Building the new Wikipedia

Wikipedia seems to be thrown around as a potential app that could be built using Ethereum. I love the idea! Ethereum seems like the idea platform for an encyclopedia that is truly run democratically.

How does every one see this working? I would love to jump into the code and make a MVP, has anyone already done this?

Comments

  • avsaavsa Member Posts: 68 ✭✭
    I believe this would probably work better for wikidata, at least at first. The scary part, for me, is that you can build a meta-data layer on everything, even including some people's personal private matters that they rather keep hidden, and there's nothing anyone can do to stop it.
  • mquandallemquandalle FranceMember Posts: 50 ✭✭
    What's wrong with the current Wikipedia?
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @mquandalle for this kind of thing, dont think 'what is wrong with it' think 'what could be better about it?'.

    You could make a decentralized version of it, for instance. I dont think ethereum itself is a platform for that in particular. But maybe it could monetize aspects like serving it and paying authors.

    You could also try figure out how to make a wikipedia that branches more, and then tries to recombine branches. (or that A/B tests, but i'd be scared of creating a 'human-pleasing-pedia') One that instead of encoding consensus(with some 'criticism' etcetera sections indicating branches), one that encodes the ideas that are put in there.

    For trying for more 'transparent' monetization... I think going for authorship in general first is a better approach. (but i might be wrong)
  • MichaelSmithMichaelSmith Member Posts: 79 ✭✭
    edited April 2014
    Wouldn't it bloat the blockchain?
    Post edited by StephanTual on
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @MichaelSmith? the wikipedia thing, i wasnt really talking about blockchain stuff.

    The 'earning from authorship', well keeping the content itself off is completely neccesary. 1) crypto commitments H(A) and a system where derivative articles are identified, and 2) a system with 'author caretaker' groups where they make deals with authors, and it is up to the viewers to ignore groups that are non-productive, plagiarizing, or that have bad deals with authors.

    1 probably heavier on the blockchain because it tracks each article and claims about it. It might be able to drop data once if there is no interest in it anymore, if there are good solutions for using judges or something to show that two articles are derivates of each other, it may well be better as it possibly takes less judgement from viewers.
  • spencermountainspencermountain torontoMember Posts: 3
    I think this is a serious idea, the zero-permissions text-editing bit. Wikipedia lacks an economic model, a web-of-trust system, a transparent authority, and a voting system. These are pretty real flaws.
    cheers
  • spencermountainspencermountain torontoMember Posts: 3
    edited April 2014
  • CerereanCererean Member Posts: 10
    I think going for a torrent system for the articles themselves, and tracking it and handling logins with the blockchain - similar to how (I think) Twister works - would be able to work. How you merge changes in I don't know, though. Perhaps multiple versions of articles could be maintained, and those which are judged to be less worthy trimmed from the system?
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @Cererean it is already the intention to couple Ethereum with a DHT, it is quite likely what you are thinking about, or even where your thinking is leading is already planned to be possible..

    I wouldnt call it a 'flaw', its good ol' wikipedia, there was _no_ way they could have done better when they made Wikipedia. It is just an enormous improvement that we may be able to make now. :)

    @BlueMeanie where is that coming from? I suppose, i havent looked at the code of it, but i suppose probably there are currently nodes that help you discover the network, but you can tell it to point at a particular node..

    The stuff about NameReg is really a complete issue itself.. The problem is that, well, personally i think that if you do not point out an initial NameReg(which is not a node, its a contract, and entity that lives on the blockchain..) there is going to be chaos in the naming system, at it will be at the detriment to most people. The point of URLs is that they uniquely point to something. It may lead to some selfish bastards holding the keys to the naming contract.

    Better let the Ethereum team, probably with discussions from others make critical bits like that. They are weary about their future selves, both as potentially being corruptable, or because people will try influence them. We will be very careful about these contracts.

    <b>About 'data storage limits'</b> in contracts, yes, very limited. On the DHT, barely any limitations whatsoever. (also with the combination, of contracts and DHT, you get more than storage)
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    edited May 2014
    Yes it is.. The bottleneck is storage, not CPU. But yes, longer term, scalability is an issue. <a href="https://forum.ethereum.org/discussion/676/youtube-hard-problems-in-cryptocurrency">We are aware of it</a>.(https://forum.ethereum.org/discussion/676/youtube-hard-problems-in-cryptocurrency, wiki page https://github.com/ethereum/wiki/wiki/Problems)

    Probably if scalibility becomes an issue, probably some activities that become too expensive will shut down, and people will start using more bare-bones contracts for things, and/or use tricks to decrease storage usage.(edit: or somehow make use of storage outside the blockchain, like via Merkle trees) There may also be fewer full nodes,(expect there to be lightweight clients) but the intention is that it is always feasible to run a full node. Averting centralization because of that reason is high on the list, so probably storage costs would go up..
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    I am assuming there is a price scaling system. If you program ethereum to increase the pricing with increasing load, eventually contracts become too costly to run. Considering centralization of Ethereum is one of the things they fear most, it seems likely the price will go up dramatically.. And the effect will ultimately be increasing efficiency and decreased use of contracts.

    Also, of course, they are thinking and working about those problems.
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    @BlueMeanie If you find a solution to something, it isnt necessarily perfect. We found a solution to doing quite a lot of things with contracts without central authority. It cant handle a billion people yet. The first airplane couldnt fly across the Atlantic.

    Really, when i say 'we found a way to do contracts without central authority' we really opened a can of worms. All sorts of open problems in there, everyone is invited to look at them.

    Of course we live, always have lived in this can of worms. Contracts have always been around us. Before they flew across the Atlantic, they used boats. Before cars, there were bicycles(still an excellent alternative, if your country hasnt screwed up the vehicular environment), horses, carts and walking.
  • chris613chris613 Member Posts: 93 ✭✭
    @BlueMeanie‌ Even a reasonably secure oligopolic public ledger is still better than a trivially falsifiable private ledger for many use cases.

    I think ultimately what you will see is more pooling. You can sign up to a pool where you simply contribute CPU shares for portions of the work that get farmed to you, which is not the entire work of the block, but just the execution of say 10 contract messages in that block. Or you contribute storage, which is not ALL the storage, but a subset which the pool operator can request from you as they need it. These sorts of schemes still erode the decentralization, but they are a step forward from the status quo.

    I think we might be OT for this thread.

  • chris613chris613 Member Posts: 93 ✭✭
    @BlueMeanie‌ I share your concerns, and absolutely do NOT, in any way, believe in the one-chain-for-all-things model.

    To answer the technical question about computing the contracts, the full nodes resolve the SLOADs by querying the pool's p2p network.

    I definitely don't feel like we have all the answers to these sorts of concerns. I'm watching ethereum primarily because I'm interested to see what the hell happens when it goes live. Will it grow fast enough to strangle itself?
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    Anyway, but as long as the activity on the network is not too much, people will be able to run full nodes, and eventually this problem will be solved.

    Bitcoin has a lot of activity, and I can still easily run it on this laptop. (which is hardly high-end) Ethereums does more storage, yes, but it isnt really much if you look at the typical contract, and protocol has some advantages aswel.
Sign In or Register to comment.