Slashdot Mirror
Trackerless BitTorrent Beta Posted
jgarzik writes "BitTorrent development is occuring at a furious pace. At the beginning of May, an Azureus update added distributed tracker and database features. Yesterday, Bram updated BitTorrent to include support for trackerless torrents in the new BitTorrent 4.10 beta."
What, exactly, does this mean for the state of legal and illegal torrents? How long would this take to fully implement?
If you lower the cost of entry to producing a BT release, won't that mean more .torrent file swimming around? With the increase of different torrents everywhere, won't that dilute the power of BT?
Is it legal to post only in questions?
More specifically, how can you connect to a torrent download if you don't know where to start? Isn't the starting point the same as a tracker?
-dave
http://millionnumbers.com/ - own the number of your dreams
I wonder what's going to stop **AA from shutting down the login servers. Sure, there might not be trackers to shut down, but a network is no good if nobody can join it. How do you expect to find out who your "peers" are otherwise?
...what happened to btefnet et al? I mean the MPAA could still shut the site down b/c they were hosting the torrent file right?
I think we'll see two things:
1) **AA will squirm for a while
2) **AA will work harder than before to moniyor and restrict user rights on the internet, via congressional purchasesing, er, I mean lobbying.
I think #2 will ultimately be futile in that it will not slow their loss of control over media content distribution (and copyright violation) but it will make life unpleasant for many...
uR iGn0ranc3, Their Power
This is realy the cat and mouse game at it's best. BitTorrent is getting better each day. While the RIAA and MPAA is closing the hosting website, Attacking ISP from around the globe, etc.
Is this a combat to the death ?
I guess nothing will beat private exchange ? (DRM)
Went to download an upgrade bittorrent.
I was a bit surprised that the download for the upgrade didn't have a bittorrent option. Isn't that ironic? or did I miss the link on bittorrent.com?
is the publisher traceable? like is the ip address in the .torrent, cause that might be a bit of a giveaway.
not sure how it'd work otherwise, but this gives each torrent a single responsible party for its uploading. on the plus side they could limit who has access to the download client tables to people who need it and upload valid.
curious, and no im not just using it for legitimate torrents, but i pay for my cable and id rather keep stuff on my file server than a tivo with a crappy interface.
The first rule of USENET is you do not talk about USENET.
What's needed is some kind of distributed HTTP overnet that works; that can handle dynamic content semi-intelligently, and MUCH faster than freenet/frost sites.
Something like i2p?
Update: it seems bt mainline uses khashmir instead of the azureus protocol. This is a bad thing. If this reaches a release, we'll have a case where two bittorrent clients are truly incompatible, and the result may cause difficulties for the technology itself.
The BSA, MPAA, RIAA only has to go after a handful of very large network providers before then can put a large dent into various P2P networks.
Hitting some of the larger college campuses would be a good start. Some colleges will fight, but until the precedent is set, others will block, and the highest bandwidth users will be offline.
Video Phone Blogs send video messages straight to the web.
All this work for a less than honorable cause. Just think what could be if all this human effort had been channeled through a charity, say Habitat for Humanity, your local food bank, or teaching someone to read.
Inefficient network use also leads to waste of money - which could be used for charity. And you're forgetting of a fundamental right that all humans must have: Freedom of speech.
There are a lot of people--I can't say whether this is true of the BT developers or not, as I don't know them--who are interested and drawn to projects that have a hint of subversion as well as technical challenge to them. Given the popularity and rate of development of such projects, this seems rather obvious.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
A much more interesting but similar system is the dijjer project at dijjer.org.
Like this it's a distributed publishing system without any sort of tracker, but without torrent files either. In dijjer you make requests from your web browser through a proxy server that's your interface to the rest of the system.
It's different in that all of the data being distributed exists in a single system, not in grouped systems of people interested in the same file. Therefore there's a lot less concern about there being too few peers signed on to make the system work.
Before it took time, patence and know-how to get a release up and going. Now it's suddenly going to become so easy to distribute stuff with BitTorrent that people will start putting up fake virus/spyware/corrupt files because it won't take any time or knowledge to do so. Releases distributed with BitTorrent has always excelled in their quality when comparred to their P2P (think Kazaa) counterparts. Now BitTorrent will suddenly become as bad as Kazaa, bogus files, destorted music... it was good while it lasted, BitTorrent.
Anyone who has read up on DHTs will know that there a solid, theoretically proven, distributed storage system. However, they also have two flaws: neither fuzzy searches nor load-balancing can easily be done. For bit-torrent only the latter matters, but Id still like to know how the nodes (A constant n number of nodes, according to the linked article) that are assigned the torrent for Star Wars Ep III are supposed to survive the onslaught of downloaders.
It really won't be that big of a problem. Traditional centralized trackers will still be used most of the time, and distributed tracking as found in the new bittorrent and azureus is used when the centralized tracker goes down. Hence in the short term, there will be little problem. Furthermore, the more popular protocol will be the one adopted by all. Azureus has a head start. The question is, since Bittorrent came out with a distributed tracking system after Azureus, why didn't the developers just postpone Bittorrent's new release until it was compatible with the Azureus protocol?
It sounds like they are both doing nearly the same thing, so if somebody beat you to the punch, why release a slightly different but just different enough to be incompatible implementation of distributed tracking?
True, ideally they'll hold back their release. Of course, the azureus devs have yet to document their protocol, which makes interoperability difficult. I've also heard that mainline's implementation was already essentially done when azureus released.
If he's doing things exactly like Azureus, then a torrent file can be retrieved from anyone known via DHT to be part of that swarm... it's called a magnet url.
This is the reason why DHT, as the monkeys released it, is a Bad Thing(tm). They should've err'd on the side of caution and assumed torrents were "private" unless explicitly marked otherwise. Because they added the "private" flag to the info dictionary, sites cannot retroactively privatize their torrents -- it changes the info_hash, which is the exact reason why the monkeys put it there (where it technically doesn't belong.)
I wouldn't touch that with a 50ft pole. My money says that swarm is being monitored for the next round of John Doe lawsuits. (esp. with the recently inacted laws.)
(all you have to do is join the swarm and sit back and log all the IPs reported by the tracker and from all the inbound connections.)
Well, I've done habitat for humanity too.
A week in 96 degree sun building houses for the homeless.
AND I also like BT.
I agree the artists need some money to keep working. I disagree that they won't write or create new art unless they get millions of dollars. I really disagree that the middlemen who do nothing that can't be replaced by BT should get rich. I donate money to artists (via magnatune among others) where I know the artists are actually going to see a majority of the money and I've established that I like the art.
I also try some stuff, don't pay for it, don't bother to delete it but never listen to it again.
There is now more quality songs/art/tv shows/movies than I could watch/listen to if I spent every day from waking to sleeping consuming it. Only monopolies are holding up the prices- but the glut is coming and prices will drop.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
I don't know why a trackerless mode was chosen, I thought that the efficiency of BT is due to the centralized tracker. I think it would be better to provide redundancy to the tracker function by adding a super tracker functionality.
.torrent file is the real problem in hosting files. Its not as easy as just providing one directory and every file in that directory gets shared. Ofcourse there are benefits also to the .torrent file when we want to serve a whole directory as a single torrent. An approach where both kinds of things can be done will be better than a single method.
Actually the centrallized tracker is a very important thing. It decides who downloads what. Without the central tracker the effort will not be that synchronized.
I was expecting the development to be towards making the tracker redundant, with creating a super tracker, that would track the tracker.
Also the
Also the Emule has it better that it can determine that multiple names of a file are actually the same file, based on the same Hash.
I would think it would be better to have super trackers track the trackers, with multiple super-trackers tracking the same tracker. And each super tracker would be tracking multiple trackers. Super trackers would provide the search capabilities, and would share tracker information among themselves. They would also provide tracker redundancy. They would also be able to determine if the different file names are in fact the same file, and merge several trackers into one.
I think the peers with good bandwidth and with maximum completed parts would become the tracker. The benefit of being the tracker would be that you get the file faster, because the tracker would obviously give itself the benefit. Then when the tracker has completed its own file. A new tracker would be selected.
What do people here think?
You have just given yourself the answer you were looking for. Freenet makes it *very* difficult to track down the sources of files. If you're downloading music or videos, it is sufficiently anonymous for what you're doing.
But as is pointed out on several sites discussing Freenet, if you're a dissident trying to release information, you could still be in for a whole lot of trouble...
http://jcsnippets.atspace.com/ - a collection of Java & C# snippets
The MPAA already has a semi-automated system for notifying network admins of allegedly infringing bitorrent traffic on their networks. I've seen this on a University network. They even have a standard XML format for submitting the notifications. AFAIK, they haven't sued anyone yet over bitorrent, however, but they *are* watching end users.
You might think that -- and that's the BS answer Azurues devs give too. However, such an attitude is pure blind ignorance. How many private trackers are their in the world? How many torrents are on each of those sites? How many users are part of those communities? Applying the azureus specific hack only to new torrents does nothing to protect all the existing torrents. Regenerating every torrent does nothing to protect the existing torrents -- it stops them from being accepted by the tracker, which actually makes it worse as the fallback is DHT.
Requiring, nay, demanding sites regenerate all their torrents is a lame answer and a dangerous precedent -- do you want to go through this again every 3 months when some other idiots do something stupid like this? It's also impossible. You're talking about recoding millions of torrents, forcing every single user to re-download each torrent (not just the az users, every f'ing user), and deal with the information leakage and "lost stats" for users who don't grab the new torrents before DHT hands out their personal and unique torrent. The Azureus developers really failed to give this shit any thought at all (which is all too common with them.) [where's the support for specifying peer sources per torrent, for example?]
(a) "You first. If you eliminate all your activities of which I disapprove, I'll reciprocate."
(b) No one is obligated to give. That's one of the things that freedom means.
(c) I give already in other ways. I have given enough.
"We reject as false the choice between our safety and our ideals." --The American President (20.1.2009)
True anonymous p2p filesharing will never be possible - it is ALWAYS possible to find out who you are downloading from.
...
Suppose, using some new hypothetical p2p program, my client uses one network, say, Gnutella, to search for a title. Using Gnutella, my client downloads a file of instructions that describes how to reassemble what I want using various numbered blocks. (For example, a block's number might be its SHA-256 hash) Next, my client searches the network, maybe using a completely different network or protocol, for each of the block numbers. The downloaded blocks are labeled with a B, as in B58273838922837389. The reassembled content file, the file I originally searched for, is made up of blocks labeled with a C, as in C1, C2, C3, etc.
So the file I want is reassembled, according to the list of instructions, like this....
C1 = B166 xor B224
C2 = B338 xor B426
C3 = B872 xor B998
C4 =...
C5 =
etc.
(Drawback, I used double, or triple or more, of the bandwidth necessary to download the file.)
So which IP did I get the infringing content from?
Remember, each block could be found using a different mechanism, Gnutella, OpenNap, Http, etc. Each block is just a bunch of random bits, indisginguishable from noise.
Well, the beginning of the file, C1, was created from blocks B166 and B224. (Of course, they would have much longer block numbers.) But block B166 combined with some other block on the network results in part of The Declaration of Indepencance. And block B224 combined with yet another block, results in part of The Bible. So was B166 or B224 infringing?
And which IP address gave me the infringing content?
The gnutella node that gave me the reassembly list didn't give me any actual infringing content, just a bunch of numbers. I suppose that the reassembly list could also have been a file that was recursively shared using the Blocks scheme I describe here. Thus I might have to reassemble something, only to find out that I have reassembled a new reassembly list (as long as I knew up front that this would be the case).
I'll see your senator, and I'll raise you two judges.