Ok, Camilo, let me get this straight.
1) Strict Exit Relay Management is for limited the exit node to one country.
2) Exclude Relay Management, is for excluding certain countries, but does it exclude it as a exit node or as a TOR node?
Also, will it still form a connection yet not use it? Or should it not even form a connection?
Here's my problem. I have excluded USA and UK from the dropdown list using ctrl key as you said and added them, but i see that connections are still formed with them as a TOR node and also as exit nodes.
NOt only that, I also noticed that it even routed my request through an exit node through UK, specifically:
achaycock (Online)
Location:
London, GB
IP Address:
78.86.55.121
Platform:
Tor
0.1.2.19 on Windows Server 2003 Service Pack 2 [server] {terminal services, single user} {terminal services}
Bandwidth:
111 KB/s
Uptime:
13 hours 20 mins 22 secs
Last Updated:
2008-08-21 04:03:38 GMT
What am i doing wrong?
On Thu, Aug 21, 2008 at 5:18 AM, Camilo Viecco
<cviecco@xxxxxxxxxxx> wrote:
Hello M
Thanks for giving it a try. I have comments inline
M wrote:
Hello friends I'm new on the list. I hope you bear with my questions and problems.
I just installed Camilo's version of Vidalia, and it seems i have a couple of problems:
1) You can only exclude one country from the "invalidnodes" settings.
You can do multiple selections by pressing the 'Control (ctrl)' button when selecting
the second (or next) country.
2) You have to exclude it every time you start vidalia (it does not save the settings)
Are these bugs in my installation, or is the program like this?
It is a bug in the program. Exit countries should be saved. Thanks for finding it. I will fix it in a few days.
3) Also, how much does this reduce anonymity?
Placing any restrictions on the nodes most likely will reduce your anonymity. In particular limiting
the exit country significanly reduces your anonymity as it is much cheaper for an attacker to place
nodes in that country and thus your probability of selecting a 'bad' exit is higher.
Reducining the number of other nodes could 'possibly' be bad for your anonymity. Part
of Tor's attacker model assumes that there are many attackers that will not cooperate with each other.
There might be more academic studies about these effects, but none come to my mind at the moment.
Will let the list give you the pointers.
(I think 2007 PETS IX attack on Tor would be a place to start
(http://www.freehaven.net/anonbib/cache/murdoch-pet2007.pdf))
Thank you for noticing AND submitting about the bug
Camilo