Hi Moritz, DFRI [1] is running two relays, DFRI0 and DFRI3, which should qualify as fast relays as defined by sponsor j. (They were just recently bumped From 10 MB to 12.5 MB. They both have an exit policy allowing the ports explicitly required on https://trac.torproject.org/projects/tor/wiki/org/sponsors/SponsorJ.) We'd love to be able to get some funds for this which would make it possible for us to run more exit traffic. We only have one /24 network to run exit traffic from atm though so this is all we can do for now without breaking the stated diversity rules. (Anybody sitting on an IPv4 /24 PI that we can borrow for a year or so? We'd handle it with care!) We could technically run bridges too but we would have to discuss this internally some more first. Do we really want to take money for running entry _and_ exit relays? Isn't that exactly how you'd attack Tor users if you had the power, by controlling entry and exit? If we've missed some public discussion on this topic, please point us at it. [1] https://dfri.se/index.en/ Thanks, Linus, DFRI
Hi Linus,
DFRI [1] is running two relays, DFRI0 and DFRI3, which should qualify as fast relays as defined by sponsor j.
As far as I know, DFRI as a registered non-profit qualifies for reimbursement directly from the Tor Project. Andrew is in the process of writing up a contract with the German Wau Holland Foundation, through which individuals and other not registered organizations can get reimbursed. Andrew, my monitoring shows DFRI0 and DRFI3 as qualifying "fast exits" since today, Nov 26.
We could technically run bridges too but we would have to discuss this internally some more first. Do we really want to take money for running entry _and_ exit relays? Isn't that exactly how you'd attack Tor users if you had the power, by controlling entry and exit? If we've missed some public discussion on this topic, please point us at it.
I understand your worries and would love to see this discussed more. -- Moritz Bartl https://www.torservers.net/
On Mon, 26 Nov 2012 01:27:39 +0000, Linus Nordberg wrote: ...
We could technically run bridges too but we would have to discuss this internally some more first. Do we really want to take money for running entry _and_ exit relays? Isn't that exactly how you'd attack Tor users if you had the power, by controlling entry and exit?
You'd be doing so by running entries and exits, and not telling the world about it, like by putting in the family list. So, I don't see a point in a single entity not running entries & exits as long as they are declared. Andreas -- "Totally trivial. Famous last words." From: Linus Torvalds <torvalds@*.org> Date: Fri, 22 Jan 2010 07:29:21 -0800
Andreas Krey <a.krey@gmx.de> wrote Mon, 26 Nov 2012 17:16:16 +0100: | > We could technically run bridges too but we would have to discuss this | > internally some more first. Do we really want to take money for running | > entry _and_ exit relays? Isn't that exactly how you'd attack Tor users | > if you had the power, by controlling entry and exit? | | You'd be doing so by running entries and exits, and not telling the world | about it, like by putting in the family list. So, I don't see a point in | a single entity not running entries & exits as long as they are declared. The sponsor wants bridges. You're not supposed to set MyFamily for bridges. Supposedly because they'd show up in the descriptors of the other family members.
On Mon, Nov 26, 2012 at 05:22:23PM +0100, Linus Nordberg wrote:
Andreas Krey <a.krey@gmx.de> wrote Mon, 26 Nov 2012 17:16:16 +0100:
| > We could technically run bridges too but we would have to discuss this | > internally some more first. Do we really want to take money for running | > entry _and_ exit relays? Isn't that exactly how you'd attack Tor users | > if you had the power, by controlling entry and exit? | | You'd be doing so by running entries and exits, and not telling the world | about it, like by putting in the family list. So, I don't see a point in | a single entity not running entries & exits as long as they are declared.
The sponsor wants bridges. You're not supposed to set MyFamily for bridges. Supposedly because they'd show up in the descriptors of the other family members.
I can simplify your dilemma by saying that we're not currently reimbursing people for operating bridges. The sponsor does want bridges, but for now we have enough, and they're not going through them very quickly. So we're saving the funding for a time when they do (and in the mean time, we're working on better bridge address distribution strategies and better blocking-resistance approaches). --Roger
participants (4)
-
Andreas Krey -
Linus Nordberg -
Moritz Bartl -
Roger Dingledine