and like regular comments on news websites that approach is not conducive to civil, intelligent debate in the least.
  - tommorris 15y ago
    http://en.wikipedia.org/wiki/Wikipedia:Polling_is_not_a_substitute_for_discussion http://en.wikipedia.org/wiki/Wikipedia:Polling_is_not_a_subs... And there are many structured discussions on Wikipedia that follow set patterns and rules. Take deletion: at Articles for Deletion (the area of the site where deletion is debated), there are a set of common patterns and anti-patterns of arguments that you should use and avoid using, and there is a time period where you can submit comments and something vaguely like quorate required before a decision can be made. At the end of a week, the discussion is wrapped up: it's obvious, it gets kept or deleted, and if not, the admin has to sum up the arguments and accept them in a sort of quasi-judicial manner. It's rather more civilized than YouTube or newspaper comments pages...
    - stdbrouw 15y ago
      Good points, cheers. It's messy (at the very least just visually) but there are some standards. I'll stick to my assertion, though, that more people need to think about this stuff, because the organically grown processes don't always cut it — otherwise, you wouldn't have frustrated rants like those of Zed Shaw (http://sheddingbikes.com/posts/1297662169.html http://sheddingbikes.com/posts/1297662169.html) and Jason Scott (http://ascii.textfiles.com/archives/808 http://ascii.textfiles.com/archives/808).
      - tommorris 15y ago
        And there are people thinking about this and trying to work on software solutions for it. The Wikimedia Foundation are trying to create a variety of tools for new page patrollers (I used to do it) to aid them in better handling the flow of new articles coming in (it's called the "zoom interface"). Another thing that is being tested is making the automatic templates used to notify people that articles are up for deletion or whatever a lot more friendly. For the last month or so, the Foundation and the community have been working together to A-B test better warning templates. As for discussion threads, there's LiquidThreads, which has been in development for a while but would provide a threaded discussion system sort of like what Reddit and HN have but for Wikimedia talk pages, but still keeping the flexibility of being able to refactor discussions and move stuff about without an admin. LiquidThreads is in use on English Wikinews and on mediawiki.org, but English Wikipedia doesn't have it because it isn't quite mature enough yet (or something, I haven't kept track). Finally, there's the big one: the visual editor. That's in development at the moment, hopefully with testing starting by the end of the year. This is going to be an opt-out WYSIWYG-style editor. See http://www.mediawiki.org/wiki/Visual_editor/software_design http://www.mediawiki.org/wiki/Visual_editor/software_design for a rough idea how it is being built. There is lots of consensus that things do need to change and Wikipedia does need to have more technical solutions to the problems that are showing with some of the decade-old organically grown social processes.
jessriedel 15y ago
If there's one piece of framework Wikipedia needs, it's a centralized database for citations. That database should contain every book, article, poem, etc. that's cited in any Wikipedia article, and I should be able to re-cite an entry with a single number or tag...without having to re-enter the name of the darn publisher myself. How much money would it take to make this happen?
- Thrymr 15y ago
  As the article points out, Wikipedia really needs lots of its information in standard forms. Lots of things are done manually that ought to be able to be automated. I think the citation one is huge. Even the same article often has a "Notes" (for footnotes) section and a "References" section that is a bibliography unlinked to the text. They often both contain citations in different formats. If this is hard even within an article, how is it going to happen across all of Wikipedia? I hope someone takes it on, though.
- wpietri 15y ago
  Part of the problem here is that Wikipedia has traditionally had way more labor than money. So the notion of spending money to save labor -- common everywhere else -- isn't really the default Wikipedia reflex. That seems to be changing some in the last few years, but between the code and the data there's a lot of inertia there.
  - Thrymr 15y ago
    In this case, what's really needed is high-level labor (organization, and maybe programming), and not low-level labor (typing in the damn references by hand). I guess it's easier to get volunteers to do the latter, and the "high-level" volunteers are mostly editors and wiki-policy wonks, not thinking about appropriate data structures and implementation. It is striking how little Wikipedia has advanced technically since it started, compared with its enormous growth of content.
    - wpietri 15y ago
      I think the enormous growth of content (and usage) is exactly why Wikipedia hasn't advanced much technically. In its early years it was a total shoestring operation, mainly focused on keeping the site up under incredible traffic. Just surviving that is a titanic victory. However, that victory came at a cost: the software fell way behind. The community picked up the slack, making a fantastic reference site despite the handicaps. But many of the ways they did that (e.g., duplicating references all over the place) make improving the software even harder. Now that Wikimedia Foundation staffing is starting to catch up, I'm sure some of these things will get better. But it's definitely going to take a while.
rmc 15y ago
Wikipedia is Web 2.0. User generated content online, crowd sourced, the whole lot. However for sources, they only look at early 20th centuary tech, like newspapers or books. They don't value internet sources nearly as much.
- adrianN 15y ago
  I think it's very reasonable not to value internet sources as much. With print sources usually you have at least an editor and at best some kind of peer review that strongly increases credibility. Citing blogs just doesn't cut it. With web sources you also have the problem of bit rot and broken links, print media can usually be obtained in any decent library for way longer than websites typically survive.
  - DanBC 15y ago
    That's true, but it's alarming to think that (some) UK newspapers are deemed trustworthy just because they're in print.
    - tommorris 15y ago
      Not as much as you might think. After the Daily Mail (think Britain's answer to Fox News but with the additional benefit that you can use it to wipe yourself if you are out of toilet roll) published a story saying that Amanda Knox's appeal had been rejected, we had an enormous discussion on the Reliable Sources Noticeboard about whether to ban the Daily Mail as a source. We didn't reach consensus on whether to do so, but you can bet that the next time they do something equally stupid, said thread will come up in discussion. A few more major blunders and they will quite probably stop being considered a reliable source on WP.
      - rmc 15y ago
        Here's an article from the Daily Mail saying Facebook Gives You Cancer: http://www.dailymail.co.uk/health/article-1149207/How-using-Facebook-raise-risk-cancer.html http://www.dailymail.co.uk/health/article-1149207/How-using-... It's the blanket "all newspapers good, all blogs bad" simplistic idea that I think is nonsense.
DanBC 15y ago
The article doesn't mention the huge number of adolescents using twinkle (and other tools) to make very many changes very quickly. Seeing editors being proud of the number of articles they've had deleted, or the number of reverts they've made, is pretty sickening.
- tommorris 15y ago
  A lot of articles get deleted because they're crap (CSD etc.). Yes, deletionism can go too far, and crowing about how much one has deleted is hardly the epitome of virtue, but keeping the site clean of spam, vandalism and crap is a good thing.
  - hugh3 15y ago
    Sometimes when I'm really bored I read the Articles for Deletion. It's always enough to satisfy myself that... yep, the articles being deleted really are crap. I admire wikipedia editors, the gnomes of our age, working tirelessly on a boring, unglamorous task which gives them satisfaction.
twopounder 15y ago
The true achievement of Wikipedia is not in its content, but its groundwork. Every company should maintain its own internal and external wikis, even if only employees can access or modify them. It simplifies distribution of documentation and files in a very user friendly way. And let's face it. Whack jobs write books and news paper columns too. Wikipedia has a surprising amount in common with college text books.
- eru 15y ago
  We know that Wikis are useful. Wikipedia wasn't the first wiki.
  - twopounder 15y ago
    Unless its 2001 and your company's documentation is stored on a shared drive. Heck, when I graduated in 2009, my college was still keeping all of its documents in a maze of ntfs shares. You don't have to be the first to lay the groundwork either.
hugh3 15y ago
To me, wikipedia is like a dog walking on its hind legs (as Samuel Johnson said of women preachers). The point isn't that it isn't done very well, the point is that it works at all. If you'd asked me, at the start of the wikipedia project, whether it would ever grow to be a useful resource I'd have said "hell no, it'll just get taken over by idiots, bad content will drive out good, you can never put together a useful resource by collaborative editing, it's a pipe dream". And yet here we are, in a situation where wikipedia is actually a damn good resource on just about any subject of interest. Almost all the people who complain about wikipedia are people who are attempting to edit it. And some of them may have a point... while others are attempting to insert stuff that doesn't fit the guidelines. But as long as you're using it in read-only mode, what's to complain about?
- nitrogen 15y ago
  But as long as you're using it in read-only mode, what's to complain about? The deletion of noteworthy information about popular programming languages is one example.
  - hugh3 15y ago
    Which one is that? Anyway, like I said, wikipedia ain't perfect, but it's amazing how far from terrible it is.
    - nitrogen 15y ago
      Alice (http://www.hnsearch.com/search#request/all&q=wikipedia+alice http://www.hnsearch.com/search#request/all&q=wikipedia+a...), plus other important pages: http://www.hnsearch.com/search#request/all&q=wikipedia+deletion http://www.hnsearch.com/search#request/all&q=wikipedia+d... I realize there's a tremendous amount of spam, vanity pages, vandalism, and useless content the editors have to deal with, but it's still disturbing to read about the occasional outbreaks of deletionism that affect Wikipedia, while corporate and political whitewashing goes apparently unchecked (I don't have a specific reference in mind).
  - tommorris 15y ago
jerf 15y ago
Wikipedia's secret is that it manages to make the average contribution very slightly positive on average, then makes up the rest in volume. When everything is cumulative like that, the fact that the sum total of Wikipedia contributions do indeed contain a lot of crap doesn't matter, since "good - bad" still comes out quite positive. It's a delicate balance, though.

62 points by stdbrouw 15y ago

The problem with Wikipedia -- one that stems from, and compounds, all of the problems mentioned in this piece -- is that the wilderness has been cleared. The fun, groundbreaking contribution work is done. Much of -- if not most of -- the big, foundational, crystallized knowledge base on the site has been written. All that remains now are small(er) updates to limited subfields, rote listmaking, current events updates (never Wikipedia's strong suit in the first place), and editing. Plenty of editing. None of those things is inherently fun for most people. Your average internet user -- someone who would have loved contributing a fresh article ten years ago -- probably has little desire to devote an appreciable amount of time to curating and categorizing other people's work. But yet, there's a vocal minority of users who do enjoy such things, and those are the "procedural whackjobs" of which you speak. Call this the rise of the Editorial Class. Wikipedia is now dominated by a small percentage of editors, who revert new work by contributors, and whose arcane rules, lists, procedures, politics, and supercillious disdain make contributing fresh material more a labor than a labor of love. And, lest we think this is noble work put forth by self-sacrificing custodians of knowledge, let me say that my encounters with such people have rarely left an impression of maturity, or even competence. No, for a lot of them, it's a petty power trip. (To wit: one of my articles -- thoroughly researched and painstakingly written -- was eventually lobbied for deletion, successfully, by a little clique of editors calling itself the "Counter Vandalism Unit." As far as I could tell, this is a group of grown-ass adults playing "24" dress-up on Wikipedia, complete with their own badge-like logo, who sadly wield a decent amount of editorial control over the site). On the flipside, as you point out, some of this works in Wikipedia's favor. More timely sources of knowledge, like Twitter, Tumblr, Facebook, search engines, etc., can beat out Wikipedia's relatively antiquated categorization systems and contribution mechanics fairly handily. But, by and large, most people contributing "content" or "information" through such channels are not really cognizant that that's what they're doing (witness the big uproar over Facebook's data collection; people are inherently surprised that they've been contributing data to any sort of repository of such). Rarely does someone tweet, or post an FB status update, thinking he's contributing to the grand collection of all human knowledge. So the sum total of all the knowledge-data generated outside of Wikipedia, more current and possibly even more interesting though it may be, is still a lot more flab than muscle. That will change eventually, as data mining and contextualization technologies improve. For the time being, however, the lack of an organizing principle similar to Wikipedia's, outside of Wikipedia, will keep Wikipedia relevant. Not indefinitely, but for quite a few years to come.

stdbrouw 15y ago
Here's where I think Wikipedia could benefit from a good look at how other social systems handle their editorial and decision making processes — for example, how something as simple as +1/±0/-1 makes resolving tough problems so much easier for open-source project, or just the fact that we bundle problems into issue reports. Wikipedia, on the other hand, has just one endless thread of discussion comments, and like regular comments on news websites that approach is not conducive to civil, intelligent debate in the least.
- tommorris 15y ago
  http://en.wikipedia.org/wiki/Wikipedia:Polling_is_not_a_substitute_for_discussion http://en.wikipedia.org/wiki/Wikipedia:Polling_is_not_a_subs... And there are many structured discussions on Wikipedia that follow set patterns and rules. Take deletion: at Articles for Deletion (the area of the site where deletion is debated), there are a set of common patterns and anti-patterns of arguments that you should use and avoid using, and there is a time period where you can submit comments and something vaguely like quorate required before a decision can be made. At the end of a week, the discussion is wrapped up: it's obvious, it gets kept or deleted, and if not, the admin has to sum up the arguments and accept them in a sort of quasi-judicial manner. It's rather more civilized than YouTube or newspaper comments pages...
  - stdbrouw 15y ago
    Good points, cheers. It's messy (at the very least just visually) but there are some standards. I'll stick to my assertion, though, that more people need to think about this stuff, because the organically grown processes don't always cut it — otherwise, you wouldn't have frustrated rants like those of Zed Shaw (http://sheddingbikes.com/posts/1297662169.html http://sheddingbikes.com/posts/1297662169.html) and Jason Scott (http://ascii.textfiles.com/archives/808 http://ascii.textfiles.com/archives/808).
    - tommorris 15y ago
      And there are people thinking about this and trying to work on software solutions for it. The Wikimedia Foundation are trying to create a variety of tools for new page patrollers (I used to do it) to aid them in better handling the flow of new articles coming in (it's called the "zoom interface"). Another thing that is being tested is making the automatic templates used to notify people that articles are up for deletion or whatever a lot more friendly. For the last month or so, the Foundation and the community have been working together to A-B test better warning templates. As for discussion threads, there's LiquidThreads, which has been in development for a while but would provide a threaded discussion system sort of like what Reddit and HN have but for Wikimedia talk pages, but still keeping the flexibility of being able to refactor discussions and move stuff about without an admin. LiquidThreads is in use on English Wikinews and on mediawiki.org, but English Wikipedia doesn't have it because it isn't quite mature enough yet (or something, I haven't kept track). Finally, there's the big one: the visual editor. That's in development at the moment, hopefully with testing starting by the end of the year. This is going to be an opt-out WYSIWYG-style editor. See http://www.mediawiki.org/wiki/Visual_editor/software_design http://www.mediawiki.org/wiki/Visual_editor/software_design for a rough idea how it is being built. There is lots of consensus that things do need to change and Wikipedia does need to have more technical solutions to the problems that are showing with some of the decade-old organically grown social processes.

jessriedel 15y ago

If there's one piece of framework Wikipedia needs, it's a centralized database for citations. That database should contain every book, article, poem, etc. that's cited in any Wikipedia article, and I should be able to re-cite an entry with a single number or tag...without having to re-enter the name of the darn publisher myself. How much money would it take to make this happen?

Thrymr 15y ago
As the article points out, Wikipedia really needs lots of its information in standard forms. Lots of things are done manually that ought to be able to be automated. I think the citation one is huge. Even the same article often has a "Notes" (for footnotes) section and a "References" section that is a bibliography unlinked to the text. They often both contain citations in different formats. If this is hard even within an article, how is it going to happen across all of Wikipedia? I hope someone takes it on, though.
wpietri 15y ago
Part of the problem here is that Wikipedia has traditionally had way more labor than money. So the notion of spending money to save labor -- common everywhere else -- isn't really the default Wikipedia reflex. That seems to be changing some in the last few years, but between the code and the data there's a lot of inertia there.
- Thrymr 15y ago
  In this case, what's really needed is high-level labor (organization, and maybe programming), and not low-level labor (typing in the damn references by hand). I guess it's easier to get volunteers to do the latter, and the "high-level" volunteers are mostly editors and wiki-policy wonks, not thinking about appropriate data structures and implementation. It is striking how little Wikipedia has advanced technically since it started, compared with its enormous growth of content.
  - wpietri 15y ago
    I think the enormous growth of content (and usage) is exactly why Wikipedia hasn't advanced much technically. In its early years it was a total shoestring operation, mainly focused on keeping the site up under incredible traffic. Just surviving that is a titanic victory. However, that victory came at a cost: the software fell way behind. The community picked up the slack, making a fantastic reference site despite the handicaps. But many of the ways they did that (e.g., duplicating references all over the place) make improving the software even harder. Now that Wikimedia Foundation staffing is starting to catch up, I'm sure some of these things will get better. But it's definitely going to take a while.

rmc 15y ago

Wikipedia is Web 2.0. User generated content online, crowd sourced, the whole lot. However for sources, they only look at early 20th centuary tech, like newspapers or books. They don't value internet sources nearly as much.

adrianN 15y ago
I think it's very reasonable not to value internet sources as much. With print sources usually you have at least an editor and at best some kind of peer review that strongly increases credibility. Citing blogs just doesn't cut it. With web sources you also have the problem of bit rot and broken links, print media can usually be obtained in any decent library for way longer than websites typically survive.
- DanBC 15y ago
  That's true, but it's alarming to think that (some) UK newspapers are deemed trustworthy just because they're in print.
  - tommorris 15y ago
    Not as much as you might think. After the Daily Mail (think Britain's answer to Fox News but with the additional benefit that you can use it to wipe yourself if you are out of toilet roll) published a story saying that Amanda Knox's appeal had been rejected, we had an enormous discussion on the Reliable Sources Noticeboard about whether to ban the Daily Mail as a source. We didn't reach consensus on whether to do so, but you can bet that the next time they do something equally stupid, said thread will come up in discussion. A few more major blunders and they will quite probably stop being considered a reliable source on WP.
    - rmc 15y ago
      Here's an article from the Daily Mail saying Facebook Gives You Cancer: http://www.dailymail.co.uk/health/article-1149207/How-using-Facebook-raise-risk-cancer.html http://www.dailymail.co.uk/health/article-1149207/How-using-... It's the blanket "all newspapers good, all blogs bad" simplistic idea that I think is nonsense.

DanBC 15y ago

The article doesn't mention the huge number of adolescents using twinkle (and other tools) to make very many changes very quickly. Seeing editors being proud of the number of articles they've had deleted, or the number of reverts they've made, is pretty sickening.

tommorris 15y ago
A lot of articles get deleted because they're crap (CSD etc.). Yes, deletionism can go too far, and crowing about how much one has deleted is hardly the epitome of virtue, but keeping the site clean of spam, vandalism and crap is a good thing.
- hugh3 15y ago
  Sometimes when I'm really bored I read the Articles for Deletion. It's always enough to satisfy myself that... yep, the articles being deleted really are crap. I admire wikipedia editors, the gnomes of our age, working tirelessly on a boring, unglamorous task which gives them satisfaction.

twopounder 15y ago

The true achievement of Wikipedia is not in its content, but its groundwork. Every company should maintain its own internal and external wikis, even if only employees can access or modify them. It simplifies distribution of documentation and files in a very user friendly way. And let's face it. Whack jobs write books and news paper columns too. Wikipedia has a surprising amount in common with college text books.

eru 15y ago
We know that Wikis are useful. Wikipedia wasn't the first wiki.
- twopounder 15y ago
  Unless its 2001 and your company's documentation is stored on a shared drive. Heck, when I graduated in 2009, my college was still keeping all of its documents in a maze of ntfs shares. You don't have to be the first to lay the groundwork either.

hugh3 15y ago

To me, wikipedia is like a dog walking on its hind legs (as Samuel Johnson said of women preachers). The point isn't that it isn't done very well, the point is that it works at all. If you'd asked me, at the start of the wikipedia project, whether it would ever grow to be a useful resource I'd have said "hell no, it'll just get taken over by idiots, bad content will drive out good, you can never put together a useful resource by collaborative editing, it's a pipe dream". And yet here we are, in a situation where wikipedia is actually a damn good resource on just about any subject of interest. Almost all the people who complain about wikipedia are people who are attempting to edit it. And some of them may have a point... while others are attempting to insert stuff that doesn't fit the guidelines. But as long as you're using it in read-only mode, what's to complain about?

nitrogen 15y ago
But as long as you're using it in read-only mode, what's to complain about? The deletion of noteworthy information about popular programming languages is one example.
- hugh3 15y ago
  Which one is that? Anyway, like I said, wikipedia ain't perfect, but it's amazing how far from terrible it is.
  - nitrogen 15y ago
    Alice (http://www.hnsearch.com/search#request/all&q=wikipedia+alice http://www.hnsearch.com/search#request/all&q=wikipedia+a...), plus other important pages: http://www.hnsearch.com/search#request/all&q=wikipedia+deletion http://www.hnsearch.com/search#request/all&q=wikipedia+d... I realize there's a tremendous amount of spam, vanity pages, vandalism, and useless content the editors have to deal with, but it's still disturbing to read about the occasional outbreaks of deletionism that affect Wikipedia, while corporate and political whitewashing goes apparently unchecked (I don't have a specific reference in mind).
- tommorris 15y ago

jerf 15y ago

Wikipedia's secret is that it manages to make the average contribution very slightly positive on average, then makes up the rest in volume. When everything is cumulative like that, the fact that the sum total of Wikipedia contributions do indeed contain a lot of crap doesn't matter, since "good - bad" still comes out quite positive. It's a delicate balance, though.