11:59:40 <hellais> #startmeeting OONI community meeting 2017-03-21
11:59:40 <MeetBot> Meeting started Tue Mar 21 11:59:40 2017 UTC.  The chair is hellais. Information about MeetBot at http://wiki.debian.org/MeetBot.
11:59:40 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
11:59:45 <slacktopus> <eiko> well can take turn for the timezone if wanna be fair
11:59:56 <slacktopus> <agrabeli> Hello everyone!
12:00:02 <slacktopus> <sukarn> Halfway between this time and the time at which the last meeting started?
12:00:07 <slacktopus> <sweemeng> awesome
12:00:08 <slacktopus> <agrabeli> Welcome to the March OONI community meeting! :slightly_smiling_face:
12:00:14 <slacktopus> <agrabeli> Thanks to all of you for joining
12:00:25 <slacktopus> <agrabeli> we will aim to keep this meeting within an hour
12:00:29 <slacktopus> <hellais> Agenda link: https://pad.riseup.net/p/ooni-community-meeting
12:00:58 <anadahz> Hello everyone
12:00:59 <slacktopus> <hellais> If you have more topics to discuss, feel free to add them to the pad :slightly_smiling_face:
12:01:19 <hellais> #topic 1. Checking accessibility of websites from Tor
12:01:29 <slacktopus> <hellais> @boter are you here?
12:02:31 <slacktopus> <hellais> First apologies for not getting back to you via email. I will write here some of the thoughts I have on this topic.
12:02:47 <slacktopus> <hellais> From what I gather the problem definition is the following:
12:03:41 <slacktopus> <hellais> You want to determine which italian public administration websites are blocking Tor users and make a site that shows a list of such sites.
12:04:55 <slacktopus> <hellais> I think that the `http_requests` test is for sure the most well suited for this purpose that will fetch the site in question once over Tor and another time over the network of the user. In your case you will not use Tor as the control, but rather the network of the “user” which is a location that you know to not be filtered or black listed.
12:05:24 <slacktopus> <hellais> This study is in a way similar to: https://www.cl.cam.ac.uk/~sk766/publications/ndss16_tor_differential.pdf
12:05:50 <slacktopus> <hellais> You will probably encounter a bunch of false positives so you will probably have to do some post-processing of the data.
12:07:32 <slacktopus> <hellais> I can see you tackle the post-processing in 2 ways: 1. If the sites you want to monitor are not that many, you can have a set of regular expressions for each page that define what you expect the response to look like and if it deviates from that you flag it as blocked 2. Based on the blockpages you build a set of rules for how they look like (probably on a site-by-site basis) and use that
12:08:18 <slacktopus> <hellais> I don’t think you can use the strategies we are currently using for mining blockpages in the pipeline since we consider a blockpage something that for different URLs is the same. Since the blocking is happening server-side the blockpages will probably all be different for every site
12:09:12 <slacktopus> <hellais> This repository: https://www.bamsoftware.com/git/ooni-tor-blocks.git/ could also be a good starting point for the server-side blockpage fingerprints
12:09:56 <slacktopus> <hellais> That is all I have for the moment on this topic, we can maybe move onto the next item, unless somebody has something to add
12:10:24 <slacktopus> <darkk> I can also add couple of repositories hosting blockpages sources:
12:11:05 <slacktopus> <darkk> https://github.com/NullHypothesis/blockpages and https://github.com/citizenlab/blockpages
12:11:57 <slacktopus> <hellais> yup, though those generally have government or ISP issued blockpages and not server-side blockpages (i.e. “you can’t access this page, because your IP got badlisted”)
12:12:00 <hellais> #topic 2. Brainstorm ways to improve measurements in at-risk countries.
12:12:11 <hellais> oops, I didn't mean to paste that so fast
12:12:53 <hellais> anyways to keep things moving we can move onto the next topic and if there is more to add to topic 1., please feel free to do so nonetheless
12:13:00 <slacktopus> <mahsa> ah, i made it just in time. this was my input
12:13:03 <slacktopus> <mahsa> :slightly_smiling_face:
12:13:27 <slacktopus> <agrabeli> For reference on this topic, here's OONI relevant documentation on potential risks associated to the use of ooniprobe: https://ooni.torproject.org/about/risks/ and here's OONI's Data Policy: https://ooni.torproject.org/about/data-policy/
12:13:38 <slacktopus> <agrabeli> @mahsa great! :slightly_smiling_face:
12:13:56 <slacktopus> <agrabeli> @mahsa perhaps share some words on the topic you'd like us to discuss?
12:14:37 <slacktopus> <mahsa> i just wanted to know how people who use OONI data go about this in countries were there are fewer probes :slightly_smiling_face:
12:16:12 <slacktopus> <agrabeli> While we are interested in collecting network measurements from all of the countries in the world (thus gaining a better understanding of internet censorship globally), we try to ensure that users (especially in high-risk countries) are aware of what ooniprobe tests are designed to do and that they are aware of potential risks associated to them
12:16:33 <slacktopus> <tafiti> We have had some people run tests in at-risk environments by constantly being on the look out for openings - like someone traveling briefly to such a country and can afford to run the tests from a public network
12:16:42 <slacktopus> <sweemeng> @mahsa probe deployment is part of our plan of using ooni data
12:17:16 <slacktopus> <agrabeli> To this end, when possible, we engage in discussions directly with users, explaining the various ooniprobe tests, and the various choices they can make in terms of what to run, what types of data to submit, and how they can opt-out (when necessary). We also engage in discussions with local lawyers, when/if possible.
12:18:12 <slacktopus> <agrabeli> @tafiti yep! There are various strategies we carry out depending on the relevant risk and what users feel comfortable with. This changes on a case-to-case basis.
12:18:34 <slacktopus> <eiko> @agrabeli Good point. I'm thinking to pull some lawyers on board.
12:19:37 <slacktopus> <agrabeli> In many cases, we engage in discussions with individuals and groups for probe deployment and in many of the cases, probes don't get deployed in the end because the individuals involved don't feel comfortable with the potential risks involved - and that's fine. In all cases, our goal is to ensure that potential users are as informed as possible, and they acknowledge the potential risk involved.
12:19:59 <slacktopus> <agrabeli> Over the last years we have received extensive legal consultation in regards to potential risks
12:20:33 <slacktopus> <agrabeli> OONI's documentation kind of outlines the "worst case scenarios": https://ooni.torproject.org/about/risks/ which is more likely to occur in high-risk countries
12:20:50 <slacktopus> <agrabeli> To our knowledge though, no one who has ever run ooniprobe has ever gotten into any trouble
12:21:06 <slacktopus> <agrabeli> That said, as OONI's userbase expands, the potential for risk might also expand
12:21:13 <slacktopus> <agrabeli> Though this also remains quite unclear
12:21:44 <slacktopus> <agrabeli> The risk is not only affiliated to the types of tests run and the legal and political conditions of a country, but they are also associated to the personal threat models of the users
12:22:25 <slacktopus> <eiko> Yes, but I'm little worried about friends in Thailand. The Gov is crazy. :S It's not illegal to deploy probe but you will never know they using other reason to against you over checking the censorship
12:22:28 <slacktopus> <darkk> There is also a question to ask individuals who were uncomfortable with the probe -- does ooniprobe app make them more comfortable or not. Is it the measurement itself that makes people uncomfortable or is it presence of some physical evidence in their "land".
12:23:16 <anadahz> @mahsa another approach is to host probes in entities such as internet freedom organizations that will be more likely to handle legal/safety issues better than individuals.
12:23:44 <slacktopus> <sukarn> @eiko @agrabeli I'm a lawyer at SFLC.in in India. If you'd like, perhaps we could talk about the legality in India.
12:23:58 <slacktopus> <eiko> That's also a question from a Thai friend: does there any risk concerning running it from his/her own personal device (the phone is a post-paid subscription)
12:24:06 <slacktopus> <agrabeli> @eiko yes that's a good point... :( As you rightfully point out, even though running ooniprobe might not be illegal per se (I'm not aware of any country that legally prohibits collecting network measurements), the mere use of ooniprobe could potentially be prosecuted under broader, national security laws if it is viewed as a form of anti-government activity....
12:25:03 <slacktopus> <agrabeli> @eiko so this is where it's important for users to know their options
12:25:13 <slacktopus> <agrabeli> Not all of OONI tests carry the same weight of risk
12:25:20 <slacktopus> <agrabeli> some tests are riskier/safer than others
12:25:32 <slacktopus> <agrabeli> Also, users have options in terms of the data they submit
12:25:54 <slacktopus> <agrabeli> For example, if a user doesn't feel comfortable with her data being published, she could opt-out from sending her measurements to OONI's collectors
12:26:15 <slacktopus> <agrabeli> Furthermore, users now have different options in terms of the platform through which they run ooniprobe
12:26:43 <slacktopus> <agrabeli> these including: Android, iOS, Linux or macOS desktop, or Raspberry Pis
12:26:54 <slacktopus> <agrabeli> there are different trade-offs depending on the platforms
12:27:41 <slacktopus> <agrabeli> Running ooniprobe from the mobile app, for example, is by far easier.... but you don't have the ability to choose which URLs to test, which is something that you could do from OONI's desktop version via the command line (by passing your own test list)
12:28:05 <slacktopus> <agrabeli> (that said, one of the upcoming features for the mobile version is to also enable users to choose which URLs to test)
12:28:29 <slacktopus> <agrabeli> @sukarn We'd love to hear about the legality in India
12:29:33 <slacktopus> <sukarn> @agrabeli I'd prefer to talk about this in a side-discussion. Is that okay?
12:29:47 <slacktopus> <agrabeli> @sukarn sure, sounds great
12:29:52 <slacktopus> <sukarn> Let's let this discussion go on without this interruption
12:30:13 <slacktopus> <agrabeli> would anyone like to add anything on this topic before we move to the next agenda item?
12:30:14 <anadahz> To the best of my (our?) knowledge no individual or entity has ever been prosecuted, threatened or questioned for running ooniprobe measurements or other network measurements.
12:31:06 <slacktopus> <eiko> open a channel for Legal :slightly_smiling_face:
12:31:22 <slacktopus> <agrabeli> @anadahz yep
12:31:43 <slacktopus> <mahsa> thanks guys, helpful!
12:31:47 <slacktopus> <darkk> Yeah. But there were people prosecuted for playing pokemon go in church, so you never know what's more dangerous :slightly_smiling_face:
12:31:59 <slacktopus> <agrabeli> @mahsa in short, we address risks involved on a case-by-base basis, and we strive to ensure that users are as informed of risks and options as possible.
12:32:00 <slacktopus> <sweemeng> let's move on?
12:32:13 <anadahz> sure
12:32:16 <hellais> #topic 3. Discuss how one can run custom tests, decks and inputs in ooniprobe. (anadahz)
12:33:32 <hellais> anadahz: do you want to say something about this item?
12:33:58 <anadahz> I don't want to go that technical but I will describe a short example of a use case that many of us would like to use in ooniprobe; running custom tests and custom decks with inputs that (such URL lists) that are not included by default in ooniprobe.
12:37:28 <anadahz> I would like us to discuss of a standard way to run custom ooniprobe measurements and then based on the comments I will add this as part of ooniprobe's documentation that will describe the canonical way (best practice) for adding and running custom decks, inputs and tests in ooniprobe.
12:38:28 <anadahz> Does anyone here had the need to run a custom experiment in ooniprobe?
12:39:49 <slacktopus> <kiran> We could do it in this thread as well.
12:40:21 <slacktopus> Action: darkk was happy with desktop ooniprobe CLI for that so far
12:41:05 <slacktopus> <sukarn> No, but I have a side-question that is purely from a curiosity point of view: Can the Android terminal be used to make scripts similar to the desktop ooniprobe CLI? Perhaps by running the desktop CLI version with the help of some dependencies?
12:41:16 <slacktopus> <sweemeng> we just propose test if it don't exactly work the way we want. the check works for us most of the time
12:42:17 <slacktopus> <hellais> @sukarn no at the moment there is no way to automate testing via the terminal or otherwise on Android. Do you have an example of a use-case you have in mind?
12:42:51 <slacktopus> <agrabeli> @anadahz which custom tests are you referring to? I guess those could be added under the same processes that OONI tests are integrated. Under probe orchestration users would be able to add custom inputs (URLs), but not sure about custom tests...
12:43:04 <anadahz> My use case is deploying a number of probes and testing for a number of URL entries that are not included in the current test-lists
12:43:47 <slacktopus> <hellais> no, custom tests are out of scope for probe orchestration, but we are very happy to accept pull requests for custom tests to be integrated into ooniprobe or measurement-kit :slightly_smiling_face:
12:43:50 <slacktopus> <darkk> @sukarn I see no obvious blockers to run desktop ooniprobe in debian chroot on android device (well, without root-requiring measurements), but that sounds a bit hardcore :slightly_smiling_face:
12:43:52 <slacktopus> <agrabeli> *correction: under probe orchestration we would be able to push URLs, but users would also have the ability to test specific input through the new features that will be added
12:43:54 <slacktopus> <sukarn> Integration with Tasker by passing a shell-script from Tasker, reading the results, and using those results to figure out what's going on, but in a more targeted way. This is purely out of curiosity, though.
12:44:29 <slacktopus> <sukarn> @darkk That would be one way to go about it :)
12:45:02 <slacktopus> <hellais> @sukarn I didn’t know about Tasker. Maybe there is a way to make the android app “Tasker Friendly”.
12:45:59 <slacktopus> <sarath_ms> @hellais As part of a discussion at IFF last week about detecting blanket shutdowns, we discussed running a test that performs a traceroute.
12:46:07 <slacktopus> <darkk> I also thought of Tasker (or something like that) to run measurement for every new wi-fi network. But I don't know if we should use rely on Tasker for that or embed the logic in the main app.
12:46:24 <slacktopus> <sweemeng> nice for cron job for phone
12:46:32 <anadahz> Regarding custom tests I would like to use the http-requests test that is not included in the default ooniprobe tests; something similar to the first agenda topic
12:46:50 <slacktopus> <sweemeng> or automatically test network when connect to a wifi
12:46:57 <anadahz> *that @boter mentioned
12:46:58 <slacktopus> <sukarn> @hellais Tasker has information on how to develop plugins for Tasker here: http://tasker.dinglisch.net/developers.html and here: http://tasker.dinglisch.net/plugins.html I can't develop plugins, though I can use them if they exist. I don't have those technical skills to develop a plugin.
12:47:16 <slacktopus> <hellais> @sukarn I will look into that
12:47:17 <slacktopus> <agrabeli> @anadahz ah so existing ooniprobe tests that aren't by default included in the mobile apps and in the raspberry pi distributions...good point!
12:47:50 <anadahz> @agrabeli exactly!
12:47:53 <slacktopus> <hellais> if anything it can be useful for inspiration on how to design our UI to support tests that are dependent on one another or trigger based
12:48:07 <slacktopus> <sukarn> @hellais Thank you :slightly_smiling_face:
12:48:41 <slacktopus> <hellais> Do we have more to discuss on this item?
12:49:47 <slacktopus> <boter> sorry, I've just kept up with all the discussion, I was late
12:49:51 <slacktopus> <hellais> I mean currently if you want to do what @anadahz describes you can define your custom deck and pop it inside of the deck directory and it will work. Since the user that has this sort of requirement is probably technical I think the current UI for doing this is sufficient
12:50:31 <hellais> #topic 4. Discuss setting up a large deployment in India to detect region-specific mobile censorship (kiran)
12:50:55 <slacktopus> <hellais> @boter no worries, if you have anything to add feel free to chime in at any time
12:51:52 <slacktopus> <boter> just before moving to topic 4, regarding the possibility of browser plugins, would it be possible to develop a plugin that checks directly if the website you're on is available via Tor?
12:51:53 <slacktopus> <agrabeli> @kiran: perhaps share a few words on your suggested topic?
12:53:15 <slacktopus> <hellais> @boter it would have to be developed, probably it would need some feed of which sites it knows to be blocked by doing some parts of the analysis I described above.
12:53:57 <slacktopus> <kiran> @agrabeli sure
12:56:17 <slacktopus> <kiran> India’s censorship scene is a bit complicated:  1. The Department of Telecom is authorised to block any site, but blocking orders are secret. 2. Courts may also issue blocking orders, almost always to block torrent websites prior to a big movie release. 3. ISPs are required by law to comply with blocking orders, but may not reveal DoT orders. 4. ISP compliance is not tested anywhere. IP-based blocking is rather common. 5. When it’s not IP-ba
12:56:17 <slacktopus> DNS-based blocking, by poisoning DNS records. 6. When it’s not that either, it’s Deep Packet Inspection, but only the larger ISPs can afford this.
12:56:38 <slacktopus> <kiran> Net result, any random website may be accidentally blocked by some regional ISP with no one being the wiser.
12:57:03 <slacktopus> <kiran> And this does not cover malicious blocking by the ISPs themselves. Neither government nor public have the tools to detect such blocking.
12:57:05 <slacktopus> <agrabeli> @kiran this is super interesting info!
12:57:56 <slacktopus> <kiran> So we need a really wide deployment of OONI in India, covering as many ISPs as we can. The only practical way to make that happen is via the Android app.
12:58:01 <slacktopus> <agrabeli> @kiran well I guess one of the perks of running ooniprobe is that (hopefully) it will enable the user to detect most forms of blocking (and maybe including malicious forms as you describe)
12:58:03 <slacktopus> <kiran> I’m really happy to see OONI has one now.
12:58:31 <slacktopus> <agrabeli> @kiran Great! are there any features that you would like to see in the mobile apps that could enable the deployments in India?
12:58:42 <slacktopus> <agrabeli> or that would work best for the Indian context
12:58:47 <slacktopus> <kiran> What we need is a reporting mechanism for adding a new website to the tests.
12:59:22 <slacktopus> <kiran> And automated testing within the app, so I can set it up to run a test whenever it receives a push notification of some new website (subject to user consent).
12:59:50 <slacktopus> <kiran> We also have a shutdowns problem in India, which IFF and SFLC are running campaigns against.
13:00:10 <slacktopus> <agrabeli> Currently new sites are added to the Indian test list via github pull requests: https://github.com/citizenlab/test-lists/blob/master/lists/in.csv
13:00:25 <slacktopus> <kiran> So one region within India could lose mobile data with barely any notice to users, leaving no way to submit test reports except via SMS.
13:00:28 <slacktopus> <agrabeli> A guide on how to contribute to test lists is here: https://ooni.torproject.org/get-involved/contribute-test-lists/
13:00:55 <slacktopus> <agrabeli> But we aim to add a feature to the mobile apps that enables users to test single URLs, regardless of whether or not they are included in the Citizen Lab test list
13:01:08 <slacktopus> <kiran> I suspect the shutdowns problem will get worse before it gets better, with telcos limiting shutdowns to narrow geographical regions, making it that much harder for anyone else to notice there’s been a shutdown
13:01:19 <slacktopus> <agrabeli> We also aim to add support for push notifications
13:01:24 <slacktopus> <kiran> Or by blocking specific services such as WhatsApp to keep local governments happy.
13:01:42 <slacktopus> <agrabeli> On Android, users can have tests run automatically by enabling the "Run tests automatically" option in the Settings
13:02:02 <slacktopus> <sukarn> We did discuss an easier way to add websites to the list last time, right?
13:02:20 <slacktopus> <agrabeli> We aim to integrate OONI's WhatsApp and Facebook Messenger tests in the mobile apps.... but not sure when that will be done
13:02:28 <slacktopus> <sukarn> This would be very useful for both @kiran and I
13:02:31 <slacktopus> <kiran> I’m happy to help get a larger deployment base for OONI Probe within India.
13:02:41 <slacktopus> <agrabeli> In the mean while, those tests can get run through OONI's web interface for macOS or Linux or via Raspberry Pis
13:03:07 <slacktopus> <boter> @agrabeli for the "Run tests automatically" are you planning to provide the option of running tests more than once per day?
13:03:23 <slacktopus> <agrabeli> @kiran We'd be happy to support you in a large deployment of ooniprobe in India :slightly_smiling_face:
13:04:47 <slacktopus> <agrabeli> Regarding shutdowns (the total internet blackout), those are not measured by ooniprobe, since ooniprobe requires connecting to a network in order to perform testing. That said, we usually refer to third-party data sources (such as BGP announcement and Google traffic data) that in some cases helps provide insight when internet traffic is disrupted from a country. This though is harder to track when internet shutdowns only occur in a specific n
13:04:47 <slacktopus> country...
13:05:36 <slacktopus> <kiran> Happy to discuss this in person at RightsCon next week.
13:06:13 <slacktopus> <hellais> @boter support for running tests automatically, how you see it now in the app, is going to be superseeded by the probe orchestration system. That is designed to support also more frequent scheduled, but we are yet uncertain about how much of this we want to expose to end-users, because it can get quite confusing.
13:06:17 <slacktopus> <darkk> @kiran btw, are these wide-range blackouts mobile-only or do they affect landlines as well?
13:06:19 <slacktopus> <kiran> Given Android’s install base, Google is probably best placed to notice when an entire region goes offline.
13:06:23 <anadahz> @kiran Regarding adding URLs from users have a look on a similar project https://blocked.org.uk/, this can be integrated with probes deployed in country and test (almost) in real time the requested URLs.
13:06:31 <slacktopus> <kiran> @darkk mobile only.
13:06:32 <slacktopus> <sukarn> @kiran @agrabeli I'd love to be involved in the discussion too
13:06:46 <slacktopus> <sukarn> I'll be there at RightsCon
13:07:27 <slacktopus> <sukarn> @darkk Most shutdowns affect only mobiles; some shutdowns include broadband connections too.
13:09:33 <slacktopus> <agrabeli> @kiran @sukarn looking forward to discussing further with you both next week at RightsCon! :slightly_smiling_face:
13:09:34 <slacktopus> <sarath_ms> In general there is great enthusiasm in the Indian digital rights communities to increase the presence of ooni measurements from India. Many groups have expressed interests in organizing workshops to train their members on running ooni, especially after the announcement of the mobile app.
13:09:35 <slacktopus> <darkk> @kiran that's unfortunate. Unmetered landlines with routable IPv4 addresses may be scanned for availability from "outside" of the end-user's network and that will not affect end-user that much.
13:10:20 <slacktopus> <agrabeli> @sarath_ms great to hear! We'd be happy to support you with these trainings :slightly_smiling_face:
13:11:04 <slacktopus> <agrabeli> @sarath_ms Which reminds me --- we have some workshop slides that we can share with you and others interested in doing OONI trainings. I'll share them with you.
13:11:53 <slacktopus> <sarath_ms> @agrabeli You mentioned about it when we spoke. I am writing an email to you right now.
13:12:13 <slacktopus> <kiran> @darkk mobile data is also almost always behind carrier grade NAT
13:12:18 <slacktopus> <agrabeli> @sarath_ms Thanks! So sorry I forgot to send them earlier :(
13:14:01 <slacktopus> <darkk> @kiran exactly -- it's usually metered and NATed, so it's not scanable. That technique works well in Turkey, where landline is disconnected as well as mobile during shutdown events.
13:14:39 <slacktopus> <agrabeli> So given that we're 14 minutes past time, I'd like to ask you all to please fill out your non-availability for next month's community meeting in the following spreadsheet: https://docs.google.com/spreadsheets/d/1n9Z_U5fdI5ZMn109y0UchSTncIVau88CaHiWjC8sSoQ/edit#gid=0
13:15:02 <slacktopus> <agrabeli> That way we can aim to have the April community meeting at a time that hopefully works best for all :slightly_smiling_face:
13:15:44 <slacktopus> <agrabeli> Is there anything else that someone would like to add on this final topic of our meeting?
13:16:05 <slacktopus> <darkk> @sukarn do you ever sleep? :slightly_smiling_face:
13:16:19 <slacktopus> <sukarn> ^_^
13:18:01 <slacktopus> <agrabeli> Thank you everyone for joining this meeting!
13:18:29 <slacktopus> <agrabeli> I'll circulate an email on the #keepiton, #gv, #ooni-talk mailing lists for the date and time of the April community meeting
13:19:10 <slacktopus> <agrabeli> In the mean while, if you have any preferred dates/times, please don't hesitate to contact me either here or by dropping an email to maria@openobservatory.org
13:19:13 <anadahz> @agrabeli For completion can you add the links to subscribe to the lists #keepiton and #gv ?
13:19:31 <slacktopus> <agrabeli> @anadahz indeed! One sec :slightly_smiling_face:
13:19:44 <slacktopus> <agrabeli> the gv is invite only
13:20:07 <anadahz> Subscription link for the #ooni-talk mailing list: https://lists.torproject.org/cgi-bin/mailman/listinfo/ooni-talk
13:20:11 <slacktopus> <agrabeli> (I think)
13:20:28 <slacktopus> <agrabeli> keepiton: https://www.accessnow.org/keepiton/
13:20:43 <slacktopus> <agrabeli> (though many of the people here today are from that mailing list I think :))
13:21:39 <slacktopus> <agrabeli> In the mean while, if any of you would like to discuss anything further prior or independently of next month's meeting, please feel encouraged to join this channel and ping the OONI team anytime :slightly_smiling_face:
13:21:56 <anadahz> Also all OONI lists can be found find here: https://ooni.torproject.org/get-involved/
13:22:07 <slacktopus> <agrabeli> @anadahz thanks :slightly_smiling_face:
13:22:18 <hellais> excellent, thank you all again for attending!
13:22:20 <hellais> #endmeeting