0:00:03.471,0:00:04.471
<i>rc3 preroll music</i>

0:00:07.943,0:00:14.598
Herald: Our next speakers are Gus and GeKo[br]from the Tor project. They both came on

0:00:14.598,0:00:19.952
onto the project. A couple have been[br]working with the project for a long time

0:00:19.952,0:00:25.641
now, and a couple of years ago, they both[br]came on as employees. Gus, as the team

0:00:25.641,0:00:30.761
leader, as the community lead of the[br]project and Georg as the network team

0:00:30.761,0:00:36.005
leader, who has been working on improving[br]the health of the network and making sure

0:00:36.005,0:00:41.883
that bad relays are removed. Give them all[br]a great round of applause from home and

0:00:41.883,0:00:49.621
welcome to the stage, guys. Take it away.

0:00:49.621,0:00:54.480
Georg: Hello, everyone, hello. This is[br]Georg from the Tor project, and I have got

0:00:54.480,0:01:02.400
with me today to talk about the State of[br]the Onion, a yearly thing, and we are

0:01:02.400,0:01:08.480
really happy to be here at the CCC and[br]think about providing an update, what we

0:01:08.480,0:01:13.280
did, what we are excited about next year[br]and what is basically in the pipeline.

0:01:14.000,0:01:20.160
Before we start, assuming we have some[br]folks watching this talk, wondering what

0:01:20.160,0:01:26.160
this Tor thing is? We thought about[br]picking them up, getting them up to speed

0:01:26.160,0:01:31.520
and talking about what we are actually[br]talking about here. So, Tor is concerned

0:01:31.520,0:01:37.280
with the online anonymity and censorship[br]circumvention. It's referred to as free

0:01:37.280,0:01:42.560
software, and we actually have an open[br]network of relay operators and relays and

0:01:42.560,0:01:49.120
operated by volunteers. But that's not the[br]only meaning of Tor. You find you are as

0:01:49.120,0:01:55.280
well, you know, in a community of[br]researchers, developers, users, and you

0:01:55.280,0:02:02.720
mentioned relay operators. As a project.[br]We are a US 501c3 nonprofit organization.

0:02:03.840,0:02:11.760
So, that's the different notions of Tor[br]you might encounter. So, what is actually

0:02:11.760,0:02:17.120
the Tor design? How does it help with the[br]anonymity goal or censorship circumvention

0:02:17.120,0:02:23.360
goal? So, I assume you have two parties[br]who want to communicate over the internet,

0:02:24.720,0:02:32.400
and they want particular. Alice wants to[br]hide the location of their IP address, so

0:02:32.400,0:02:36.720
they can connect directly to Bob because[br]that would be obvious where they are

0:02:36.720,0:02:44.960
coming from. So, they try to get their[br]traffic through multiple relays. So, no

0:02:44.960,0:02:51.680
single relay can actually betray Alice[br]here and find out now what Alice is up to,

0:02:51.680,0:02:59.708
or actually, where she is coming from. So,[br]what Alice is doing, or actually Alice's

0:02:59.708,0:03:07.386
Tor-client on her machine is picking a[br]path through the network where through

0:03:07.386,0:03:14.121
relays mentioned here with R1, R2 and R3[br]before she's finally reaching Bob. So,

0:03:14.121,0:03:23.138
this looks like some something like this[br]here, and at the end, Alice is asking the

0:03:23.138,0:03:32.050
exit relay or relay three on this slide to[br]connect to Bob, and then they can talk to

0:03:32.050,0:03:39.448
each other. That's the basic underlying[br]concept of Tor. Then there's the problem

0:03:39.448,0:03:47.168
that we sometimes see censorship in the[br]wild, which means that adversaries trying

0:03:47.168,0:03:53.653
to prevent Alice from actually reaching[br]the Tor-Network and so that she can

0:03:53.653,0:03:59.348
benefit from the privacy properties that[br]the Network is providing. And in this

0:03:59.348,0:04:07.167
case, the direct connection to the cloud[br]above there with the public relays as

0:04:07.167,0:04:13.772
presented. And what Alice needs to do is[br]to connect to so-called bridges, which are

0:04:13.772,0:04:19.307
nonpublic relays in this case, which[br]bridge work as a first hop. And then she

0:04:19.307,0:04:24.521
is picking the usual remaining two hops[br]before connecting to Bob. So, this is a

0:04:24.521,0:04:32.400
rough idea of how Tor is trying to prevent[br]censorship. Or to bypass censorship to be

0:04:32.400,0:04:40.280
more correctly and which will play a role[br]in the coming slides because we talk a

0:04:40.280,0:04:47.558
bunch about censorship, work we do and[br]have done and want to do. So, that's

0:04:47.558,0:04:53.840
basically Tor in a nutshell. That's there[br]are many more things to Tor, but that's

0:04:53.840,0:04:59.737
hopefully enough to understand what the[br]following updates are about. So, if you

0:04:59.737,0:05:06.800
recall the previous slides, that was[br]basically trying to provide privacy at the

0:05:06.800,0:05:12.996
network layer for users hiding the IP[br]addresses. But as we know, the web, in

0:05:12.996,0:05:17.945
particular browsers, are large beasts, and[br]that's by far not enough anymore to

0:05:17.945,0:05:22.940
guarantee any meaningful privacy on the[br]internet because of all of the tracking

0:05:22.940,0:05:28.992
mechanisms and arrays of fingerprint[br]users. So, a couple of years ago, we

0:05:28.992,0:05:36.532
essentially started to provide a tool[br]called Tor Browser, which is essentially a

0:05:36.532,0:05:43.508
fork of Firefox and has dozens of patches[br]on top of that. So, we can actually

0:05:43.508,0:05:50.154
provide the privacy guarantees we think[br]are important. And this tool got some, you

0:05:50.154,0:05:56.051
know, some meaningful updates over the[br]year. And one of these is that we

0:05:56.051,0:06:02.776
overhauled the Tor connection experience.[br]Some of you who are already familiar with

0:06:02.776,0:06:09.902
Tor browser, know about this weird modal[br]dialog popping up once. This is (virtual)

0:06:09.902,0:06:17.292
browser, which was, up until the Tor[br]browser 10.5, the default way of

0:06:17.292,0:06:23.810
connecting to the tunnel broker program,[br]the Tor browser. And this is gone because

0:06:23.810,0:06:28.052
that's a really weird experience if you[br]have any other browser, what is happening

0:06:28.052,0:06:32.942
once you started? You get a browser window[br]and then start searching or typing or

0:06:32.942,0:06:38.496
whatever. You never get any modal dialog,[br]which is a UX experience, which is not

0:06:38.496,0:06:44.757
really the best. So we fixed that. There's[br]no modal dialog during startup anymore,

0:06:44.757,0:06:50.347
and there are easy ways to an easy way to[br]connect automatically now. So, you don't

0:06:50.347,0:06:55.487
even see this particular sort of screen[br]anymore, or was giving you much smoother

0:06:55.487,0:07:03.805
experience for your Tor browser usage,[br]which is pretty exciting. Then we finally

0:07:03.805,0:07:10.360
deployed Snowflake, which is a means for[br]helping censored users on the internet,

0:07:10.360,0:07:17.857
which is, you know, kind of next, next,[br]next-level step in the arms race against

0:07:17.857,0:07:25.360
censors. And this has been in the works[br]for a couple of years and has been testing

0:07:25.360,0:07:31.070
for months in our alpha release series and[br]finally made it earlier this year and

0:07:31.070,0:07:39.520
stable. And you can see in this on this[br]graph how the usage grew over time,

0:07:39.520,0:07:47.501
starting with the initial launch and the[br]stable series at the beginning of July

0:07:47.501,0:07:56.200
this year. You see, there's a continually[br]growing numbers of snowflake users you see

0:07:56.200,0:08:03.235
at the right side, the despite up and[br]down, and we'll talk about this a bit

0:08:03.235,0:08:08.010
later. But it's a growth, and we can see[br]this, and we can hear the feedback for

0:08:08.010,0:08:16.924
users. So, what you can help is. Running[br]snowflakes, how this was going to work is

0:08:16.924,0:08:22.129
a thing Gus will explain later on. But[br]that's already a thing you can try to

0:08:22.129,0:08:27.896
remember and getting out of this talk, so[br]you can help censored users. Um, yeah,

0:08:27.896,0:08:35.002
that's two of the high notes for this year[br]for the next year and upcoming years, we

0:08:35.002,0:08:40.575
plan to make it even easier to help[br]censored users around the world, for

0:08:40.575,0:08:46.240
instance, by faster updating the D4[br]bridges. we ship with the Tor browser.

0:08:46.240,0:08:52.705
Usually, what's happening right now is[br]that once we want to bundle new bridges to

0:08:52.705,0:08:57.952
Tor browser, we have to have a new[br]release, which is pretty cumbersome and

0:08:57.952,0:09:03.512
slow, and we want to make this faster that[br]you can keep your Tor browser but get

0:09:03.512,0:09:10.192
updated bridges if there are any available[br]which we can ship. And then we continue

0:09:10.192,0:09:17.087
working on the general idea of just[br]helping users bypassing the censorship,

0:09:17.087,0:09:21.836
though they should have a button like "I[br]am censored" and then Tor browser should

0:09:21.836,0:09:27.545
figure out everything it needs to provide[br]working bridges for the user and the

0:09:27.545,0:09:34.644
particular region where they are. That's[br]the kind of the golden standard we want to

0:09:34.644,0:09:41.789
get to. So, this will be pretty exciting[br]work then for another project, actually a

0:09:41.789,0:09:47.588
multi-year project, which we recently[br]started, I want you to give an update. The

0:09:47.588,0:09:53.780
Tor browser thing is pretty cool in the[br]sense that you have an app, and then you

0:09:53.780,0:10:02.330
have per app settings kind per app means[br]of providing privacy properties, but

0:10:02.330,0:10:07.602
particularly on Mobile, where you have[br]kind of dozens or hundreds of apps. It's

0:10:07.602,0:10:14.653
pretty cumbersome if it's usable or[br]possible at all to configure. Every app to

0:10:14.653,0:10:20.691
every app to use Tor as a proxy, so what[br]we want, or we actually want to what you

0:10:20.691,0:10:27.752
just want on mobile at least, is a way to[br]him to route all safe traffic and specific

0:10:27.752,0:10:35.282
safe applications through Tor. You don't[br]want to configure this per app, though.

0:10:35.282,0:10:41.680
That's that's not the way to go. That's a[br]pretty "VPN" like functionality to do. I

0:10:41.680,0:10:48.720
put "VPN" in quotes here because that's[br]kind of a working, you know, concept we

0:10:48.720,0:10:56.537
would probably want to come up with the[br]better term at the final product, because

0:10:56.537,0:11:03.703
VPN is kind of tainted and people have[br]particular understandings of what this

0:11:03.703,0:11:11.103
means. VPN is, and you have kind of a new[br]tool here which was trying to fill the

0:11:11.103,0:11:17.716
niche and provide better guarantees than[br]regular VPNs do. So, we want probably come

0:11:17.716,0:11:22.900
up with a different term. But that's[br]pretty close from the functionality point

0:11:22.900,0:11:29.840
of view. What we want to do and the bonus[br]points here as well are that, We can

0:11:29.840,0:11:36.025
easily expand our censorship circumvention[br]means to the whole device and don't have

0:11:36.025,0:11:41.172
to deal with that on a per app basis,[br]either. The work is done with our friends

0:11:41.172,0:11:46.320
from the Guardian project and the LEAP[br]Encryption Access Project, which is

0:11:46.320,0:11:50.632
exciting, and we plan to have this[br]available on Android first, likely

0:11:50.632,0:11:58.136
starting in 2023. Maybe already at the end[br]of next year, we'll see. As I said, it's a

0:11:58.136,0:12:08.914
multi-year project spanning different[br]teams at Tor. It's using Arti the new rust

0:12:08.914,0:12:14.280
based (talk line) we are currently[br]writing. So, that's a pretty exciting

0:12:14.280,0:12:20.761
project, and we hope you make serious[br]progress over next year. So let me leave

0:12:20.761,0:12:26.232
the application part right now and talk a[br]bit about what we could call network

0:12:26.232,0:12:30.954
health. The one of the points which[br]frequently comes up, which is important,

0:12:30.954,0:12:39.016
is our work in the bad relay area. All the[br]dealing with malicious relays remains hard

0:12:39.016,0:12:43.848
with our limited resources. We removed,[br]for instance, several large groups of

0:12:43.848,0:12:49.200
actually relays in early 2021 and used[br]this actually as kind of a wake-up call to

0:12:49.200,0:12:54.880
seriously invest in this area, which means[br]writing new scanners for detecting

0:12:54.880,0:13:02.641
malicious behavior and do a better[br]monitoring for malicious behavior at the

0:13:02.641,0:13:07.600
network. And I think over the year. I'm[br]confident to say that we actually are

0:13:07.600,0:13:12.400
going to have a safer Tor network and[br]compared with previous years, I think it's

0:13:12.400,0:13:18.251
fair to say as well that we right now have[br]a safer Tor network as well compared to

0:13:18.251,0:13:25.070
what we had in the previous year. So, that[br]is exciting progress. Worth mentioning

0:13:25.070,0:13:32.477
here, but that's not enough, right? So,[br]what we actually want to do to provide an

0:13:32.477,0:13:40.349
even safer experience and tackling the the[br]the problem of malicious relays more at

0:13:40.349,0:13:47.560
the core, is leveraging trust in our relay[br]community, helping with those problems.

0:13:47.560,0:13:53.368
And the key points to take away here is[br]that is. It mixed approach in the sense

0:13:53.368,0:13:59.360
that we have technical tools helping, that[br]really work. But as well this is a social

0:13:59.360,0:14:06.041
approach, which is important here because[br]we can't solve the problem of malicious

0:14:06.041,0:14:15.876
relays on the technical means alone. And[br]this is the thing we take into account

0:14:15.876,0:14:21.920
right now already started successfully, I[br]think with experiments, for instance, we

0:14:21.920,0:14:28.080
removed like three weeks ago, two large[br]groups of relays which we deemed to be

0:14:28.080,0:14:35.200
malicious, which were perfectly configured[br]from a configuration perspective. Then all

0:14:35.200,0:14:41.520
the my family settings, and they had a[br]contact info information side, which was

0:14:41.520,0:14:46.960
supposed to be non-spoofable. So, they did[br]all the technical parts right, but still,

0:14:47.520,0:14:52.720
once we start to contact them and tried to[br]talk to them, it was pretty clear they

0:14:52.720,0:14:57.200
were very likely malicious, and we removed[br]them quickly from the network, which

0:14:57.200,0:15:02.080
showed us once more that there's a social[br]component here too, which is important.

0:15:02.960,0:15:07.520
And this will be the priority for the[br]network health team, not only for the

0:15:07.520,0:15:11.520
team. I mean, yes, the community team[br]involved as well, and other teams too. But

0:15:11.520,0:15:18.080
it would be important for the Tor project[br]in 2022. And what this means at the end,

0:15:18.080,0:15:23.680
you know, taking trust into account is not[br]set yet. That could be the idea that we

0:15:24.400,0:15:29.040
say, OK, we have here a large group of[br]trusted relays, and they get more traffic

0:15:29.040,0:15:37.440
to see a lot more traffic to see from uses[br]compared to the non-trusted group. This

0:15:37.440,0:15:41.040
has performance implications and many[br]other implications, which we need to

0:15:41.040,0:15:46.800
explore in detail. Starting this year, but[br]more next year, and probably for the

0:15:46.800,0:15:53.600
coming years, which actually brings me to[br]my final point for my part, which is

0:15:53.600,0:15:58.320
talking to you a bit about Tor performance[br]and the work we did this year and what's

0:15:58.320,0:16:04.400
coming up next. So, if you look at these[br]and this graph of those two graphs, you

0:16:04.400,0:16:10.800
see a growing gap between the bandwidth,[br]which is virtualized on the network and

0:16:10.800,0:16:16.080
the actually used bandwidth over the[br]years, starting from, you know, kind of

0:16:16.080,0:16:24.640
2011 and continuing up until today. This[br]is kind of counterintuitive because one of

0:16:24.640,0:16:31.040
the things we usually get, as, kind of[br]most of the most important complaint, is

0:16:31.040,0:16:37.520
that Tor is slow? So, so what's the issue[br]here? If you have so much kind of surplus

0:16:37.520,0:16:41.920
bandwidth, but it's not getting used, but[br]on the other hand, users are complaining

0:16:41.920,0:16:47.680
Tor is slow. So, we have a project which[br]is trying to solve those problems. We

0:16:47.680,0:16:54.320
think that a big part of this equation is[br]coming up in that good congestion control

0:16:54.320,0:16:58.640
for the Tor Network, which was lacking so[br]far. So, that we have an overall better

0:16:58.640,0:17:04.560
bandwidth usage. And this could be[br]implemented this year, which is exciting,

0:17:04.560,0:17:10.480
and will be deployed next year. And we[br]hopefully see not this growing gap

0:17:10.480,0:17:18.000
anymore, but a shrinking gap.[br]Additionally, one thing we sorely missed

0:17:18.000,0:17:22.880
was feedback for relay operators, whether[br]their relays are doing well, whether they

0:17:22.880,0:17:29.680
are overloaded and whether they can[br]improve settings and make the proper

0:17:29.680,0:17:36.160
modifications. So, we implemented a series[br]of kind of warnings or triggers which

0:17:36.880,0:17:41.920
relay operators can monitor and we from[br]the Tor Project side can monitor as well.

0:17:42.480,0:17:50.480
And then we can ping relay operators and[br]helping them figure out their stuff and

0:17:50.480,0:17:56.800
getting those issues fixed. Resolving the[br]overload they see on their relays and

0:17:58.480,0:18:07.760
planned for 2022 as well is that we start[br]to do better load balancing by figuring

0:18:07.760,0:18:14.800
out which relays are seriously overloaded[br]and moving traffic from them back to less

0:18:14.800,0:18:19.440
overloaded relays, giving an overall[br]better performance and user experience for

0:18:19.440,0:18:26.320
all users. So, I think that's all I had to[br]say from my side. Thanks for listening and

0:18:26.320,0:18:32.470
our Gus will pick this up.[br]Gus: Thank you, Georg. So, hello. This is

0:18:32.470,0:18:40.509
Gus from the Tor project. And today I will[br]talk a little bit about the Community Team

0:18:40.509,0:18:49.680
and our work on the Tor community, so we[br]will cover the new user support forum, our

0:18:49.680,0:18:55.955
new gamification project. The "run a[br]bridge" campaign that we started last

0:18:55.955,0:19:04.308
month, and we are also going to talk about[br]the Tor censorship in Russia. So, for the

0:19:04.308,0:19:12.382
third forum, we at the beginning of this[br]year, we start to think about having a

0:19:12.382,0:19:19.910
place where people can ask questions. That[br]is not the mailing list. So, in 2021, what

0:19:19.910,0:19:26.605
looks like a support forum? You know how[br]where users can do questions and receive

0:19:26.605,0:19:32.287
help. So, email and use of the[br]communication are nice, are cool and

0:19:32.287,0:19:37.076
important because people in certain[br]regions, they can access this resource.

0:19:37.076,0:19:43.390
They can send an email from Iran, from[br]China, from Russia now, and they can

0:19:43.390,0:19:49.401
access our documentation. But you are[br]thinking about, are there other ways to

0:19:49.401,0:19:54.240
reach out to this community to find[br]places, to find a way, for them to

0:19:54.240,0:20:00.366
communicate and ask questions? So, part of[br]GS plan is to,..., The first part of this

0:20:00.366,0:20:06.315
plan is to have a Tor forum, so people can[br]access this information and ask questions

0:20:06.315,0:20:12.353
on your support forum. That's friendly,[br]and you can store an app on your phone and

0:20:12.353,0:20:18.181
contact and talk with others. And later,[br]we'll talk about the second part of this

0:20:18.181,0:20:24.973
plan. So, we launched the Tor Forum[br]jazzier in October, and it has been very

0:20:24.973,0:20:32.348
nice, and I invite everyone to join our[br]forum. The other project that we are doing

0:20:32.348,0:20:37.309
in the community team is the gamification[br]project for relay operators. So, the idea

0:20:37.309,0:20:42.556
is to understand what, what are the[br]motivations, how we can incentivize better

0:20:42.556,0:20:48.217
the Tor network, how we can grow, the Tor[br]network, basically, or why people are

0:20:48.217,0:20:54.011
stopping children relays. So, we are doing[br]this as part of our internship, and Nico

0:20:54.011,0:20:59.440
is our intern, and she is doing this work,[br]and we have a survey online, so people can

0:20:59.440,0:21:06.707
ask some questions and give feedback about[br]their experience, running relays. And last

0:21:06.707,0:21:13.816
month in November, we launched our[br]campaign to get more bridges and in as far

0:21:13.816,0:21:19.320
as ... Well, Bridges are very important[br]for users, living in censored countries.

0:21:19.320,0:21:25.003
This is how they are going to connect to[br]the Tor network. So, our plan was to have

0:21:25.003,0:21:33.092
200 new obfs4 bridges. obfs4 is a[br]pluggable transport that can obfuscate

0:21:33.092,0:21:43.389
your Tor connection. And we, ... so the[br]plan was 200 new bridges and the campaign

0:21:43.389,0:21:52.539
staffs at now are at 947 new running[br]Bridges. 847 new obfs4 bridges, and the

0:21:52.539,0:22:02.328
network size about from 1200 to 2000 new[br]bridges overall. So, the campaign was a

0:22:02.328,0:22:11.298
real success and we ... and you can see on[br]the graph here on the screen how the

0:22:11.298,0:22:19.851
campaign changed the course of the network[br]size here. And so, this campaign started

0:22:19.851,0:22:27.843
in November and December, a situation just[br]happened. So, at the beginning of

0:22:27.843,0:22:34.681
December, we received a lot of users[br]asking for support in Russia and what it

0:22:34.681,0:22:40.488
was not? Well, we usually have some users[br]asking for help, but this time was

0:22:40.488,0:22:46.921
different. We received, like a lot of user[br]support requests, basically emails asking

0:22:46.921,0:22:53.052
for Tor bridges, and that was very strange[br]because we didn't know anything happening.

0:22:53.052,0:22:58.624
So, we start to investigate with OONI[br]which is the "Open Observatory of Network

0:22:58.624,0:23:04.760
Interference" to understand what was[br]happening. So, we start to see some

0:23:04.760,0:23:12.230
anomalies on the Tor net in Russia,[br]basically blocking not just our website,

0:23:12.230,0:23:19.634
but also the Tor network and not only the[br]Tor network, but also some Tor bridges.

0:23:19.634,0:23:25.969
And that was like, ... we started to look[br]into that to understand what was

0:23:25.969,0:23:33.195
happening. So, we start to collect[br]information, and we put together (...)

0:23:33.195,0:23:39.496
Ticket and a few days later, we received[br]an email from Russian authorities saying

0:23:39.496,0:23:45.379
that they were going to block the[br]Torproject domain, and basically, failed

0:23:45.379,0:23:51.660
to give us a reason, and we didn't[br]understand what was happening, so we, ...

0:23:51.660,0:24:00.537
I'm going to skip the lawyer part and the[br]reason that they are blocking the Tor

0:24:00.537,0:24:05.600
project website and I will focus on what[br]they are actually doing and how that is

0:24:05.600,0:24:11.600
impacting the Tor network and the Tor[br]community. So, Russia is the second-

0:24:11.810,0:24:17.840
largest country of Tor users, after users[br]in the United States, Russia, Germany,

0:24:17.840,0:24:24.160
Netherlands and other countries that are[br]the top 10 top 20 countries that are using

0:24:24.160,0:24:32.680
Tor. In the end, as we start to look at[br]the metrics and see that the numbers of

0:24:32.680,0:24:39.760
our users were decreasing in December. And[br]we also saw that the bridge users also

0:24:39.760,0:24:44.880
increasing. So, you can see clearly the[br]impact of the censorship on just a graph

0:24:44.880,0:24:50.640
here and just a graph is available on the[br]metrics portal too. So, the summary here

0:24:50.640,0:24:55.520
is, well, On December 1st, the Russian[br]authorities they blocked Tor Directory

0:24:55.520,0:25:00.000
Authorities. So if you have Tor followed[br]on your computer, you cannot bootstrap

0:25:00.000,0:25:06.000
Tor. They block Tor Browser Bridges. So if[br]you have Tor browser installed, you cannot

0:25:06.000,0:25:10.992
use these bridges. They also block a[br]domain fronting with Azure. So if you try

0:25:10.992,0:25:15.800
to bypass censorship, that was not going[br]to work. They also blocked Snowflake,

0:25:15.800,0:25:20.990
which we will talk about a little bit[br]later. And they also blocked a bunch of

0:25:20.990,0:25:27.740
Tor bridges in different internet[br]providers. So, it depends on where you are

0:25:27.740,0:25:33.886
in Russia, you can use Tor. But in other[br]places, that was going to be more

0:25:33.886,0:25:39.098
complicated. And the only way to bypass[br]the censorship at the time on December 1st

0:25:39.098,0:25:44.430
was to use a bridge from[br]https://bridges.torproject.org or from our

0:25:44.430,0:25:50.510
email. And so, we start to fight the[br]censorship, we launched our Telegram bot

0:25:50.510,0:25:55.726
that you can get a bridge and that the[br]bridges is not blocked in Russia. And we

0:25:55.726,0:26:00.043
tasked these bridges on all of these[br]points on Russia to see if they are

0:26:00.043,0:26:05.120
blocked, if they are blocked we ask for[br]relay operator to hold that IP address.

0:26:05.120,0:26:10.523
So, Tor Bridges are working, and we are[br]checking if they are checking in,

0:26:10.523,0:26:17.556
recording if they are working. That are[br]community also fought back and that our

0:26:17.556,0:26:23.860
committee spin up like more than 400 new[br]Tor bridges in just a few days. I mean, we

0:26:23.860,0:26:32.014
have amazing volunteers translating Tor[br]user support guides in Russian, and doing

0:26:32.014,0:26:38.668
after the first block on December 1st. The[br]anti-censorship thing also provide a fix

0:26:38.668,0:26:45.296
for snowflake, and just fix what's[br]available on Tor browser, the last

0:26:45.296,0:26:51.522
release. So, you can see onto the graph[br]that Snowflake was around like less than

0:26:51.522,0:26:59.045
2000 users, but after December, you can[br]see it take a while, but then such

0:26:59.045,0:27:04.859
increase the number of snowflake users,[br]basically because of Russia. And you can

0:27:04.859,0:27:12.728
see just a graph here. There's a decrease[br]here, is because the server crashed after

0:27:12.728,0:27:19.057
too many users. So, we fixed the server,[br]and we start to get more users. So, if you

0:27:19.057,0:27:24.440
want to help people inside this country,[br]you can run a Tor bridge, or you can run a

0:27:24.440,0:27:31.425
snowflake proxy and that that will be very[br]helpful for Tor users in Russia. And a new

0:27:31.425,0:27:38.280
update, during Christmas, we also had a[br]new round of censorship in Russia. More

0:27:38.280,0:27:45.514
bridges were blocked between December 23[br]and 24. We are going to reach out to relay

0:27:45.514,0:27:51.360
operators, and we are going to contact[br]them and say, OK, you need to rotate your

0:27:51.360,0:27:56.561
IP address if you want to get back in the[br]game and fight censorship. And we are

0:27:56.561,0:28:02.906
going to do that and just (check) if[br]snowflake is working fine, and we have

0:28:02.906,0:28:09.548
been working with doing the other support[br]with Russian users. And we already

0:28:09.548,0:28:16.130
answered more than 1300 Help requests[br]since December 1st. Just for comparison,

0:28:16.130,0:28:24.482
we resolved 1400 support tickets between[br]January and November. So, in one month, we

0:28:24.482,0:28:32.114
already have more user support request[br]from Russia than, you know, in 12 months,

0:28:32.114,0:28:39.840
basically. So, uh, so I will do a call[br]here for the international community to

0:28:39.840,0:28:45.659
spin up a Tor bridge or run a snowflake[br]proxy. If you can't, if you cannot run a

0:28:45.659,0:28:51.687
bridge, you can donate to relay[br]associations. If you cannot donate, you

0:28:51.687,0:28:58.143
can help and teach our users about Tor[br]bridges. Or you can help localize Tor in

0:28:58.143,0:29:03.588
Russian. Or you can do. We can apply[br]pressure like if you are part of a digital

0:29:03.588,0:29:08.732
rights organization or your organization[br]and help us to make pressure on the

0:29:08.732,0:29:15.532
Russian government. And stand up and start[br](a directory) like Edward Snowden did and

0:29:15.532,0:29:22.144
publish messages calling the Russian[br]government to stop blocking Tor. How to

0:29:22.144,0:29:29.280
get involved. We are available on our IRC[br]and Matrix channels. You can join us, our

0:29:29.280,0:29:35.383
mailing list. They are public and you can[br]see what we are talking, and you can help.

0:29:35.383,0:29:41.484
You can also join the Tor Forum and you[br]can contribute on GitLab. And for next

0:29:41.484,0:29:46.411
year, we are going to improve. We are[br]going to continue to improve our user

0:29:46.411,0:29:52.289
support tools for users living in censored[br]countries or regions. So one of our ideas

0:29:52.289,0:29:59.455
is to provide a Telegram chat channel, so[br]users can communicate and have and get

0:29:59.455,0:30:04.323
user support on Telegram. We are going to[br]continue to develop the Tor relay

0:30:04.323,0:30:10.129
gamification project, and continue to[br]organize our trainings in the global

0:30:10.129,0:30:16.120
south, in Latin America and Africa, and[br]organize relay operators meetups. Today we

0:30:16.120,0:30:24.225
are going to have our relay operator meet-[br]up at 10:00 p.m. German time. And the link

0:30:24.225,0:30:28.188
you can find on the Tor relay mailing[br]list. And also, if you search on Twitter,

0:30:28.188,0:30:37.265
on social media, you can also find that,[br]um. And today we just covered some topics

0:30:37.265,0:30:43.721
from the state of the onion. One month[br]ago, we did a huge presentation like two

0:30:43.721,0:30:50.435
and a half hours about anti-censorship[br]from the rising UX SysAdmin team and many

0:30:50.435,0:30:57.720
other updates about Arti, about virtual or[br]non deprecation and many other topics. And

0:30:57.720,0:31:04.400
you can watch that on YouTube. So, I think[br]that's it from my side, and we are open

0:31:04.400,0:31:07.997
for more questions.

0:31:07.997,0:31:12.105
Herald: Thank you so much, guys. Like[br]obviously, Tor is a really important

0:31:12.105,0:31:16.995
project and that's honestly great to see[br]how dedicated you are to basically helping

0:31:16.995,0:31:21.520
everyone. I was actually. Now we're going[br]to go on to the question, and I was

0:31:21.520,0:31:26.720
actually wondering something myself before[br]we head over to taking the ones coming in

0:31:26.720,0:31:32.066
from the internet. Basically, I as far as[br]I understand like when you working with

0:31:32.066,0:31:36.240
bridges and making sure to like, avoid[br]this censorship and everything like as far

0:31:36.240,0:31:41.004
as I understand, an important tool in this[br]process are the meek-bridges where you use

0:31:41.004,0:31:46.248
huge cloud providers to basically mask[br]traffic to Tor. It's like regular HTTPS

0:31:46.248,0:31:51.709
website traffic. Does that not work in the[br]case of Russia or like what does the

0:31:51.709,0:31:57.423
attack threat situation look like at the[br]moment? And that's the landscape.

0:31:57.423,0:32:04.669
Gus: I can answer in two parts. The first[br]part is that some cloud providers, they

0:32:04.669,0:32:12.469
don't like domain fronting. And so, Amazon[br]and others, they change their policy, and

0:32:12.469,0:32:21.832
they start to block, well, not just block,[br]but to remove projects that were using

0:32:21.832,0:32:29.926
domain fronting. So, the only cloud[br]provider that allows Tor or allow Tor to

0:32:29.926,0:32:38.920
do that was Azure, and we had to limit the[br]bandwidth on that. So if you use meek-

0:32:38.920,0:32:45.666
Azure on Tor browser, it's going to be[br]very slow. And one thing that we saw, just

0:32:45.666,0:32:51.860
as the first part, like the providers,[br]they don't like that they were enforcing

0:32:51.860,0:32:59.155
us to stalk, or we will remove just[br]support. The other thing is that the bill,

0:32:59.155,0:33:06.571
like the cost of running a meek-Azure[br]bridge or a meek-Amazon bridge, but it

0:33:06.571,0:33:15.440
that was too high and too costly. So,[br]snowflake is the next step here because it

0:33:15.440,0:33:25.647
uses domain fronting to connect you to a[br]Tor proxy. It's not like proxy, and the

0:33:25.647,0:33:35.440
cost will be like very cheap. So, you can[br]get the benefit of domain fronting, and

0:33:35.440,0:33:42.886
you can use a lot of proxies to connect[br]Tor users. And that will not cost a lot of

0:33:42.886,0:33:48.826
money for the Tor project or for Tor[br]users. So, that is the way to go here is

0:33:48.826,0:33:55.560
not to look back, but look forward.<i>Laugh</i>[br]Herald: It sounds so cool. Like obviously

0:33:55.560,0:34:01.063
it seems that this was very important and[br]actually hearing like some of the problems

0:34:01.063,0:34:05.231
that you guys are facing in your fight, I[br]think that's very interesting for all of

0:34:05.231,0:34:12.326
us. So questions from the audience. The[br]first one is that the apps that you're

0:34:12.326,0:34:17.746
making like the question is, whether they[br]would make you identifiable. So basically,

0:34:17.746,0:34:22.520
if exactly those five apps are always[br]calling home over the same Tor nodes, the

0:34:22.520,0:34:28.519
question is if that if someone could link[br]that back to you?

0:34:28.519,0:34:33.840
Georg: Hmm. Do you want to talk about this[br]Gus? Or should I?

0:34:33.840,0:34:38.893
Gus: Go ahead.[br]Georg: Yeah, I think this should not be

0:34:38.893,0:34:48.775
the case. I mean, depending on what kind[br]of apps you have, how they are configured

0:34:48.775,0:34:55.774
and such and potential, you know, timing[br]signatures and stuff. So, that's one of

0:34:55.774,0:35:01.964
the things we're concerned, for instance,[br]with Tor browser and trying to really make

0:35:01.964,0:35:09.140
sure to break this up in the sense that[br]folks can't learn anything about those

0:35:09.140,0:35:16.946
patterns you have. It's hard, in[br]particular, if adversaries can monitor,

0:35:16.946,0:35:25.346
you know, exit nodes or endpoints over a[br]long period of time. But generally, you

0:35:25.346,0:35:30.060
should be protected from this kind of[br]threat.

0:35:30.060,0:35:39.440
Herald: Right. That makes sense. So, the[br]next question is that if they understand

0:35:39.440,0:35:43.680
correctly, the Tor organization is[br]registered in the United States, could the

0:35:43.680,0:35:47.600
project be in danger of any government[br]pressure to be discontinued, And have you

0:35:47.600,0:35:57.840
guys have a plan to move to more neutral[br]countries like Switzerland or similar?

0:36:02.640,0:36:09.840
Gus: So from my point of view, I don't[br]think we suffer any pressure right now

0:36:10.800,0:36:19.120
from US government. So, I think. Would[br]what would you be interested? Well, one

0:36:19.120,0:36:24.240
thing that is important is one thing is[br]that the Tor project and the other thing

0:36:24.240,0:36:30.880
is the Tor network. The Tor Network is,...[br]we have directed authorities in different

0:36:30.880,0:36:38.404
countries and that just to avoid this kind[br]of government pressure against the Tor

0:36:38.404,0:36:49.400
network. So, I think the question would be[br]more like finding different ways to fund,

0:36:49.400,0:36:59.390
..., make Tor sustainable, not just. Like[br]diversifying our funds, so we don't be so

0:36:59.390,0:37:07.297
connected with a government, are one[br]source provider of resource. I think just

0:37:07.297,0:37:15.600
it's happening right now. Isabella, the[br]executive director, has changing a lot of

0:37:15.600,0:37:24.200
our money income. And if you look back in[br]the Tor history, US government was adding

0:37:24.200,0:37:31.720
a lot of money through to the TOR project[br]in different by different ways, you know,

0:37:31.720,0:37:38.758
like a human rights projects and internet[br]freedom projects. And just was basically

0:37:38.758,0:37:43.854
how Tor is and was funded by U.S.[br]government, but not just U.S. government,

0:37:43.854,0:37:50.298
other governments like Swedish government [br]too. So, I think I would be more concerned

0:37:50.298,0:37:57.760
about the Tor directed authorities being[br]in just one country, and that's not true.

0:37:57.760,0:38:04.388
We are in different countries and they so[br]far I don't I never heard any kind of

0:38:04.388,0:38:11.801
pressure from the U.S. government against[br]the nonprofit, call it the Tor project.

0:38:11.801,0:38:15.554
So, I think that it's basically, my answer[br]here.

0:38:15.554,0:38:21.735
Herald: That's good to hear. And now on to[br]maybe a little bit lighter question, do

0:38:21.735,0:38:29.312
Tor browser users have any chance or hope[br]to see less captchas in the future?

0:38:29.312,0:38:36.720
Georg: Yeah. Yeah. I think we do have some[br]hope, there is, ... I mean, not just only

0:38:36.720,0:38:45.520
hope. But we have work ongoing solving[br]this from different angles. The first one

0:38:45.520,0:38:53.040
is outreach to major providers trying to[br]understand why they are blocking Tor or

0:38:53.040,0:38:58.720
why they provide, captchas and working[br]with them to come up with solutions, which

0:39:00.160,0:39:07.280
are not only deployable by them, but by[br]the wider industry. So, there is a

0:39:07.280,0:39:12.640
knowledge gap here and then trying to,[br]..., based on that, trying to figure out

0:39:12.640,0:39:19.360
how we can solve this problem. And that's[br]not only from , you know, policy angle,

0:39:19.360,0:39:28.240
but we plan to look into technical means[br]as well. For instance. There's the idea of

0:39:28.240,0:39:36.320
providing tokens to Tor users, so[br]they can, which they can spend anonymously

0:39:36.320,0:39:42.960
at websites, for instance, and the[br]websites can look for that and try to

0:39:43.520,0:39:49.840
regulate the traffic, keeping the noisy[br]bots out while providing good service to

0:39:49.840,0:39:58.160
our users providing such a token. That's[br]another thing that won't be solved next

0:39:58.160,0:40:04.080
year. It's a multi-year project, too. We[br]are a small organization, so there has to

0:40:04.080,0:40:11.360
be some kind of prioritization. But that's[br]definitely on our radar and a serious

0:40:11.360,0:40:16.720
problem for us. So, we should fix this.[br]Herald: Sounds like great initiatives and

0:40:16.720,0:40:22.320
also like that going some of the way in[br]order to some extent legitimize the use of

0:40:22.320,0:40:26.320
the Tor browser. Maybe not as much in[br]common society, but also when actually

0:40:26.320,0:40:28.720
visiting different websites.[br]Georg: Yeah, exactly.

0:40:28.720,0:40:36.320
Herald: Nice. Next up is whether you guys[br]are planning to figure out some kind of

0:40:36.320,0:40:40.640
solutions for firewalls, for instance, the[br]corporate ones that are slowing traffic

0:40:40.640,0:40:55.360
down.[br]Georg: I know, Gus, do you. Do you have

0:40:55.360,0:41:01.760
some, you know, queries or complaints from[br]users for this particular issue? I'm not

0:41:01.760,0:41:11.611
sure about that.[br]Gus: Yeah, I and. I just want to be a very

0:41:11.611,0:41:15.842
specific question, I ....[br]Herald: It's also very fair to just say

0:41:15.842,0:41:19.760
that it's not a problem that you've heard[br]a lot of complaints about,

0:41:19.760,0:41:23.682
Georg: Right.[br]Gus: Yeah, sure, that's true. We I didn't

0:41:23.682,0:41:29.567
hear about that. Like the captacha one is[br]a popular one, but I never heard.

0:41:29.567,0:41:38.072
Georg: I think they're a bunch of larger[br]things to fry here, there. It's not really

0:41:38.072,0:41:43.803
in our not even our top 10. So there.[br]Herald: Right? I guess it can also be very

0:41:43.803,0:41:48.002
hard for you guys to like, work with[br]figuring out how to prioritize all the

0:41:48.002,0:41:51.240
different initiatives and wishes that that[br]people have.

0:41:51.240,0:42:01.567
Georg: Yeah, definitely.[br]Herald: Cool. So unfortunately, we don't

0:42:01.567,0:42:07.080
have time for any more questions right[br]now, but there is a break-out room that

0:42:07.080,0:42:13.002
people can come to, and you will answer[br]any further questions. For now, we are

0:42:13.002,0:42:18.525
going to have a break on this channel[br]before the next talk that's going on at

0:42:18.525,0:42:23.616
20:00, which is (in German) "Cookiebanner,[br]das Online-Werbe-Ökosystem und Google,

0:42:23.616,0:42:31.027
Preisträger BigBrotherAwards 2021" For[br]now, thank you very much, guys. Take care

0:42:31.027,0:42:35.686
and maybe we'll see you in the break-out[br]room.

0:42:35.686,0:42:47.975
<i>postroll music</i>

0:42:47.975,0:42:52.243
Subtitles created by c3subtitles.de[br]in the year 2021. Join, and help us!