Cheating?

[MaliceTheRebel]

I wanted to ask staff if this -> http://www.utorrent.com/forum/viewtopic.php?id=3531 is true and if not, how did they arrive at such claims if you are in fact aren't cheating?

Don't get me wrong, i love bitcomet, i think it's the best client out there, (though i'm getting a serious problem as of recently) i just wanna know if their accusations are founded in truth and why.

[Lucy26]

Well, I'd say some people used BitComet (a very old version) to cheat, instead of BitComet "cheated".

All the criticisms were caused by a bug in an early version several years ago, but that version was immediately replaced by another stable version, and a newer version without that bug was released in a few weeks' time.

The developers of BitComet were working on the software at that time, they were not aware of what was going on in the West (English-speaking community), so they did not come forth and explain it to the public. After that, more and more rumors were spread. Then came the post that you mentioned here.

I will add that all the claims made in this 5 year-old forum post have long been proven to be false. BitComet never falsely reported any data, and never did any kind of cheating. The original "bug" that did exist was a simple exploit used by hackers, and if you want to look at the record, BitComet has had one major exploit in it's 8-year history, whereas uTorrent has had many.

[kluelos]

Notice that this message is from 2005.

At that time, µtorrent was a new client with nothing special to offer, and couldn't get very much traction. BitComet was extremely popular, but its original author was Chinese, as are its current developers. They do not speak English well and do not hang out in English-speaking forums at all .

µtorrent, or at least Switek and Firion, evidently decided that they could get more traction for µtorrent by libelling BitComet. This largely succeeded because it was conducted on forums in the West, and most of BitComet's developers don't speak English or don't speak it very well. There is also a Chinese tradition of not elevating falsehood by answering it. Unfortunately, that runs head-on into the Western tradition that silence implies consent. The lack of an English-speaking presence and spokesperson allowed them to get much further than they should have.

Most of the charges are false, and Switek knew they were false at the time he said them. There is just enough half-truth sprinkled in to keep them alive, so this campaign worked and µtorrent gained popularity at BitComet's expense -- enough so that Ludde, it's developer, could sell it for a hefty chunk of change.

It says a lot of things about our community, most of them not good, that this campaign of lies worked so well.

One early version of BitComet had a bug that allowed hackers to bypass a torrent's "private" flag. Some hackers discovered this and did so. As soon as it was reported, that version of BitComet was withdrawn and replaced with a fixed version. Switek spun this as BitComet setting out to bypass the flag in order to cheat. He was lying about that and knew he was lying, for you had to hack the torrent file in order to do this.

A client reporting, or misreporting, or not reporting at all, statistics to the tracker does not and can not affect the amount of bandwidth that client gets from any seeder. Switek is either a fool who has no idea how bittorrent works and therefore no business talking about it, or is again spreading falsehood knowingly.

Seeders make their own decisions about whom to seed to, and this is an entirely internal software decision by the seeding client. The tracker does NOT report YOUR statistics to any other client, nor any other peer's to you. The well-defined bittorrent protocol has no mechanism for doing this. A seeder thus cannot learn what your client is or isn't saying to the tracker, and nothing your client can do affects what the seeder decides to do. But in the first place, the vast majority of your download comes not from seeders, but from leechers like yourself sharing pieces among themselves. THAT would be the place to cheat, among the leechers and not the seeders, if it were possible to do so.

Switek is either completely ignorant and has no business even commenting on the subject, or he knows all this and is intentionally lying about it, trusting that nobody else knows enough about bittorrent to see this ridiculous fabrication for what it is.

Unfortunately for him, the bittorrent spec was released to the public domain and anybody who cares to, can learn all about it.

There is no "fair" amount of choking or unchoking. Switek's use of this word just invents justification to invoke his own prejudice. BitComet is simply more efficient about detecting connections to peers who aren't very productive than other clients. BitComet already behaved this way and had for years. When so-called "super-seeding" came along, much later, its author (John Hoffman) didn't bother to take that existing behavior into account and didn't bother to persist his own data.

Superseeding makes a client look like a regular peer, and an undesirable peer at that. (It pretends not to have any wanted pieces until it's satisfied with the distribution.) When BitComet disconnected from that undesirable peer in search of somebody better, Hoffman just dropped the data about that connection. When BitComet re-connected to him to see if he'd improved any, Hoffman treated this connection as brand new (because he had dropped the history instead of preserving it). He blames BitComet for that instead of kicking himself for his own error and fixing his own software. He owes a huge, grovelling apology to the entire community for this.

Public tests of BitComet have demonstrated that it shows no preference for connecting to any client. Most recently a set of tests was performed by Rob Topolski, a network engineer, working independently and out of curiosity. Instead of some shadowy, unidentified "college lan" and some unidentified researcher that Switek alludes to, Topolski explained exactly what he did under what circumstances, and invited anyone to duplicate it. Topolski found that these charges were entirely baseless. His results are widely published on the net. Published contradiction doesn't seem to bother Switek or Firion.

Switek suggests that the way to combat BitComet cheating is for the sophisticated and advanced µtorrent client to do "tit-for-tat" trading with BitComet. This is exactly what every client already does - it searches for desirable connections to trade with and does not trade with undesirable connections. Switek's solution is to do what every client had already been doing from the beginnings of bittorrent.

Firion has invented the charge that BitComet misreports its statistics to the tracker. As a result most tracker software ships with BitComet as a banned client by default. We try to get tracker admins to test BitComet for themselves, and when they do, they find there is no truth to the rumor. The µtorrent client itself has had multiple versions which DO so misreport, and which are banned on many private trackers because they have been tested and shown to do this.

You should understand from this that it's not just the lack of merit in these charges which is so problematical. It's the clear evidence of actual malicious wrongdoing -- of people knowingly lying through their teeth so that they could pump and dump their client for profit, that is so personally distressing.

[METALBOSSS]

oh wow i red that and was like " i use that bitcomit.." an was like how am i cheeting if i dont even know what you are talking about, soo how am i cheeting.. i just got here ! lol but ya what that persion said the post was 2005.. by the way did that fix that?:blink:

[cassie]

To repeat (*sigh* one more time *sigh*) what was already said... that bug (the one that existed so many, many, many versions ago) was immediately fixed, soon after its discovery.