Distributed Authoring Mash-up

I want to bring to your attention an idea that can solve the main problem of Web 2.0 projects: a drop in the quality of published materials and discussions over time.

If you personally do not see a drop in the quality of topics and comments on Habré, then please postpone reading and commenting on this topic until the moment when you also find that the quality is still falling.

Quality. Authors Web 2.0 vs Democratic Web 2.0.

I want to be able to read interesting, high-quality content, without wasting time sifting piles of garbage. Some are even willing to pay for this opportunity.

In principle, a subscription to individual blogs provides such an opportunity, but ... there are few such blogs, they are difficult to search for, new entries in them do not appear often . In addition, people often write on their personal blog, mixed together , and not all topics are of interest to you . As a result, with a good average quality of materials, their quantity is too small . Another problem with individual blogs is the lack of a unified commenting system , which greatly complicates the discussion and, therefore, reduces the number and quality of comments .

However, I believe that the only way to guarantee a constant level of quality material is the name of the author . If you are satisfied with what and how this author writes now, then, most likely, it will suit you in the future. Of course, you can “outgrow” the author, or he can “go astray” ... :) But firstly, all this cannot be compared with the fluctuations in the quality of materials that democracy and plus points with karma lead to, but - secondly you can always solve the problem in one click by unsubscribing from this author.
')
The same applies to the quality of the discussion material. Only the presence of a moderator gives certain guarantees that the quality level of comments will be maintained at a constant level. Naturally, for this, the moderator must have full control over the discussion: the ability to use pre-moderation, black / white lists of people, the ability to delete / change comments themselves, etc.

When discussing the pros and cons of manual moderation, two questions usually pop up: where to find (and what money to hire) good moderators, and what to do if the quality of moderation does not suit you. I have answers to both questions: the authors should moderate the discussion of their materials themselves, and if you are not satisfied with their moderation quality, do not read the discussion or organize your own alternative discussion, which you will moderate on your own.

Amount. Search mash-up against the search for interesting authors.

It takes a lot of time to find interesting individual blogs. Therefore, in particular, sites like habr are popular - here you can quickly find a large amount of materials from different people on topics that interest you. In fact, habr simply makes you mash-up blogs of various authors, and not one, but several different mash-ups (main page, habrant, live broadcast, new, etc.).

But, despite the attempts to prepare a mash-up for every taste (habranth), Habr tries to please everyone at the same time (more precisely - to the majority), and as a result, the quality of materials decreases ... and those who notice it first (and leave) are unhappy coincidence usually those who were able to create high-quality content (i.e., after they leave, the quality drops even more).

I believe that a specific person should make a mash-up of interesting blogs. In this case, the average quality of blogs in his mash-up will be constant . And having connected to such a mash-up you will quickly receive a large amount of materials of interest to you, the quality of which will be constant .

Implementation (Stage 1)

This is how the work of “Author's Mash-up” may look (we will add the word “Distributed” in the next step):

• Any person can create any number of thematic sections, and post in them what he wants - his own copyrighted materials, reprints from other sites, links, etc. He is the only moderator of these sections and these materials.
  Any person can try to leave a comment on any material, but the moderator fully controls both the ability of specific people to leave comments and the comments themselves (including choosing between moderation and pre-moderation). In other words, the author of the material decides what will be in the comments, i.e. it is the only factor that determines both the quality of the material and the quality of comments.
  You can make a mash-up: “add” someone else's material (including his discussion) to your section (and even “connect” the entire someone else's section to yours). In this case, this material will also be seen by your readers, and will be able to connect to its discussion (comments are still moderated by the author of the material).
  If you are not satisfied with the moderation of a discussion by the author of the material, you can create your own, parallel thread of discussion of this material, which you will moderate on your own. In this case, your readers will be able to see both branches (and participate in both), and readers of the original blog of the author of the material will see only the first thread of the discussion. Thus, one material can have several independent discussions moderated in completely different ways, and readers will be able to choose which discussions they will participate in.
  
  This approach will lead to the fact that everyone will have the opportunity to discuss any material at the level at which he wants to discuss it, among people interested in this material and wanting to discuss it at the same level. Without disturbing other people to discuss the same material on another level.
  
  Thanks to mash-up it will be possible to connect blogs in batches (perhaps, later forming your own mash-up from these blogs), and these packs will be chosen by specific people “for themselves”.
  
  And thanks to moderation, the quality of both the materials and the discussions will be kept at a more or less constant level (which initially arranged for you when you connected to this section).
  
  

  Examples

  
  

  Reader

  
  For example, you are not a writer, you are a reader. In this case, you create one or several sections, and connect sections (individual blogs or whole mash-up) to these empty sections of other people that you want to read. Forming thus the selection of the information grouped as it is convenient for you. And then you, in fact, go into your account and read your sections, as though they are filled with other people who do not even know that you have connected them to yourself. Something like RSS. Accordingly, someone can connect to your sections too, if he likes your selection. And no one bothers you, if you wish, also to write something in their sections - but these materials will be seen only by your readers, and they will not get "up" in any way - in sections of other authors that you connect (unless they, of course, connect your material to your own section).
  
  As for commenting, this will be the case: you can try to comment out any material, but your comment can be moderated by the author of the material. In this case, you can save your comment in your own, in a separate comment thread on this material. For yourself and your readers.
  
  

  Moderator

  
  There is a strong section on some topic, it is led by a tough and qualified moderator. Even if he does not write, but only selects articles from his point of view, and monitors the quality of comments - his section will be of very high quality. And if, for example, the author of the article doesn’t follow the comments, the moderator connecting the article can choose the option “connect without comments”, and the readers of his section simply will not see the original comments on this article in the author’s section and will discuss it from scratch at the level specified by this moderator.
  
  

  Questions and answers

  
  Questions asked: Vox , dinamyte .
  
  Q: Weak Link - Moderator. He can get sick, change attitudes, etc.
  
  A: Maybe. And the author of articles can also get sick / change his views. As soon as their section becomes not interesting - we unsubscribe, that's all. As soon as there is nothing to read, we study existing sections / people, subscribe.
  
  Q: We must also provide for the possibility of transferring the powers of moderators ...
  
  A: It is not necessary. If a certain moderator is cool and drives a quality stream, then he will have subscribers. And subscribers - automatically - a local copy of its stream. Thus, they are all his potential doublers and followers of his bright cause. Moreover, they can cross-link this section to each other so that the added one appears to the others.
  
  Q: An author may consider criticism (even if constructive) for a stone in his garden and reject such comments.
  
  A: Has the right. The critic has the right to place in his section of his critical comments. If they are interesting, readers can subscribe to this section not from the author (who has good articles but comments are weak), but this critic (who has the same articles, but comments are more rigid - plus those, or instead of those comments which the author has).
  
  Q: How do users find out about the existence of criticism?
  
  A: Who needs it - he will know. He will try to criticize, he will be moderated, he will go to look for a place where you can discuss this article at his level (which, by the way, can be really extremely low and his criticism can be complete nonsense, this is also not excluded).
  
  

  Implementation (Stage 2)

  
  And now let's make this system distributed.
  • When you connect to someone else’s material / section, a copy of this material is downloaded to you, along with comments (which are updated regularly).
    All your partitions can be located on a common service that allows you to create an account and manage your partitions, moderate them, etc., or on your own computer / hosting. To do this, you simply install one of the available implementations of this system (all implementations will use a common open protocol).
    Each, by definition, will have its own unique url, by which it will be possible to get access to its materials / sections, and the same url can be used to authorize via OpenID, for example. However, there are enough ways to make authorization in a distributed system, it is not necessary to dwell on OpenID.
    
    With this approach, each user of the system will keep a complete copy of all materials of interest to him, completely identical to the author, including comments.
    
    

    Security. Possession of your information against naivety.

    
    If your information lies on a foreign server, you no longer own it. (As everyone knows, LJ got into the SOUP, Habr deprived some authors of access to their materials ... and the further, the more such examples will be.) The proposed system allows you to choose to store your information in yourself, or on one of the available services.
    
    If your paranoia is sufficiently trained, then you may not like that anyone can connect to your sections - i.e. actually to the collections of information that interested you. But I described above just one of the options for the system, so in principle it may be possible to create "closed" partitions to which no one can connect.
    
    Another interesting feature is the guarantee of the integrity of the content (digital signatures). Being connected to section A of person B, you received material from the author C (which he, in turn, could also receive through a chain of other people). But nothing prevents you from connecting directly to the system of this author C, extorting his public key and verifying the signature on the received material.
    
    

    Availability. Distributed P2P systems against one project.

    
    The proposed system will not depend on a single service (as a habr), instead there will be something like Jabber.
    
    Since the content to which you are connected is automatically copied to you, there will no longer be a situation when an interesting article suddenly becomes unavailable after you have read it.
    
    Popular content is automatically backed up on all subscribers' machines, since even if the original author's site closes forever, a copy of its materials will be available in a significant number of sources.
    
    Distribution of identical content across different servers, along with the described flexible subscription mechanism (when many will connect not from the author of the original material, but through different mash-up or well-known moderators), will allow to distribute the load throughout the system. In addition, you can easily create mirrors for especially loaded nodes.
    
    

    Search. Google

    
    All content in this system will be available in several ways: through the internal API for downloading / updating content and comments, sending comments, etc. and through the website. Accordingly, search engines will be able to easily index the entire content of the system for full-text search.
    
    Regarding the search for authors, sections, getting statistics on activity, etc. - for this you can make separate web services.
    
    

    Salvage. Earnings for authors, not publishers.

    
    Honestly, I initially thought that this project would be completely non-commercial, due to its distributed nature. But then I realized that the thirst for profit and the desire to earn extra money would still find a way out. For example, authors can insert banner ads directly into the text of their articles. Well, well, in the end, in this case, the authors themselves will earn on their materials, and not, as now, systems that simply store and display their articles.
    
    

    Features Want a plus sign? No problem.

    
    Although this system should work fine without the need for pluses, you can enter them without problems: just each comment will have additional meta-information about the number of different pluses. And since all this is optional information, the user may have a choice - to see / control these pluses, or not to fool around with this nonsense.
    
    Similarly, since article formatting will use a predetermined minimum of html markup, you may have the ability to control via CSS the appearance of other people's articles that you have connected to your section.
    
    

    Todo

    
    Content structuring. I know for sure that working with high-quality content in the style of blog entries (like in Habré) is extremely inconvenient. The format of the blog is suitable for entries in the style of "what I ate for breakfast today" and for news. And how to organize content more conveniently - you have to think.
    
    It may be worthwhile to realize the support of group moderation, so that there is not an authoritarian opinion about what is happening, but a multipolar one. But I still can not imagine how this can be implemented, and whether it is necessary.
    
    PS I remind you that all of the above is just an idea that has been thought out for less than a day. I hope you will help identify weaknesses and develop an idea - or suggest an alternative solution.
    
    Update: Initially, this idea arose and was discussed in the closed topic CurlyBrace "Tired" . Now the topic is open, if you're interested - read, see where the "legs grow."
    
    Update2: Read continuation: Mash-up, tags and bloody gebnya (tm) .