12:38 Upgrade | |
Hi, ALL! For sure many of you have noticed that recent attempts to control the Internet are magnified and intensified, that there are many examples - from the closing torrents.ru without trial (and in fact they cooperate with copyright holders!) To demonstration catches of various hackers, what would others sat silently. Yeah. Fuck you. I propose to organize and begin to develop our new (national, hehe) data exchange system, which will be the development of existing torrents, emuley and others. Just say, probably all of this somehow implemented in the various instruments of exchange of information - the goal to make it more affordable, more convenient and more alive. Watch out - a lot of text without pictures. The current problems of network
What can we do about it?I just try to offer something different, but I will be glad and any additions to my thoughts. Cause I can, and fundamentally not be right, right?) The general ideologyAt the moment (it's still only an idea and can do anything cool to trade) has been suggested that the XMPP protocol is very good and use it for the interaction between clients and servers + to use some (well, most likely BitTorrent) technology to exchange data directly. Accordingly, we have a basis:
For us this will be sufficient basis for the eyes and the ears - can be implemented almost everything that your heart desires, so let's try to solve problems using these tools. Problem 1 - duplication of contentHere I propose to just sit and think about structuring hands and mx. At the moment everything is on rutrekere and where many more stores, for example, BB-codes, and mx and this is often not very convenient for cars) It all still need to create distribution on rutrekere fill out the form below. So why not make it all better and not to keep the structure instead of plain text? Problem 2 - Content is not always checked and often zavirusovannymTo do this, I see quite the introduction of digital signatures for different content (mostly -. Torrent files, etc). For this signature can understand what people in the network test it and in fact if anything - questions to him. Perhaps there is an option to organize a multi-level signature - say, the first small signs moderator, and then he should sign someone more responsible, etc. or can be arranged so levels of confidence to the content. To help protect against viruses is the next idea - to make a simple bot that will send content to scan for viruses and they will sign the content, if everything checks out. Problem 3 - Several different formats of content scattered across a networkWhen entered into the search box at rutrekere «House MD» you get just a hundred different hands ... it is sad and need to sit down and choose what you would be conveniently and yet look at the title, like the quality of something, but then it turns out that this is not the season, but not always a good idea to search on more and «Season 3" - can once again have missed something. You simply create a single hand at everything and have the opportunity to select you. Torrent file in one place and do not delve into all this. You can impose all sorts of tools next to - what subtitles are needed, which translate what season, what quality of pictures, etc. For what you could combine hand, simply separate them all to sign - if there is some distribution, then there is more content, then later it adds to your existing file and it gets into the queue for moderation, and This new content (not the whole distribution) will be marked as "not verified" (or, as will be better) until it will not sign a moderator and just everything in step 2. you just sign it all separately and individually to realize functional in step 2. Looks good. Issue 4 - To a more or less rapid jumps still need trackerEverything is harder. Need to strengthen the resilience of the network - it needs to implement the following algorithm to find a new tracker. To do this we need the following functionality:
The first mechanism is needed to ensure that used to understand through what connected when the tracker will fall, and the second when the current tracker is working, but does not want to handle this client. With the second everything is simple and clear - just in announce his return to something like «redirect to ****» all. But the first requires a special server setup - you need to create a network of some of servers (say, will nexus), it is desirable that they were geographically distant. Each server of that Nexus will be a prioritized list of all the little network that he would send all its customers. If the tracker suddenly stop responding, then customers will go to the bottom of the list of servers and connect somewhere. At this time within the network is data and determine who will actually distribute the content now (this happens very quickly - a network of small and simple "agree", and to realize you can estimate some heuristics and come up with some simple and unambiguous rules to determine the successor of the deceased tracker), and then after decided that all redirect to the new tracker. This should not cause any problems setting up - the people, and so set up a backup of your torrents, and now just need to talk to people and the mail server does not need to set up - they all do it yourself. Problem 5 - Not so easy is to distribute on a torrentIt's still important to remain anonymous and to make the network more distributed. Here we propose to make a hierarchy of complexity, but complexity must be such that it only strengthens the network, rather than weakens it. Our trackers have already joined the Nexus and may vary from one another a certain area which can operate until at least one member of the Nexus is alive and it will be stored, respectively, the entire contents of the dead members of the Nexus. Yes, there is a problem with the load, but if you have 20 servers in the Nexus, then such a problem, perhaps, never will be. If so be that the load is too large, you can always pick up the server support, and to redirect part of its customers, etc. In order to identify the server for its participation in the Nexus will use digital signatures, etc. This will protect your network from external influences. A child support server may have a different key and, accordingly, to create Nexus all share a pair of keys - from the primary server and from the subsidiary. Auxiliary can be any number, and the principal is only one. This mechanism is designed to make connections and expand network by one party to the network - that would not have to all communicate and exchange keys. Essentially auxiliary servers are only a caching proxy. But in general they need rather illusory. Issue 6 - No possibility of automatic renewal torrentIt's very easy to implement through the answer to announce, so do not be distracted by it. Issue 7 - The complexity of the initiation of distributing content without the use of hand on trackerTo do this, you can simply implement for each user its own section on the distribution of which he may provide links that do not open without, for example, the generated password. In principle, there is also nothing particularly complicated - trackers always exist and will continue to work. In principle, we can not even keep all the servers and make the tracker directly to the client - whatever the user of this private distribution makes a request to that client through the XMPP server. And it's all the ideas?No, not all, of their weight and just over the goal (or in excess of desire - as anyone) - is a perfect tool for sharing information. In addition to be bound by these there are lots of introductions to all this. For example, create an anonymous semi-automatic method of merging the information into the network (for a-la vikiliks), to create opportunities for private trackers (which used only a few members of this super-network could participate in file-sharing). As for the interface - would like to make a single basic interface and does not want that, it was still just a browser. In a browser, there are few opportunities for real file sharing. And sledstivii this - have a certain hierarchy of nexus that would be seen as the palm of your hand "the entire Internet." For example, a partition on a territorial basis, in the direction of (IT, music and mx), etc. And in the very nexus also create your own some hierarchy, etc. Even think that this would be interesting to start. A MB that would not have participated in the nexus of the hierarchy, and themselves trackers. It is likely that will have to do something like DNS names in the issuance of this hierarchy. But even here already have a lot to think about it like that to organize what would have been no kiberskvotterstva etc. Create an opportunity to create their own free private tracker community. But here the issue of security becomes - the server is all the information there is, in principle, he can send others to the network ... then you can use a mechanism - the "tracker on the client with synchronization of the network over an encrypted channel. In general, lots and lots of good =) Ok. And then what?In general, I propose to assemble a small team and start it all to create. If not us, then others. People are already scurrying around at the expense of distributed different things (eg diaspora). 'll Find all :-)
Waiting for all the comments :-) | |
|
| |
| Total comments: 0 | |