8 ms·
Wiki inventor Ward Cunningham develops federated wiki
- derleth 14y agoI don't see how this is anti-Wikipedia. But I suppose conflict sells, so linkbait (to be clear, linkbait on the part of Wired, not the submitter) titles like these are inevitable. Also, for a lot of people, 'wiki' is simply short for 'Wikipedia'. That's unlikely to change at this point. I don't know what Ward can do to make a Federated Wiki (great name, BTW) achieve any kind of success with that kind of barrier.
- klint 14y agoI'm the author of the article, and I can see why people got confused. I don't think Max Ogden was implying that it's a bad thing that you have to give up your view point on Wikipedia - and Ward Cunngingham definitely didn't imply that in my conversation with him. I regret that readers thought that. I don't think either Ogden or Cunningham are expecting this to replace Wikipedia, but they are hoping that it will be useful for other wikis.
- mark_l_watson 14y agoA little off topic, but a friend of mine had a dinner conversation with Ward Cunningham and refactoring expert Ralph Johnson a few months ago and posted it on youtube: http://www.youtube.com/watch?v=jqGYoKvekik http://www.youtube.com/watch?v=jqGYoKvekik The video could have used some editing (shorter!) but has some interesting parts.
- J3L2404 14y agoForking does not lead to spooning. This sounds like it will quickly become an impenetrable rat's nest. Better to fight it out on the one true source, Wikipedia.
- sp332 14y agoThe Internet is already made of "rat's nests" like these. DNS records, BGP route tables, cross-domain embedding, all of these are pulling together information from lots of inconsistent sources to render a page in your browser, and it works pretty well.
- J3L2404 14y agoGoogle is a rats nest but you can eventually find what you want, I go to Wikipedia first.
- sp332 14y agoThe problem Wikipedia is having at the moment is that some people have taken over sections of it as their own "turf". And there will always be a holy war between the completionists and the deletionists, etc. With a federated wiki, you get to pick what kind of info you want to see, and who gets to curate it.
- evangineer 14y agoIf deletionists had been running the show when Wikipedia got started, it would have failed.
- p_sherman 14y agoHoly mother of misleading titles.
- kzrdude 14y agoWhen I talked with Ward in 2005, he mentioned his vision of a Wiki had a continuum of pages where he said you could take a page and browse the different "directions" people had edited it to.
- tokenizer 14y agoThat sounds interesting, but useless. I only want to see the most 'correct' version. It's almost as if saying when you use an app you can also use many of the apps versions, including its beta or alpha, regardless of how incorrect or useless they are with the current version. I also think that information needs to be organized and many of his ideas make the collection or the organization of information almost impossible. But nevertheless very interesting.
- dchest 14y agoWho decides which version is the most correct? Read the same Wikipedia article on history in different languages and you'll find out that there's no such thing as the 'correct version'.
- bglusman 14y agoI hear what you're saying about only wanting the most "correct" version, and no doubt that's how most people feel when looking for information, but about some things you have an ability and perhaps an obligation to contribute to deciding what is the most correct version... This suggests to me that Ward is onto something here, but also that you may have a point about it's limitations in its current form. I hope the hypothes.is project (funded by kickstarter, http://hypothes.is http://hypothes.is) succeeds, and I wonder if some degree of integration between the two is possible so that reputation and correctness rating across wiki-nodes (or any such federated system) can naturally "bubble up" to top visibility with some citation/indication of the bubbling via the hypothes.is client, regardless of which specific fork of the wiki you happen to be looking at... I'd love to discuss this possibility further with Ward, hypothes.is, or anyone else interested since they'll both be open source projects any of us can try to make this happen in :-)
- Leynos 14y ago
- rocky1138 14y agoAlso: wikipedia offers its entire content available as a tarball. Fork and run with it as you wish. While not "federated" it's pretty damn open.
- J3L2404 14y agoHoly smokes, how big is it?
- _delirium 14y ago8 GB bzip2-compressed, for the English Wikipedia, if you download the dump that contains only the current revisions of article-space pages. Much bigger if you want the one with every historical revision.
- J3L2404 14y agoThanks. Important to mention that is text only.
- piranha 14y agoDepends on which language you'd like: http://en.wikipedia.org/wiki/Wikipedia:Database_download http://en.wikipedia.org/wiki/Wikipedia:Database_download
- jsvaughan 14y agosmall enough to have offline on your phone, as several apps do.
- sp332 14y agoThe text of a large percentage of the current version of English Wikipedia, without discussion pages, is that small.
- morsch 14y agoThe text, anyway, I don't think a dump including the images has been available for quite some time. And that's not just due to the size -- though that's also a concern, I imagine -- but mostly due to copyright concerns. Text submitted to Wikipedia is for the most part under a copyleft license, images often are not.
- e12e 14y agoTFP: http://wardcunningham.github.com/ http://wardcunningham.github.com/ in the spirit of slashdot: the fucking project
- e12e 14y agoTFC: https://github.com/WardCunningham/Smallest-Federated-Wiki https://github.com/WardCunningham/Smallest-Federated-Wiki
- zimbatm 14y agoSo where is the link to the actual project ?
- zimbatm 14y agohttp://wardcunningham.github.com/ http://wardcunningham.github.com/
- emperorcezar 14y ago"To run The Simplest Federated Wiki, you’ll need your own web server, which Cunningham thinks is an important part of the project." Sounded great till that point. He's putting his own nerd centric bias into it. If I'm a history guy, not a tech guy, and I want to "fork", asking me to run my own "server" or whatever is a non-starter.
- nicholassmith 14y agoSmart money says if there's demand for it people will be able to buy custom instances with it for a variable amount of money.
- clinth 14y agoAgreed. Github is popular even though git itself is obtuse.
- nicholassmith 14y agoExactly, you just need to reduce the friction level to a certain point. Saying it requires it's own web server sounds like a chore, but Mac OS X comes with a web server built in. You could in theory run your own local Wiki. Or someone clever could sell cheap instances powered off Amazon/Google/Heroku/cheap VPS host with a custom web frontend that does the heavy lifting. TL;DR good companies make life easy for people who want it without hassle.
- sp332 14y agoFTA: To make Federated Wiki easier to adopt, there’s a one-click installer to deploy a server to Amazon Web Services.
- nicholassmith 14y agoAh I missed that, saved the full thing for later.
- 14y ago
- drone 14y ago... and then the spammers have a field day: on every page, a list of 10,000 forks, re-titled with names like "Luis Vitton CHEAP$!$" and "U 2 can aFf0rd Rolex!" Interesting idea when all players are playing by the same rules and with the same intent. Not as appealing when the most active are there simply to generate noise.
- SwellJoe 14y agoThis has been a problem for every Internet service everywhere ever. That hasn't stopped email from being useful, forums from being useful, traditional wikis (weird to say "traditional wikis", now that there's a new kind of wiki) from being useful, websites from being useful, etc. The spam problem always has to be solved. That said, very active wikis have shown themselves to be more resistant to spammers than most online services. Because anyone can remove the spam, there's a pretty high cost to keeping the spam flowing, particularly if there are sufficient impediments to automation in place.
- drone 14y agoOf course every non-moderated service has spam, that was kinda the point of my post... Maybe the part you missed was that they introduced a new way of spamming via forking pages... I don't think I ever made any allusions to something not happening because of it. Thanks for the pep talk though, I'll keep it in mind as I do my nightly task of helping clean up spam on the sites where I moderate.
- SwellJoe 14y agoSpamming via forking seems, to me, to be very similar to trackback spam on blogs. But, yes, it will probably require a bit of novelty in terms of how one fights it.
- nicholassmith 14y agoI think this could lead to a lot of interesting possibilities but I'll focus on one. Lets say you're on Wikipedia and for example want to know about the assassination of J.F.K, the Wiki article contains most of the stock facts and so on, but lets say you want to know about the really crazy and out there conspiracy theories of Hoover and aliens and so on, a federated system would allow people to set their own page up and you can continue through to it. The sort of content that sends most Wikipedia moderators into catatonic seizures indexed and available. There's plenty of other uses out there, I just think it's an interesting point that it'll allow anyone with enough time on their hands to create an archive of what they think is relevant as a counterpoint to what someone else thinks is relevant. Don't like the standard descriptor? Normally you'd have to fit a wiki mod, fork it off and have it as a counter point.
- derleth 14y ago> The sort of content that sends most Wikipedia moderators into catatonic seizures indexed and available. The reason it is anathema to the Wikipedia concept is because Wikipedia wants its name on good content, not Alex Jones JOOZ-DID-IT (Oh, and Aliens-Big-Pharma-Bilderberger-USPS-Did-9/11 Too) bullshit, unless such bullshit is clearly marked as coming from a specific minority viewpoint. I think that's a great reason to keep the crap on the current version of Wikipedia to a minimum: The world is confusing, especially when there are people like Alex Jones out there trying to make it more confusing to sell more videos. Having something where you have a good chance of getting the Sane People Version of controversial (or only-controversial-because-of-the-crazies) topics is invaluable if you want somewhere to start. Especially if it's a topic that books aren't really written about, like specific health scams. So the Federated Wiki universe will need some way to mark specific forks of articles as being from a given group of people, so people aren't mislead into thinking that the sane editors approve of the version of reality Alex Jones is selling the gullible.
- nicholassmith 14y agoExactly, the mods are there to keep the crazy out but occasionally you need to reference some of the more outlandish things for various reasons and letting people for things off into tinfoilhatswiki.net would be good for the mods sanity and useful for people who need to reference the over the top stuff. Not sure how the link direction would work, but I'm sure there's a way.
- Jhonbxl 14y agoIf the problem is having the knowledge hosted in one place only, controlled by "one" person only, the solution should be to decentralize, not make everyone able to copy it everywhere. Plus, this solution seems to add more work to the contributors, since you'll need to review each fork/edit and choose which to merge or not. On the other hand, a wiki working like a P2P network, where everyone hosts the same version and edits are automatically propagated accros the web would solve the "problem" without adding more job on contributors.
- pessimizer 14y agoCould this be layered on top of the current Wikipedia through browser plugins? And how would discoverability work in the context of 10,000 current versions of a wiki page? It's hard enough when there's 30 different forks on github. edit: federated search, federated social? How would the system prevent search gaming and social gaming?
- morsch 14y agoIt's a bit too early to worry about 10k forks for a single page, isn't it? The typical use cases -- certainly the early ones, when it's not widespread -- might involve numbers of forks that aren't too different from GitHub.
- webwanderings 14y agoDidn't I see something from Dave Winer doing similar thing for Twitter, personal stream?
- roguecoder 14y agoIt seems odd to me to focus on Wikipedia as the comparison point, but then again I was on C2 back-in-the-day and have always seen the quintessential wiki as a superior version of a forum rather than Wikipedia. If we are comparing to currently-active technology, I would argue that the federated wiki concept has the potential to bring tumblr or del.icio.us style contributions to a different, longer-form audience. I do worry that the federated concept will make authorship too important. The lack of permanent credit on C2 was part of what made it valuable: people more often wrote things to contribute value when they didn't have the motivation of scoring points, gaining karma or making a name for themselves.
- sdfjkl 14y agoAt last you can fork the truth and make your own version of it. I imagine this will be a hit with governments everywhere.
- themonk 14y agowebring?
- Vitaly 14y agoInstead of plain EC2 installer t needs a Heroku Installe. Then anyone can have it running for essentially free in no time
- Nux 14y ago"To make Federated Wiki easier to adopt, there’s a one-click installer to deploy a server to Amazon Web Services." <- oh yeah, super-federated.
- sp332 14y agoWhat's not federated about that? And obviously you don't have to run it on EC2, that's just the easiest way at the moment.
- duskwuff 14y agoAs I see it, one of the biggest benefits of traditional wikis' linear history was that it fostered quick back-and-forth collaboration between users. Unless there's a strong mechanism for discovering changes made by other users, a forking model seems like it'd rapidly result in a bunch of different (and potentially non-mergeable) versions of any page.
- DeanCollinsNYC 14y agowhen a third person forks a third time to makes additional changes does it feed back into server 1 as well as server 2? How many tiers can it go eg would I see tier 10’s changes? What happens when I register tier 7’s changes do I still see tier 8?
- 6ren 14y agoWith always-on, always-connected devices everywhere, we are not that far from every device being a web-server by default, and every device being a router for web traffic. Data storage, and even computation may be centralized in the cloud... but why not distributed distribution?
- seiji 14y agoYe olde, but still worth watching: Van Jacobson's Google Talk from 2006, "A New Way to look at Networking": http://www.youtube.com/watch?v=gqGEMQveoqg http://www.youtube.com/watch?v=gqGEMQveoqg
- irunbackwards 14y agoMuch better headline than the one actually on Wired, "Wiki Inventor Sticks a Fork in His Baby."
- pella 14y agosee "In praise of cooperation without coordination: Clay Shirky at TEDGlobal 2012" "The result is Git, distributed version control, which Shirky is here to explain to us. It’s a distributed workflow that brings chaos back into the system. Yet there’s a beautiful innovation to ensure that said chaos doesn’t promptly override said system: a signature that creates a unique identifier for every single exchange. This enables cooperation without coordination. “I tell you this not because it is great that open-source programmers now have a tool that supports their philosophical way of working,” Shirky says. “I tell you all of this because of what it means for how communities come together.”" http://blog.ted.com/2012/06/29/in-praise-of-cooperation-without-coordination-clay-shirky-at-tedglobal-2012/ http://blog.ted.com/2012/06/29/in-praise-of-cooperation-with...