Hi, this is pretty cool! I'm still a bit confused on the high level picture. Can you answer:
1) Is this an end-user program (eg, everyone is expected to download and run on their laptops and PCs), or a program that is run on a servers and accessed by users via a web browser? 2) Can this program run standalone (C, C++, Java, Python), or is it a web application (PHP, JSP, CGI, etc)? 3) What's the relationship between this an Jabber? Is XMPP just used "under the hood" and thus merely a design detail, or does Bouillon somehow depend on the Jabber network? Basically, I see some cool stuff and I get the sense it's a p2p wiki with trust metrics, but I'm still a bit confused. Like, why did I need to log in with the foaf/foaf account -- why not just have a "public" mode where anyone can view? Or, how would I go about creating my own account -- can I create an account on your server, or need I install my own server? So I see some neat stuff, but I just don't "get" quite yet. -david > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:p2p-hackers- > [EMAIL PROTECTED] On Behalf Of Victor Grishchenko > Sent: Tuesday, June 27, 2006 9:57 AM > To: [email protected] > Subject: [p2p-hackers] Bouillon 2 project: p2p wiki with reputations > > Hi! > > I'm presenting what could be described as "editable web" or "P2P > WYSIWYG wiki with reputations running over a social network", or just > the Bouillon project. > > - Bouillon allows P2P editing of arbitrary XML (XHTML for now). But, > there are no single global version of any page. Each user's agent > retrieves and assembles pieces of a page from user's friends, friends- > of-a-friend, etc > - If Bouillon user likes a page or a new piece of a page or a new > version of an existing piece (i.e. considers it relevant) then (s)he > may confirm that page/piece (i.e. issue an opinion) and let it > propagate further in the social network > - If a Bouillon user considers a piece to be trash/spam then (s)he > may suppress propagation of that piece issuing a negative opinion. > - Users may edit existing pieces or insert new pieces using their Web > browsers or any other client if implemented. > Bouillon employs wiki-like addressing scheme, e.g. http:// > your.bouillon.gateway/page/Page_Name. Different from wiki, pages are > not local to any server because content freely propagates via social > links. So, the actual Bouillon address is 'Page_Name'. What you see > at that address depends on you and your social vicinity (friends, > their friends, etc). > > Under normal conditions, Bouillon may retrieve pieces and pages > recommended by a friend of a friend at least (i.e. two steps away). > So ideally, due to the six degrees of separation hypothesis, we'll > need just two intermediary recommenders to get Bouillon content from > any given point of social network to any other given point. > > Bouillon is implemented over XMPP (Jabber) network and uses existing > IM contacts. The application itself is an external component attached > to a Jabber server. The core algorithm of Bouillon, oc-co engine, is > content-agnostic and able to process any XML. > > Bouillon demo site is online at http://oc-co.org:8000, see e.g. > http://oc-co.org:8000/page/Bouillon_Manual or > http://oc-co.org:8000/page/How_Bouillon_works (this page is > unfinished) > Note: the AJAX part works under Mozilla/Firefox only! You may use > guest login 'foaf' with password 'foaf'. > See the project's blog at http://oc-co.org. > > Any questions/ideas are welcome. > > > Yours, > > Victor Grishchenko > research fellow > Ural State University > Russia > _______________________________________________ > p2p-hackers mailing list > [email protected] > http://lists.zooko.com/mailman/listinfo/p2p-hackers _______________________________________________ p2p-hackers mailing list [email protected] http://lists.zooko.com/mailman/listinfo/p2p-hackers
