I am having some weird issues with a client reporting -KT0101- as it's peer ID. Is this ktorrent's peer id? I won't bother posting boring details until I ensure that I'm talking to the right devs.

Edit: I'm a tracker developer. Hope the above makes more sense now.

These are the peer ID's we are using :

2.0.2 : -KT2020-
2.0.1 : -KT2010-
2.0 : -KT2000-
2.0rc1 : -KT20R1-
2.0beta1 : -KT20B1-
1.2 : -KT1200-

Current development version is :
-KT21DV-

We never had a 0.1 release certainly not a 0.1.0.1 release. Maybe somebody else is developing a client with KT as peer ID ?

It's possible that someone is using it as well... but you may want to know there seems to be a client out there that looks like ktorrent. It connects to the tracker as a seed (even when leeching!) and doesn't report anything (not even a complete peer id.) I don't know if it is doing this because of a bug, or it's just a leeching client.

ktorrent looks interesting, I'm compiling it right now.

Looks to me like a cheating client of some kind. Don't take this for granted but I think I've read about a client somewhere that reminds me of this situation. I'm not sure if peer ID was starting with KT, though.

danomac wrote:It's possible that someone is using it as well... but you may want to know there seems to be a client out there that looks like ktorrent. It connects to the tracker as a seed (even when leeching!) and doesn't report anything (not even a complete peer id.) I don't know if it is doing this because of a bug, or it's just a leeching client.

ktorrent looks interesting, I'm compiling it right now.

Is there a User-Agent in the HTTP requests of this peer to the tracker ? We also send ktorrent/<version> as User-Agent string.

There was no user-agent string.