Invite Code:qZ1nKsma

good for 100 people per request

Enjoy TorrentGeeks.com

Made just for users of TorrentFreak.com

Thanks TorrentFreak for the great, Up to date info.

idolwind@gmail.com

THANKS AGAIN

TG is maintained by people who truly understand what it is to be a community. Although I have to agree, I have been a member at sites that are run by little guys with half mustaches.

Bittorrent, at its CORE, revolves around community minded people. Thats its WHOLE purpose, to share amongst connected “peers”.

This is why at TG, you will find Blogs, Profiles, Chat, Comments, Discussion Forums, etc. All of which help to further promote a true torrent community. NON of which reauires VIP to access. VIP only helps support growing server costs, and ultimately will only be picked up by diehard torrentiers.

Private torrent sites, (if ran by the right people) only exist to better distribute quality torrents. Ratio’s are a by-product of greedy leechers.

I have pretty much built a site that people can take advantage of for free, its my way of giving back. However leechers are not welcome, that is the ratio’s main purpose.

Well, if you want to help build a truly great torrent community, head over to TG and see for yourself how this is supposed to be done.

I wont try to speak for other sites that are heading in the wrong direction.

you will need this code, currently good for 100 people, I made it specifically for this site.

TrdoCPuc

Thanks for the great Articles.

Master Geek

Here are the codes:

iHSP4Vn7
yx4AvVmy

No it isn’t! Not if say your upload speed is a 10th of your download speed.

IMHO, your ratio should only be as good as your download:upload speed ratio, but private trackers will never work like that.[/quote]

No worries though, I just discovered versions 0.89 and 0.93 (yet to test 0.90-0.92) send back information on what you are downloading (info_hash and size), regardless of claims made in the privacy policy.

This is more bad news for BC users, on private or public trackers.

This is off-topic here, but I’ll be happy to respond in some place that is appropriate. Start a thread somewhere and invite me to it.

robb (at sign) funchords (dot) com

It is definately implemented poorly though, but the implementation may have more to do with the level of risk than anything else.
Posting a list of IPs isn’t likely to land you in court for any copyright infringement (as the main defendant or for secondary infringement) whereas tying a regularly updated list of IPs to a torrent hash may (it is how trackers operate afterall).

[quote comment="161410"]There are simple defenses that are possible for both clients and trackers.[/quote]
These forms of attack can only be stopped at the tracker end with an extension or change to the protocol.
Prevention would require a way for clients to check each incoming peer, either by communicating back with the tracker or by some other means.

Simply dropping all incoming peers until they’ve been seen on an announce return will slow down start of a torrent massively. Requiring a request back to the tracker for each peer requires more sockets open on the webserver and a lot more work for the tracker.
Selecting 200 peers from the database or file in one go is fairly light in processing. Opening the database or file 200 times and selecting one peer each time is not.
The coding on this one is extremely simple, the consequences to the trackers load and bandwidth consumption are not.

A less intensive way to prevent these attacks is to allow clients to verify themselves, using signed certificates whereby the tracker would sign a string consisting of peer_id/port/IP with a private key, allowing decryption and verification by the other peers using a public key. This one isn’t all that hard to implement if using a library such as openssl.

Either way requires all peers to use the new rules, and it isn’t that simple getting all clients to accept the changes and getting all private trackers on board.

BTW, if funchords is Robb, why did you select different bitcomet versions for “proving” it doesn’t behave as claimed? That methodology is flawed. If you want to claim something is false based on 0.90, then why not also give the results 0.7 or 0.6 as you do in other areas?
Not giving a complete picture suggests you knew what conclusions you wanted to arrive at and arrived at those conclusions any way you could, making your “study” completely worthless.

Your analysis of how superseeding works is correct, superseeds require a HAVE message from any other peer.
However, you decided not to test how BC reacts when it is that other peer and claimed it was false anyway.
TheShad0w was first to implement superseeding, and also the one who banned BC clients from connecting to bittornado, he actually studied how BC behaved with superseeds.
You also seem to forget that peers do not know superseeds as seeds, but as other peers. Superseeds do not allow other peers to know they have the complete torrent, so withholding HAVE messages from seeds is irrelevant.

If you haven’t the time to test all the versions, then don’t push a half complete study as if it’s irrefutable fact.

It works on the assumption that a substantive number of the IP addresses that they’ve been able to capture will be sharing the file that you want.

Right now, they have

* SceneTorrents – ~180 clients
* TorrentLeech – ~500 clients

This sample is too small to be effective, and they appear to be manually collected and updated. The sample collection depends on an insider leaking the information.

The lists provided are not in a useful format. You can’t use these lists, as-is, with Azureus’ Peer Injector nor with the Add Peer feature of uTorrent.

The hack works because each peer in the list will be contacted, and a request for a hashid will be performed. The client will either continue the handshake (if the hashid matches an active download for the client) or drop the connection (if the client is not active in that download).

As not all peers on a private system participate in all swarms, someone trying to use this hack is likely to get a only few peers for their desired download. Overall the download will be slow.

Immediate defenses:

1. The BitTorrent protocol is a Tit-for-Tat protocol that rewards good sharers with faster downloads. Someone using this hack is a cheater, and is unlikely a good sharer, but must use the BitTorrent protocol. Ironically as a result, they’ve just cheated their way into a poor-performing download due to the protocol involved.

2. About once or twice a day, change your incoming port. This will invalidate the information on the collected lists.

There are simple defenses that are possible for both clients and trackers. At the moment, the cost of coding (and debugging) these outweigh the low risk presented by this hack. But if this hack takes off, I’ll be happy to discuss these.

I can go at the police station and say i’m a pirate they wont arrest me!

Do block addservers in a DNS nullzone.
It will save you from being passively probed by for profit organisations
10 – 20 thousand times per host per month.

Example: I taught my cousin how to use BT finally and showed him a handful of sites to get stuff from. He tries to get Evan Almighty from MiniNova, and instead it’s a sweedish porn flick.

Now I’ve got nothing against porn :), but on most private sites they would have that file removed from the tracker, or change the label on it.

I guess the point I’m making is I like the upkeep of private over public.

However note that some private peers will monitor their leeches once they notice their ratio doesn’t improve, compare them to the tracker’s list, and subsequently ban or snub those ips that dont occur in both places on their clients.

Because, there’s a LOT of stuff I can’t find on public trackers for I live in Québec.

I built a proof of concept over 18 months ago for exactly this type of attack, only there’s no need to allow plain text downloading of IPs or use a peer injector plug in.
It even works with every single client out there.

This is as old as the protocol itself, and az already has code to protect users from being swamped by external connections. It just needs finalizing, if any private tracker admins with more time than myself would like to volunteer to help out.

The information is out there regardless of weather torrentfreak links to them or not. You can just google for this and there are 10 websites that pop up on the first page that tell you all about it.

If people are going to cheat they are going to cheat regardless of what others say or do.

pretty low…