Anatomic P2P GUI Client 0.1
Beta Release: http://developer.berlios.de/project/...?group_id=2947
p2pnet.net News:- Anatomic P2P GUI Client 0.1 Beta Released …
… “at last” it says ; )
Anatomic P2P is, “an attempt at a true decentralised BitTorrent Network,” says the site.
“It works on a (supernode) caching system based on BUT NOT HAVING ANY OTHER RELATION TO the Gnutella system of GWebCaches with modified FBT2 trackers written in PHP and a modified BitTornado client written in Python.”
Anatomic P2P
p2pnet.net News Feature:- Anatomic P2P is an attempt at a true decentralised BitTorrent Network. It works on a (supernode) caching system based on Gnutella GWebCache with modified FBT2 trackers written in PHP and a modified BitTornado client written in Python.
RespectP2P's ToM interviews Anatomic P2P's developer.
Read on >>>>>>>>>>>>>>>>>>>>>>>>
Anatomic P2P - Decentralised BitTorrent
By ToM - RespectP2P
The success of BitTorrent has been staggering for both legal distributors and those interested in less-legitimate methods of distributing works online. Torrent usage is on the rise despite some hefty legal action taken against certain torrent sites.
One drawback to BitTorrent is its use of specific “trackers” which direct users to other users sharing files in a torrent swarm.
In this interview, I speak to the developer of Anatomic P2P, covering network specifics, the future potential of this interesting project, and a some other basics.
Unlike recent attempts at cloning BitTorrent, Anatomic doesn’t exist for the sake of generating a profit for a company and doesn’t try to act as “yet another P2P network”. The ideas the Anatomic P2P project bring forward makes its success a real possibility.
Q1. First off, why decentralized BitTorrent? Which factors made you decide to base your protocol around the torrent concept?
The first time I used BitTorrent I was amazed by its speed in distributing large files across thousands of users (especially in distributing Linux ISOs). I believe BitTorrent is the best way of file distribution on the planet at the moment and its success reflects the technology involved.
BitTorrent is a proven protocol and so I based Anatomic P2P around it.
I know quite a lot about BT and it seemed like a challenge.
Q2. What interests you about file-sharing? Which factors lead you to start coding P2P-based applications?
File sharing was something amazing the first time I used it and I was amazed about the content I could find on it (legit and not legit). In the last few years I have learnt so much about file-sharing protocols and I think it is my turn to attempt to give something back.
Q3. What are the obvious similarities between Anatomic and BitTorrent? Which features will users accustomed to using BitTorrent recognize?
Anatomic P2P uses the standard BitTorrent 1.0 "Peer Wire Protocol". In coding Anatomic P2P I have tried to keep it as transparent as possible. I have kept the changes at the server-side and not added very much to the BitTornado client that Anatomic P2P is based on. The downloading system is exactly the same, just the way that you find out about trackers and users is different.
Q4. eXeem is an application which claimed to be "BitTorrent but decentralised". It, however, established a poor reputation for itself by remaining closed-source, bundling spyware and lacking any truly innovative concepts over the original BT protocol. In what ways will the Anatomic network differ from eXeem's attempt at a "decentralised BitTorrent"?
eXeem had one node and at times this node was down. I would like Anatomic P2P to have far more than just one node, perhaps even dozens. In my opinion eXeem was inefficient - every user was a tracker. I found this stupid because if one user left the network, another 100 users had to be notified for nothing. This is why Anatomic P2P has only 2-4 trackers in use for a torrent. Being a Linux user I was dissappointed with eXeem, not open-source, bloated, spyware when running on Windows - I nearly went as far as attempting to reverse engineering it.
Q5. In what way is Anatomic decentralised? Does it rely on bootstrap or is it initially based on servers? If so, is it only one, or several main nodes?
The concept behind the caching system that Anatomic P2P uses was based on Gnutella GWebCache ("http://www.gnucleus.com/GWebCache/"). Every Gnutella Client has a list of around 20-30 caches stored internally. The Gnutella client connects to these caches to find out about hubs, which are then connected to by the client. In Anatomic P2P I have kept the caches but replaced hubs with trackers and info_hashes. In essence the supertrackers or supernodes or nodes (I kept changing the name) track trackers the hashes on the trackers and the trackers tracker the hashes and the peers/seeds.
You can see my test backend at http://anatomic.berlios.de/network/.
See specification for more info. Bear in mind that there is no link with Gnutella at all.
Q6. What sort of rating system do you plan to have in place to avoid fake, corrupt and industry planted files, etc, if any?
BitTorrent rarely has fake files and the Anatomic Network will use torrents as per usual (except the torrent will contain a dictionary key {"planted" : 1}). I don't think any rating system will be needed.
Q7. How will Anatomic users find files? Will searches be confined to torrents on this particular network?
Torrents will be used (see above).
Q8. Will hash link indexing sites be a valuable or even necessary addition to the network in order to distribute verified files and new content?
After a poll on Filesoup a month or two ago when I was in concept stage, it was concluded by a huge majority to keep torrents and not have a search system.
Q9. Will any anonymity or encryption features be included for privacy protection? How about protection for seeds (the first to share a particular file)? If the network takes off, do you believe its users will be as venerable to industry warnings as on other networks?
The trackers and supertrackers know little about the files they are sharing - I used FBT2 and compact tracker especially because of this. In the eyes of a tracker it is only tracking a hash and this is less incriminating. The tracker/supertracker cannot control what it tracks so it therefore has the same status as a router on the internet (in theory). Encryption and seed protection is something I may need to investigate in the future.
Q10. Is Anatomic backwards-compatible with BitTorrent, i.e. will the client be able to accept regular .torrents from around the web?
Normal torrents will work perfectly well in the Anatomic Client.
At the time of this interview I was working on making 'normal' clients compatible by using a wrapper script that redirects to a tracker. I have tested this on a freeware torrent called "SpywareBlaster 3.3" and I had 16 seeds at one point but now it has gone to 4 seeds 3 peers at the moment.
It is a bit of a hack and only one tracker will be supported unlike the Anatomic client, which can use multiple trackers.
Q11. How will your project be funded? Via Bram-like optional donations or something less-savory?
My project will be a hobby for now (something to think about and hence stop me falling asleep during boring lessons ) and if it is successful I may introduce optional donations for servers.
Q12. Where do you see your project in the future? Do you plan on continuing its development for the foreseeable future or handing it over to someone else?
I'd hope that people will use Anatomic P2P for their own use. Future development is something I have been thinking about. It could be like the official BitTorrent where Bram is in true control of the protocol and everyone works around him. On the other hand it could be like the linux development process where everyone chips in. I would like to strike a balance between the two.
Q13. Finally, I noticed you have produced a torrent info application called TorrentParse. Do you have any other past projects you'd like to mention, or any projects you plan on creating in the future?
Writing a P2P protocol/network was something I wanted to do for a while now. TorrentParse GTK was a TorrentSpy like app for Linux (but then I found out there was a much better program out there called GTorrentViewer). Learning PHP-GTK was interesting and I would like to single them out and wish them well in releasing version 2.0, which should add some amazing capabilities to PHP.
Q14. Have anything else you'd like to add?
Thanks to: developer.berlios.de
Everyone at Filesoup
I should release a tarball of (useable) code very soon. From then I will make a proper Release Candidate.
I would be grateful for any hosting I offers because hosting P2P nodes on an open-source site is tough. I have to be very selective with what I seed. My email is: kunky[at]mail.berlios.de
Go Here For More Information about the client ]http://p2pnet.net/story/4390.
Source: http://p2pnet.net/story/4390
Download Client : http://developer.berlios.de/project/...?group_id=2947
Reply With Quote
Bookmarks