The use of indirect evidence as "proof" of downloads is known, the interesting bit here is that in spite of pushback from ISPs and users, industry practices have not changed.
Perhaps this (and the widespread lack of privacy in cloud-based services generally) will drive more users to privacy-preserving data sharing options, such as OneSwarm.
Indeed, it's true that BitTyrant will not always improve performance, even when directly compared to other clients on the same torrent at the same time! There are several reasons for this:
Data availability: Suppose seeders are sending out data at 20 k/sec, and a complete copy does not yet exist in the swarm. Then, even if you start downloading quickly, eventually you will catch up to the data production point and download at no faster than 20 k/sec.
Luck of the draw: sometimes, you'll get a lucky peer (or seed) from the tracker. Because most trackers only return a subset of the peers available, BitTyrant can't make its peering selections with global information. If one client happens to get several high capacity peers that BitTyrant never sees, relative performance may be worse.
Seed dominated performance: if there are a lot of seeds for a file, their altruistic contribution might dominate performance. BitTyrant is designed to improve download performance when tit-for-tat controls download speeds. This is typically the case when there are few seeds for a file, but many leechers.
This is why we conducted an evaluation on not just one torrent, but more than 100, drawn from popular aggregation sites like mininova and piratebay. Aggregate statistics are necessary to have an idea of performance in general, as opposed to special cases that can arise.
Slashdot Mirror
perhaps someday privacy will be a concern, in addition to piracy. Until then, use OneSwarm.
As the article points out, these practices are identical to the lax enforcement practices described last year on slashdot and elsewhere.
The use of indirect evidence as "proof" of downloads is known, the interesting bit here is that in spite of pushback from ISPs and users, industry practices have not changed.
Perhaps this (and the widespread lack of privacy in cloud-based services generally) will drive more users to privacy-preserving data sharing options, such as OneSwarm.
Time to switch to OneSwarm
Actually, the technique used to frame the printer is not IP spoofing. You should read the walled text 8).
Indeed, it's true that BitTyrant will not always improve performance, even when directly compared to other clients on the same torrent at the same time! There are several reasons for this:
- Data availability: Suppose seeders are sending out data at 20 k/sec, and a complete copy does not yet exist in the swarm. Then, even if you start downloading quickly, eventually you will catch up to the data production point and download at no faster than 20 k/sec.
- Luck of the draw: sometimes, you'll get a lucky peer (or seed) from the tracker. Because most trackers only return a subset of the peers available, BitTyrant can't make its peering selections with global information. If one client happens to get several high capacity peers that BitTyrant never sees, relative performance may be worse.
- Seed dominated performance: if there are a lot of seeds for a file, their altruistic contribution might dominate performance. BitTyrant is designed to improve download performance when tit-for-tat controls download speeds. This is typically the case when there are few seeds for a file, but many leechers.
These provide just a glimpse at the many factors that drive BitTorrent download performance. For a more thorough treatment, check out the paper:http://www.cs.washington.edu/homes/piatek/papers/
This is why we conducted an evaluation on not just one torrent, but more than 100, drawn from popular aggregation sites like mininova and piratebay. Aggregate statistics are necessary to have an idea of performance in general, as opposed to special cases that can arise.