14:00:02 <hellais> #startmeeting OONI Community Meeting 2021-04-27
14:00:02 <MeetBot> Meeting started Tue Apr 27 14:00:02 2021 UTC.  The chair is hellais. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:02 <MeetBot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:00:48 <slacktopus> <agrabeli> <here> Hello everyone! Welcome to the *April 2021 OONI Community Meeting*! :):ooni::party_parrot:
14:00:58 <slacktopus> <agrabeli> How are you?
14:01:20 <slacktopus> <agrabeli> I'm Maria, working with OONI, and happy to see you here today. :)
14:01:34 <slacktopus> <agrabeli> Please feel encouraged to introduce yourselves (either now or later async). :)
14:02:09 <slacktopus> <cohosh> hi! I'm Cecylia from Tor :)
14:02:26 <slacktopus> <agrabeli> Thank you @cohosh for joining us! :)
14:03:06 <slacktopus> <sbs> Hi, I'm Simone and I work at OONI!
14:03:23 <slacktopus> <rajudev> Hi, I am Raju, volunteering with SFLC.in
14:03:37 <slacktopus> <agrabeli> Great to see you @rajudev! :)
14:04:05 <slacktopus> <rajudev> great to see you as well @agrabeli
14:04:25 <slacktopus> <agrabeli> As a reminder, please add topics you'd like to discuss today in this pad: https://pad.riseup.net/p/ooni-community-meeting-keep :)
14:04:47 <slacktopus> <rajudev> looks like we have minimal things on our agenda.
14:04:55 <slacktopus> <agrabeli> While you think about topics, I could start off with sharing some brief updates from the OONI team
14:05:08 <slacktopus> <rajudev> I have been submittine tests from app and ooniprobe cli since last two days
14:05:10 <slacktopus> <rajudev> from Uganda
14:05:36 <slacktopus> <agrabeli> @rajudev fantastic, thank you! And thank you for sharing updates on the channel. :pray:
14:06:20 <slacktopus> <agrabeli> You might be interested in our last report on the internet disruptions that occurred in Uganda during the January 2021 elections: https://ooni.org/post/2021-uganda-general-election-blocks-and-outage/
14:06:47 <slacktopus> <agrabeli> But we also document the social media + VPN blocks (as part of the OTT tax) here: https://ooni.org/post/uganda-social-media-tax/
14:06:57 <slacktopus> <rajudev> I read about this, found it through a web search
14:07:10 <slacktopus> <rajudev> good work there
14:07:26 <slacktopus> <agrabeli> @rajudev how is your internet experience currently in Uganda?
14:08:01 <slacktopus> <rajudev> I have been here since 2 days now.
14:08:16 <slacktopus> <rajudev> Internet is expensive, most services heavily blocked
14:08:37 <slacktopus> <rajudev> I did the Circumvention test on the app
14:08:55 <slacktopus> <rajudev> it says that Riseup is accessible, but Riseup VPN did not work for me
14:09:04 <slacktopus> <agrabeli> ah interesting
14:09:07 <slacktopus> <rajudev> nor did Calyx Institutes
14:09:47 <slacktopus> <rajudev> what worked for me now is an app called `InviZIble Pro`
14:10:03 <slacktopus> <rajudev> available on F-Droid.
14:10:29 <slacktopus> <rajudev> It routes traffic over the Tor network and encrypts your IP with DNSCrypt
14:10:55 <slacktopus> <agrabeli> great! And speaking of Tor -- are you able to use it in Uganda (without paying the OTT tax)?
14:11:06 <slacktopus> <rajudev> but features like WhatsApp Calls or Telegram calls do not work
14:11:10 <slacktopus> <agrabeli> (I'm also asking since we have @cohosh with us today)
14:11:23 <slacktopus> <rajudev> About tor, yes I am able to use it, but not with the default settings
14:11:37 <slacktopus> <rajudev> You need to tell tor to use bridges for connections.
14:11:44 <slacktopus> <rajudev> without bridges it does not connect
14:12:02 <slacktopus> <sbs> ideally we would like to discuss this with cyberta!
14:12:07 <slacktopus> <cohosh> @rajudev do you use default bridges that are built into the app?
14:12:22 <slacktopus> <rajudev> yes, I do
14:12:29 <slacktopus> <cohosh> I haven't heard of this tool, looking it up now...
14:13:12 <slacktopus> <rajudev> It works surprisingly well where VPNs do not work
14:13:21 <slacktopus> <agrabeli> It's great to hear that Tor works with the default bridges! <3
14:13:51 <slacktopus> <cohosh> It's interesting because according to our metrics we have more directly connecting users in Uganda than bridge users
14:14:05 <slacktopus> <cohosh> But, I'm not familiar with the ISPs/ASes there
14:14:29 <slacktopus> <rajudev> Network configuration might be changing, but that is how it is working for me now
14:14:32 <slacktopus> <agrabeli> It could be that most of the directing connecting users have paid the OTT tax, in which case Tor is probably not blocked.
14:14:33 <slacktopus> <cohosh> https://metrics.torproject.org/userstats-bridge-country.html?start=2021-01-27&end=2021-04-27&country=ug
14:14:51 <slacktopus> <cohosh> https://metrics.torproject.org/userstats-relay-country.html?start=2021-01-27&end=2021-04-27&country=ug&events=off
14:15:09 <slacktopus> <cohosh> Ah, interesting! Thanks
14:15:29 <slacktopus> <agrabeli> https://ooni.org/post/uganda-social-media-tax/#social-media-tax
14:16:21 <slacktopus> <rajudev> In Tanzania Twitter was blocked but worked flawlessly over RiseupVPN.
14:16:45 <slacktopus> <rajudev> In Uganda, a lot of websites/apps around 30-40 of them are still blocked
14:16:53 <slacktopus> <rajudev> unless you pay OTT tax
14:17:37 <slacktopus> <db1984> Hi all, Davide here, getting involved in net neutrality and censorship! @rajudev do different (less known) messaging apps like signal or matrix works? (i have only small programming experience, mostly a law background) sorry for the maybe dumb curiosity/question
14:18:03 <slacktopus> <rajudev> Those are the only messaging apps working for me
14:18:06 <slacktopus> <agrabeli> @rajudev if you know people in Tanzania willing to test twitter.com further, they can do so by opening the following link with their OONI Probe mobile app (and tapping "Run"):
14:18:14 <slacktopus> <agrabeli> https://run.ooni.io/nettest?tn=web_connectivity&ta=%7B%22urls%22%3A%5B%22https%3A%2F%2Ftwitter.com%2F%22%5D%7D&mv=1.2.0
14:18:17 <slacktopus> <rajudev> Signal works,
14:18:35 <slacktopus> <rajudev> I am asking people to message me on matrix, XMPP and briar
14:20:57 <slacktopus> <rajudev> any more updates, or we move to next point in agenda?
14:21:57 <slacktopus> <agrabeli> Thank you @rajudev for sharing updates from Uganda and Tanzania! And thank you for contributing measurements! :pray:
14:22:12 <slacktopus> <agrabeli> A few brief updates from the OONI team:
14:22:30 <slacktopus> <agrabeli> 1. The OONI Probe apps now include a *new Signal test*! :party_parrot:
14:22:49 <slacktopus> <agrabeli> You can learn about this test here: https://ooni.org/nettest/signal
14:23:04 <slacktopus> <agrabeli> ^^ It's included in the "Instant Messaging" card of the OONI Probe apps.
14:23:30 <slacktopus> <agrabeli> 2. We have released a *beta for automated testing on OONI Probe Android*! :) :party_parrot:
14:23:47 <slacktopus> <agrabeli> You can try it out by joining the beta program here: https://play.google.com/apps/testing/org.openobservatory.ooniprobe
14:24:07 <slacktopus> <agrabeli> And updating to *OONI Probe Android 3.0.0-beta.1*
14:24:29 <slacktopus> <agrabeli> In the beta app, you can *enable automated testing through the Settings of the app.*
14:25:04 <slacktopus> <agrabeli> In doing so, OONI Probe will run automatically every hour (without requiring you to do anything), testing 100 websites each time. :)
14:25:27 <slacktopus> <agrabeli> (eventually, we aim to add support for the rest of the tests -- excluding the performance ones -- for automated testing too)
14:26:22 <slacktopus> <agrabeli> As you will see in the Settings for Automated Testing, you can enable it to only run tests automatically when you're connected to WiFi and while your phone is charging. But if you'd also like to have tests run when your phone is not charging, you can disable the relevant option in the settings.
14:26:51 <slacktopus> <agrabeli> We'd greatly appreciate it if you could please try it out and share any feedback you may have (to help us improve it before the official release)! :pray:
14:27:38 <slacktopus> <agrabeli> You can *share feedback* by writing to my colleague @nuke (here or by emailing lorenzo@openobservatory.org), or by opening a ticket at: https://github.com/ooni/probe/issues
14:29:20 <slacktopus> <agrabeli> 3. Are you perhaps *running the python legacy version of ooniprobe*? :)
14:29:48 <slacktopus> <agrabeli> This is the old version of ooniprobe that's been around since ~ 2012.
14:30:25 <slacktopus> <agrabeli> The problem is that this (deprecated) version of ooniprobe relies on v2 onion services, which are being deprecated by October 2021. See: https://blog.torproject.org/v2-deprecation-timeline
14:30:52 <slacktopus> <agrabeli> This means that if you're still running legacy ooniprobe, it will no longer work from October 2021 onwards, and it will no longer contribute measurements.
14:31:24 <slacktopus> <agrabeli> We'd therefore greatly appreciate it if you could *please upgrade to OONI Probe CLI*: https://ooni.org/install/cli :)
14:31:44 <slacktopus> <agrabeli> And if you know others who may be running an old ooniprobe, please share the above info with them! :pray:
14:32:11 <slacktopus> <agrabeli> (we literally have a bunch of old ooniprobe measurements coming in every day, and we have no idea who's contributing them ,:))
14:33:01 <slacktopus> <agrabeli> Those are all the updates on my end. :) Do you perhaps have questions?
14:33:12 <slacktopus> <agrabeli> (welcome @ivalentovitch! :) )
14:34:11 <slacktopus> <ivalentovitch> Hi :)
14:34:59 <slacktopus> <agrabeli> @ivalentovitch as we don't have any other agenda items atm (https://pad.riseup.net/p/ooni-community-meeting-keep), is there perhaps a particular topic (or questions) you'd like to discuss? :)
14:35:26 <slacktopus> <agrabeli> (sorry for putting you on the spot! The same question applies to everyone else as well :) )
14:35:50 <slacktopus> <ivalentovitch> I usually do have questions, but this time I do not have any. I will be testing the ooni probe beta and get back to you with feedback :)
14:36:12 <slacktopus> <agrabeli> @ivalentovitch thanks a million! Greatly appreciated. :pray:
14:37:00 <slacktopus> <rajudev> none
14:37:12 <slacktopus> <agrabeli> As a side-note wrt automated testing: To avoid over cluttering the UI of the Test Results screen, the test results are automatically published on OONI Explorer (in real-time), but not available in the app itself. Just mentioning this in case this question comes up. :)
14:38:22 <slacktopus> <ivalentovitch> is there something, you would like us to pay special attention in the process of testing the beta version?
14:39:46 <slacktopus> <agrabeli> @ivalentovitch great question! Essentially just checking if it works for you as expected, and how you find the overall experience of automated testing. For example, do you find the pop-up window (informing you of automated tests) annoying? Would you prefer to have access to automated test results? Would you like to have additional settings for automated testing (perhaps for something we have not thought of)? etc.
14:40:09 <slacktopus> <rajudev> going AFK
14:40:37 <slacktopus> <agrabeli> thanks @rajudev for joining us! Have a great evening. :)
14:41:23 <slacktopus> <hellais> In other news, we have been making a bit of progress on the URL submission platform. We have a proof of concept which is being deployed to testing infrastructure and supports the basic workflow of: 1. Allowing editing and adding of URLs to the country lists 2. Open the PR with github for review
14:42:34 <slacktopus> <hellais> In implementing this prototype we made two decisions to keep the complexity of the system down, but still retain most functions: a. A user of the URL submission platform can only have one batch of changes done at a time. That is to say they can add as many URLs as they want, make as many edits as they wish, but once they click “submit for review”, their working copy will be freezed and they will not be able to add or edit URLs
14:42:34 <slacktopus> until it’s been reviewed. This is done so as to minimise the risk of conflicts in git. b. We decided for the first iteration to keep the review directly on the github platform, without having a native review backoffice inside of the URL submission system. We might change this in the future if the need for it is strong enough (or if we start having to deal with malicious/spammy submissions), but for the moment we thought it was best to keep it
14:42:35 <slacktopus> as simple as possible and closest to the current URL review process.
14:46:02 <slacktopus> <ivalentovitch> One question in this regard: we want to keep dead URLs in the existing lists. Do you envisage a mechanism to ask reviewers check if given URL is blocked before suggesting it should be deleted?
14:46:17 <slacktopus> <ivalentovitch> keep dead URLs if they are blocked
14:46:50 <slacktopus> <hellais> Yeah this is a good question.
14:46:54 <slacktopus> <agrabeli> @ivalentovitch ah, that's a brilliant question!
14:47:23 <slacktopus> <ivalentovitch> from my experience, volunteers rarely review existing URLs so this is not going to be a major problem in the begining
14:47:26 <slacktopus> <hellais> It was recently questioned in one of the PRs if this should even be our policy for keeping or not URLs and I have started to also think if we should reconsider it
14:48:00 <slacktopus> <hellais> after all if a site is no longer alive, even if it’s blocked, we don’t really care at that point anymore to measure it and are probably better off measuring another more relevant one
14:48:24 <slacktopus> <hellais> This is the recent thread about it: https://github.com/citizenlab/test-lists/pull/758
14:49:01 <slacktopus> <agrabeli> @hellais well it depends on the site. If it's, for example, a site that's politically critical and it continues being blocked for many years later (even after the domain has expired) that may show the impact of censorship.
14:49:27 <slacktopus> <agrabeli> i.e. that perhaps the site eventually "died/expired" as a result of ongoing blocking.
14:49:32 <slacktopus> <ivalentovitch> the main idea behind keeping some dead URLs in the list is to monitor what happens with blocked URLs over time. I believe some researchers (including CitLab) still like this idea
14:50:30 <slacktopus> <agrabeli> @ivalentovitch yeah, I like this idea too. I find it more interesting if it's a dead human rights/political criticism/media site, than if it's a dead porn/gambling site (since those are usually blocked in bulk).
14:50:55 <slacktopus> <ivalentovitch> but I agree that the current process of checking if given URL is blocked might be not very appealing to volunteers and they will just not follow it
14:51:30 <slacktopus> <hellais> We can automate that part in theory, though we would only be able to detect it automatically for confirmed blocked cases
14:51:54 <slacktopus> <agrabeli> perhaps we can do that for certain categories (as opposed to all dead websites)?
14:52:18 <slacktopus> <agrabeli> For example, keep (some) dead sites that fall under HUMR, POLR, NEWS categories
14:52:29 <slacktopus> <ivalentovitch> great idea!
14:52:30 <slacktopus> <agrabeli> Because if those are blocked, it's usually the result of targeted blocking
14:52:53 <slacktopus> <agrabeli> Whereas if PORN or GMB are blocked, they're usually blocked in the thousands (as a result of relevant legislation0
14:53:30 <slacktopus> <hellais> I think the fact we are keeping in the testing lists blocked websites might also be a bit in contrast with the notion that the citizenlab testing list is not a blocklist. It also means that as the blocklists grow we will accumulate more and more dead blocked URLs in the list which are no longer relevant because they point to dead content.
14:54:06 <slacktopus> <hellais> But I agree it would probably be wise to engage also other researchers and the folks from citizenlab about this topic
14:54:28 <slacktopus> <ivalentovitch> yes, some balance need to be achieved...
14:54:38 <slacktopus> <agrabeli> Perhaps this can be determined contextually?
14:54:55 <slacktopus> <agrabeli> On a case-by-case basis, when reviewing each country-specific test list
14:55:37 <slacktopus> <agrabeli> Or rather, perhaps we could have a policy that we only keep dead HUMR, NEWS, POLR websites, if those websites are determined to be relevant/interesting to keep for that specific country by the in-country researchers updating the list
14:56:17 <slacktopus> <ivalentovitch> I agree!
14:56:48 <slacktopus> <agrabeli> (I'll add my comments to the relevant ticket discussion)
14:57:26 <slacktopus> <ivalentovitch> I will do the same. may I ask another question: the ooni probe beta version will be automatically installed if one has the regular version of the probe (on Android), correct?
14:58:24 <slacktopus> <agrabeli> @ivalentovitch no, you need to join the beta program (https://play.google.com/apps/testing/org.openobservatory.ooniprobe), updating (through the Play Store) to OONI Probe Android 3.0.0-beta.1, and then inside the beta app, enable automated testing in the settings. :)
14:59:15 <slacktopus> <agrabeli> But when automated testing is officially released, your OONI Probe will update to that with the next regular update.
15:00:25 <slacktopus> <agrabeli> I think joining the beta program in general may be useful, because it would enable you to also test other upcoming features in the future, before they're officially released. :)
15:01:13 <slacktopus> <agrabeli> As time's up, I guess we can end the meeting here.
15:01:56 <slacktopus> <agrabeli> <here> Thank you everyone for joining us today! :) I'll share updates for the next monthly meeting on Slack and ooni-talk as always. Hope you all stay safe, and have a great day/evening! <3
15:01:58 <slacktopus> <ivalentovitch> Thank you  :)
15:02:09 <slacktopus> <cohosh> thank you!
15:02:41 <slacktopus> <agrabeli> and of course, please feel encouraged to engage in discussions on this channel anytime (irrespective of community meetings) :)
15:02:49 <hellais> #endmeeting