14:01:02 #startmeeting 2017-11-28 OONI Community Gathering 14:01:02 Meeting started Tue Nov 28 14:01:02 2017 UTC. The chair is hellais. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:02 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:01:11 Let it begin 14:01:30 Great! 14:01:46 I'm Maria from OONI. Feel free to introduce yourselves asynchronously :slightly_smiling_face: 14:02:04 Reminder: Please add the topics that you'd like us to discuss in this meeting in this pad: https://pad.riseup.net/p/ooni-community-meeting 14:02:18 Welcome, and hi! :slightly_smiling_face: 14:02:37 Hello! 14:03:31 Hey , I'm Shah Rukh from Bytes For All, Pakistan 14:04:29 @shahrukh hi, thanks for joining us :slightly_smiling_face: 14:04:52 let's get started with the 1st agenda item? 14:05:11 #1. PK case: how have OONI responded the incident? What can be improved? How can it be improved? 14:05:33 Would anyone like to share a few words on this? 14:05:44 (particularly those who added it to the agenda) 14:06:51 Yep, my question is vague post-incident question -- was it done in timely fashion, was it good enough, was OONI Run useful to measure stuff (do we want to measure that), and so on -- just a generic post-incident brainstorm. 14:08:43 @shahrukh I'd love to hear some words on that as you can see that from "outside of OONI Bubble" 14:08:52 well in pakistan OONI testing showed the list of the sites which were blocked but it was only testing from the capital 14:09:08 https://twitter.com/> these were the sites 14:09:29 most are the micro blogging site others are the news sites 14:09:48 so most of the population was in dark about whats really going on 14:10:32 I suggest that big cities like Karachi and Lahore must have somewhere a probe installed 14:11:07 because of different isp were behaving strangely 14:11:16 right 14:12:05 Is start time of the incident known? When have you heard first reports of these websites being blocked? 14:12:33 to keep a keen eye on the web traffic blockade i guess it would be a great idea to have different probe connected to different isp in different cities to gather the exact information 14:12:42 yep 14:13:05 @darkk to my knowledge they started blocking them on Saturday. that is also confirmed by OONI data. 14:13:16 three days ago 14:13:45 @shahrukh are they still blocked? I'm asking as we haven't received the measurements from 27th Nov yet 14:13:48 First "they" took out the news channels and then the online news portal 14:14:02 no thank god now the ban has ended 14:14:08 awesome! 14:14:33 I told my associates who were in the field to use different VPN 14:15:09 but after a while to the best of my knowledge the automated vpn stopped responding 14:15:18 behaving strangely 14:15:28 hm, throttled maybe? :S 14:15:32 sorry English is not my first lang ! 14:15:55 (@shahrukh no worries, Eng is not our 1st lang either :)) 14:16:00 hehehhe 14:16:33 what is "the automated vpn" ? 14:17:05 after that "they" noticed that social media was still reporting adequately 14:17:16 people were posting live videos 14:17:23 so they took it out out 14:17:40 after that we were sitting ducks 14:18:12 @darkk like Zenmate , browsec etc 14:18:43 Ah, okay, I see, so you mean "VPN" running as an in-browser extension, right? 14:18:50 not like putting in ip addresses and port numbers 14:19:04 or mobile application ! 14:19:05 ye 14:19:08 yes* 14:19:13 (as opposed to OpenVPN, Tor, ssh tunnels, etc) 14:19:46 Hm, we should probably consider including such circumvention tools in the testing... 14:20:07 most of the population is not literate enough to use these 14:20:19 yeah, we need something simpler 14:20:53 I use Sock5 proxy chains when needed or the tor project 14:21:20 cool 14:21:42 but yeah, there is definitely the need to simplify circumvention tech, to make it more easily usable by most people... 14:21:53 #agreed 14:22:12 So I can share some updates from the OONI analysis if you like 14:22:44 From an article I have come to known about a vulnerability in tor sys too :( 14:22:51 Thanks to the PK OONI Probe community, measurements were collected last weekend, testing the sites @shahrukh reported to be blocked 14:23:17 The data analysis of these measurements confirms that all of those sites were indeed blocked by means of DNS tampering 14:23:30 And that they were only blocked on Saturday and Sunday, 25th and 26th Nov 14:23:39 @shahrukh is this consistent with your experience? 14:23:56 yes , 14:24:23 yesterday morning the internet came to norm 14:24:50 Today we also received the measurements from 26th Nov, and they show that ISPs started blocking Facebook Messenger and the WhatsApp web interface as well 14:25:25 Have you been able to access Facebook Messenger and the WhatsApp web interface yesterday and today? 14:25:35 (we haven't received the measurements from 27th yet) 14:26:24 yesterday morning I couldn't connect to sites without vpn but in some of my friends from another city said they could connect without vpn 14:26:36 now its normal 14:26:37 interesting 14:26:50 this might indicate a lag between ISPs in terms of lifting blocks 14:27:01 good to hear that things work now 14:27:17 yea its a relief 14:27:33 Then, based on the measurements, it looks like they might have temporarily blocked Facebook Messenger and the WhatsApp web interface -- just on Sunday 14:28:02 We also found that the Telegram app appears to be blocked, since at least 20th Nov... 14:28:47 This screenshot shows Telegram blocking as of 20th Nov. (though the potential blocking may have started earlier than that) 14:29:01 @shahrukh do you use Telegram? Do you know if it's accessible locally? 14:29:11 (wondering whether our results are false positives) 14:29:55 I'm using a windows based phone telegram is not available on windows store ! 14:30:51 ok 14:31:07 do you perhaps have friends in PK who use Android or iOS? 14:31:27 it would be useful to know if people can access Telegram (since our measurements aren't very consistent) 14:31:42 yea sure I can ask them and mail you the responses 14:31:52 @shahrukh is this something local to PK? 14:31:55 @shahrukh thank you :slightly_smiling_face: 14:31:59 Because I see Telegram has a windows phone app: https://www.microsoft.com/it-it/store/p/telegram-messenger/9wzdncrdzhs0?rtc=1 14:32:22 My phone is not compatible 14:32:36 Ah ok I see 14:32:55 I was wondering if it had been removed from PK due to some regulations or something 14:33:55 @shahrukh I've heard, that it was army that forced ISPs to unblock those websites. Is it more or less correct and how did they justify the unblocking? Was it the army that ordered to *block* these websites just a days before? Is it something known or is it unclear? 14:34:37 I'm not asking about documented facts, I just try to understand local context a bit better. 14:35:53 In my personal opinion when the things get out of hands the gov influences the Pakistan Electronic Media Regulatory Authority to ban the site 14:36:19 sites or channel 14:36:54 Our country director can better guide you on this @darkk 14:37:17 @darkk here is the prohibition on news outlets: https://twitter.com/reportpemra/status/934324775161356289 14:37:34 (covering the protests) 14:39:02 can totally relate it to the quote "Censorship is fear of the Truth, There is no other reason for its existence !" 14:39:18 @agrabeli wow. thanks! I've expected to see alike document in Urdu so I've not even tried to look for it :slightly_smiling_face: 14:40:33 @darkk long story short, anti-blasphemy protests erupted over the last days, triggering the blocking of social media sites and news outlets 14:40:43 (and based on latest findings, blocking of communication tools as well) 14:40:59 In collaboration with Bytes for All, we have written a report that we plan to announce tomorrow 14:41:12 Great 14:41:23 Over the next hours we'll be finalizing the report based on the latest data analysis and new findings 14:41:33 *tomorrow = morning, PK time 14:42:06 we will be sharing that on our twitter/fb and website too 14:42:20 @shahrukh awesome, thank you! 14:42:35 @shahrukh are there any other ways we can support you in this? 14:43:21 Sorry idk I'm just an ICT officer here ! you may ask Shahzad thou 14:43:31 sure, thanks 14:43:31 or haroon 14:45:21 @agrabeli so, basically, unblocking happened like that -- army has completed their security operations (whatever those are), so justification to block the medias was lost and medias were unblocked. Is it something that is a least distantly correlated with perception of these events? 14:45:39 @shahrukh you rightfully pointed out that it's important to run tests across many different ISPs, particularly since censorship offer differs from region to region and from ISP to ISP 14:46:13 totally #agreed 14:46:55 @darkk that might be a hypothesis... not sure. 14:47:22 @shahrukh To that end, there are various ways of running OONI Probe, with different types of advantages and trade-offs. 14:47:50 The easiest way to run OONI Probe (and to encourage others to run it) is via the OONI Probe mobile apps (available on Android, iOS, and F-Droid) 14:48:22 if we can get official permission from different educational institutions to plant a few probes that can support or deny your hypotheses 14:48:32 You can also use OONI Run (https://run.ooni.io/) to add the sites you want to test, generate a link, and share it with OONI Probe mobile app users. They will automatically test the sites of your choice ;) 14:49:33 we can try to make a group of people from different locations to generate results and send them back ! 14:49:41 Alternatively, OONI Probe can be run on RasPis (as you know), or on macOS or Linux. 14:50:11 The results are automatically sent to our servers and published, everytime someone runs OONI Probe (unless they opt-out) 14:50:26 So no need to actively send measurements back --- all they need to do is run the software 14:50:33 oh thats that ! 14:50:54 All measurements are published in 2 places: 14:51:11 1. OONI Explorer: https://explorer.ooni.torproject.org/world/ 14:51:28 2. OONI API: https://api.ooni.io/ 14:51:48 The reason why we automatically publish all measurements is to increase transparency of internet censorship globally 14:52:01 And to support research, policy, and advocacy efforts (which can use the data) 14:53:04 So the easiest way to coordinate on testing across regions and ISPs is probably via the OONI Probe mobile apps, which are easy to install and run 14:53:23 And you can ask people to test the sites that you add to OONI Run: https://run.ooni.io/ 14:53:35 certainly 14:53:40 Our team has just started working on the development of OONI Probe desktop apps for Windows and macOS 14:53:56 The idea is that these desktop apps will (hopefully) be as easy to install and run as popular software 14:54:33 And we're currently revamping the mobile apps (to improve their design and usability) --- they will be relaunched ~ February 14:55:36 @shahrukh do you have other thoughts on how we can potentially improve our response to such incidents in the future? 14:56:02 Is there anything that we could do better from our side? 14:58:19 make a circle of people who can be trusted to run this "https://run.ooni.io/ " as these incidents are in most cases unpredictable ! 14:59:37 We can surely coordinate on this 14:59:55 If you have contacts who would be interested, please feel encouraged to connect them with us 15:00:11 most of the times they do it on the name of security but in my personal opinion they do it to hide the truth/facts , oppress them so that they cannot be out reached to the population and create a ripple 15:00:12 (I believe you also have our email addresses) 15:00:20 goood 15:00:24 good* 15:00:39 @shahrukh unfortunately that's often a motivation behind censorship in many countries around the world... :( 15:00:51 which is why we think it's important to monitor and document these incidents :slightly_smiling_face: 15:01:06 and why we find it particularly important to collaborate with local groups! 15:01:36 take iran for instance 15:01:45 they do it on the name of religion 15:01:54 All media is vetted by the Iranian government to ensure that it meets the strict religious ideals that are upheld by the country 15:02:31 yeah, there are many different types of motivations behind censorship decisions... 15:02:42 step 1 would be confirming them 15:03:02 which in itself can be quite challenging, since not everything that looks like censorship is necessarily intentional censorship 15:03:19 intentional censorship??? 15:03:25 (for example, a site might not be accessible because it's hosted on an unreliable server) 15:03:39 OH 500 error or something 15:03:44 got it ! 15:03:47 right 15:04:05 or maybe because the site itself is blocking IP addresses from a specific country in compliance with laws and regulations 15:04:32 yea that happpens! 15:05:03 for example, some websites that rely on Google App Engine are inaccessible in Cuba and Iran not because local ISPs are blocking them, but because Google blocks IP addresses from Cuba and Iran in compliance with US export laws 15:05:04 or just bug https://ooni.torproject.org/post/not-quite-network-censorship/ 15:05:10 right 15:05:36 So there are many reasons why a site might not be accessible... and not all of those reasons are because ISPs are intentionally blocking access 15:05:48 right 15:05:56 distinguishing these cases requires thorough network data analysis 15:05:57 WRT cause for censorship -- it's often "security". The political question is "who's security is it?", but it's not the question I want to ask as it makes the discussion quite close to utilitarian calculations I don't want to make :) 15:06:14 Are you guys behind the tor project ? 15:06:38 I would say the tor project is behind us 15:07:27 As in we are a sub-project of the tor project 15:07:30 @shahrukh The Tor Project Inc is our legal entity 15:07:54 As in, we are a project hosted under The Tor Project Inc. 15:08:06 awesome 15:08:20 (as a heads up, we are already 8 minutes overtime :clock10:) 15:08:52 Some of the most common motivations behind censorship (that we see around the world) include: 15:09:00 1. Security reasons (maintaining stability) 15:09:38 2. Hiding information (e.g. Malaysian case where ISPs blocked news articles and blogs covering a political scandal) 15:09:57 3. Politics (and reinforcing geopolitical dynamics of power) 15:10:05 4. Reinforcing social and cultural norms 15:10:19 5. Reinforcing laws and regulations 15:10:28 (point 5 being related to all of the above) 15:11:06 Many forms of internet censorship around the world are legally justified 15:11:35 By measuring censorship and collecting data which shows what is blocked and how, we can support public debate 15:11:56 But without the data, it's harder to an informed debate to begin with I think 15:12:26 yes without proof there is nothingness ! 15:13:11 (I forgot economic reasons in the list above) 15:13:29 Anyhow, we're running late... Is there anything else you would like to discuss? 15:13:46 No thank you for your time 15:14:08 You guys are doing an awesome job 15:14:12 @shahrukh thank *you* for joining us :slightly_smiling_face: 15:14:17 creating awareness 15:14:23 Our work wouldn't be possible without community members like you :slightly_smiling_face: 15:15:11 Most excellet 15:15:12 do ping me to give a heads up whenever the next time meeting happens 15:15:16 Thanks again for joining us, and looking forward to the next meeting! 15:15:19