About the architecture of the distributed blogosphere

    The policy of the new owners of livejournal does not leave indifferent the owners of cozy Zhzhechek. Someone without much talk starts up a sandalone blog , someone repeats the mantra " In the fagot soup. " Someone goes to Goa and is not steamed.

    Recently, the esteemed Nestor saw the counterfeit copyright badges in his journal and decided that it was time to do something. To begin with, he closed access to all his notes and called Mamut a maserfaker. But it did not help. Then he called on the progressive public to build a distributed 2.0 sphere .

    This statement of the question seems to me the most correct, and here are my 5 cents in the piggy bank of ideas.

    A distributed blogosphere should be divided into strictly delimited entities.
    • User ID
    • User content
    • The container (or server).
    • Search server.

    Let's consider them in more detail:

    User identifier (UID).


    The simplest and most key component. In UID form, this is url. In content, the UID should include a set of standard blogosphere interfaces:
    • Open user ID. (so that it can always be authorized).
    • RSS feed for the blog.
    • some other standard interfaces.

    It is very important that the user has the ability at any time to change the DNS settings for his UID and redirect it to a new server. The best guarantee that the url (or UID) will always remain with the owner is to have it in the form of a clean domain (without additional directories), which is registered to the owner of the UID. Although this does not have to be something pathetic like tema.ru. Geographic third-level domains are in bulk and free.

    User Content



    This is what is now being made from an empty LJ account - one thousandth. Namely:
    • Structured profile information. Name, nickname, emails, ICQ, birthday, etc.
    • Avatar (s)
    • A set of static pages.
    • Blog Feed
    • Comments on blog posts, pages, photos, etc.
    • Media storage. Photos, podcasts, videos, etc.
    • Friend lists and security settings. Who sees what / who what can.
    • Page design (probably in the form of CSS).

    The user content format should be standardized. Those. for example, the structure of directories should be defined where XML and media files are laid out in a strict order and which can be packed into a neat zip file.

    The container (or server).



    Iron, infrastructure and software that can host the content of one or many users. It should look something like this:
    • register on the server.
    • redirect to the DNS server of its UID.
    • upload a backup of your content to the server (the format, as we recall, is standardized).
    • we get our cozy ZhZshechka at the usual address.

    Sad, offended:
    • we take backup of content. (which we certainly did regularly)
    • register your own hosting
    • Deploy the standard (and open) server software.
    • we redirect our UID to our DNS hosting.
    • upload backup of our content to ourselves.
    • we get back our cozy ZhZshechka at the usual address.


    Who will keep the content server? I see at least three obvious scenarios:
    • a small server on your own hosting for your beloved (and maybe for a few more friends).
    • service for a monthly fee. Gentlemen pay 5 tanks a month and are not interested in availability, connectivity, business continuity and emergency recovery.
    • service for advertising. The boys allocate space for banners and text ad units and do not bathe with payment.

    Search server.



    This is a server that knows the profiles of a large number of users (all?) And can answer questions like:
    • find me Vasya Pupkina, who studied at the 13th school of the city of Mukhosransk, graduating 67 years.
    • find all users with an interest in “dried chamomiles”.
    • find all the girls aged 14 to 20 from the city of Verkhny Podgorsk who has been online for the past 7 days.

    The search engine plays the role of the unifier and some common denominator for the blogosphere. We go there to see others and to show our own. Pay attention to the enormous role of people search for current social network operators. It is this functionality that binds users to the service and "cementes" the audience. Having separated the search from the servers and content, we make it a cute and harmless helper. After all, search servers can be somewhat independent.

    Technically, a software server can be quite simple. Monetization is obvious. Over time, the search function will be gladly taken away by large search companies.

    Thus, we have a distributed network of content servers. Part works on a commercial basis, part amateur. We also have a blogosphere population. Each has its own name in the form of url. Which no one can take away. Each user grows the content of his blog on a particular server. However, the user is not attached to his server as serfs, but is free to pick up his content and go where it is better. For other participants in the blogosphere, such a transition will be completely transparent. Url has not changed. All this life is watched by search servers. Collect and cache open information. Help users find each other.

    Everyone is happy.

    PS: I wanted to publish a blogosphere on my blog, but something is broken there. gives out some kind of error about karma. : - /

    Also popular now: