16:59:48 #startmeeting 16:59:48 Meeting started Mon May 11 16:59:48 2015 UTC. The chair is hellais. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:59:48 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:00:54 so who is here? 17:01:00 here 17:03:12 hum I see 17:03:17 not many people around today 17:03:31 I guess I will still to a summary of what has happenned last week for the sake of the log 17:03:48 here 17:04:10 so last week there was this event at Stony Brook with a bunch of network measurement/internet censorship researchers that I attended 17:04:23 some of the highlights and next steps we defined there: 17:05:17 * we will work with citizenlab more on the test-lists stuff: https://github.com/citizenlab/test-lists. They like the services idea as well as the idea of making it into a web application. 17:05:28 * anadahz here 17:05:31 we also agreed to speak to herdict about getting some of their data 17:05:38 poly, anadahz: ah cool :) 17:06:07 I agreed to hack up a web frontend to allow users to submit new URLs as git pull requests 17:06:36 * I got some very good advice on the big data pipeline and the conclusion is: 17:07:00 - it's best to invest in hardware than to buy cloud stuff 17:07:13 - If we want to use cloud for something we should apply to the aws grants 17:07:52 - http://predict.org/ and http://aws.amazon.com/public-data-sets/ are good places to dump all our data onto 17:08:15 - Other tecnical stuff that I will not go into too much detail now 17:10:47 OK so from my side: 17:11:37 I'm trying to setup a "monster" server that is going to be used for the pipeline ^ 17:12:39 anadahz: how is that going? 17:13:03 The server was long time powered off so I spent some hours cleaning it (7U) 17:13:46 I made some space for it in the rack, I still haven't found the accompanied rack rails 17:14:11 the manual states that the rails are included! 17:14:46 hellais: should I talk about network meter? 17:14:47 we 'll most probably need some hard disks 17:15:28 I 'll have more updates this week 17:16:17 The good news is that we could use 2 x 1U servers next month 17:17:13 Meanwhile lepidopter release is ready it 'll be nice if we had some betatesters prior to 0.1 official release 17:17:44 poly: yes please 17:17:45 that's my update atm 17:18:28 * isabela has a quick request (can do it after ppls updates) 17:18:47 * isabela and hello :) 17:18:53 As regards my update: 17:18:56 isabela: hello 17:19:02 isabela: hi :) 17:19:29 - I'm working to integrate code to traceroute on Android devices with this PR (https://github.com/TheTorProject/libight/pull/117) 17:20:10 - I'm doing some progress as regards compiling libight on Android (compiles on ARM, patches to boost needed to compile on other archs) - https://github.com/TheTorProject/libight/pull/114 17:21:04 - we are discussing with mlab about the possibility to receive funding to develop libight 17:21:25 17:21:58 update on the tor censorship analyzer (and a bit more) 17:22:27 so I started working on implementing the gui for the tor censorship analyzer in python's tkinter 17:22:56 however, hellais suggested that we could instead work on a frontend for running both the censorship analyzer and ooniprobe 17:23:18 we also found a good cross platform toolkit to implement this in (electron) 17:23:57 the objective is to provide more visual feedback to the user as well as making running tests easier so that users are more likely to run the test more than once 17:24:38 so far, hellais has setup the code base of the project and I have started working on something of a design document 17:24:56 17:25:13 #link network meter design document: https://github.com/hellais/network-meter/blob/master/docs/design.md 17:25:45 poly: on this topic sbs also expressed the interest in collaborating on this project for having the GUI to NeuBot 17:26:00 his requirement though is that it should be possible to ship it to the app store and windows store 17:26:28 app store == mac store? 17:26:47 though for this to be possible an option should be available to create an app bundle that includes the software you need and disables the install capabilities 17:26:55 poly: yes mac store 17:27:30 their policy says "Apps that download code in any way or form will be rejected 17:27:35 Apps that install or launch other executable code will be rejected 17:28:42 hellais: sounds doable. I'll try to incorporate that into the design doc 17:28:53 poly: thanks 17:29:24 oh wait, I actually just realised that we were looking at the guidelines for the iOS store 17:29:27 not the mac one 17:29:34 :P 17:29:48 Apps that download or install additional code or resources to add functionality or change their primary purpose will be rejected 17:29:55 this is what the mac store says 17:30:16 Apps must be self-contained, single application installation bundles, and cannot install code or resources in shared locations 17:30:44 Apps must be packaged and submitted using Apple's packaging technologies included in Xcode - no third party installers allowed 17:30:48 this also need to be considered 17:30:50 hellais: uhm, I guess that means no plugins (= apps that download or install additional code to add functionality) 17:31:13 hellais: so firefox isn't allowed on the app store? 17:31:16 sbs: yes correct, so we would still have to do what we were thinkiong 17:31:19 *thinking 17:31:23 sbs: I don't think so 17:32:08 I'm thinking of having some kind of self-contained plugin and an option to disable plugin install by users when packaging 17:32:43 sbs: so you'd install the plugin for your application, disable the installation of any others and ship the app 17:33:13 this way other app developers can also include Network Meter into their code and be able to ship it to these stores 17:33:36 poly: yes, basically it's like NetworkMeter with predefined plugins shipping with it as loadable libraries that can be loaded from Node.js 17:34:31 In other words, the definition of plugin should encompass both external binaries and shared libraries 17:36:09 poly, hellais: if it could help, I was already writing something similar in pure Python for Neubot (with plugins only as external commands), you can find it here: https://github.com/bassosimone/neubot-scheduler 17:38:46 sbs: do you have any particular deadlines? 17:39:40 No, this is something that I'm currently working on on my 20% at Nexa, but I am quite interested into this thing, so I'd like to contribute :) 17:40:53 hellais: think we should have a seperate design doc for the plugins? 17:42:25 I would put the generic plugins specification in that one deign doc and one document per plugin 17:43:58 anyone else have some report backs? 17:44:22 isabela: what was your question? 17:46:58 ksorry 17:47:04 2 things 17:47:47 1. last week I helped core tor and sponsor r teams review their roadmaps 17:48:13 I am trying to make sure we keep them updated, so I wanted to ping you about it -> https://trac.torproject.org/projects/tor/wiki/org/roadmaps/OONI 17:48:39 2. now that I am out of the pile of reports I will reach out to people about hadoop etc 17:48:49 thats it 17:50:00 isabela: yes I will review the roadmap and update it. I think we are more or less in line with what is in there, though some unexpected events have occurred that lead us to do other things and move some stuff down on the roadmap 17:50:31 isabela: thanks for help in finding some big data people :) 17:50:44 sure, please keep in mind that is ok to move things around, maybe you didnt do something in april but you might want to move it up to june to pick it up again.. 17:51:15 thanks folks! 17:52:16 ack 17:54:46 so next steps? 17:55:27 As regards my work 17:55:45 I will try to make sure libight compiles on all Androids 17:55:54 while we wait for the server at the uni of vasilis to be setup I will continue testing and implementing the data pipeline on the old dedicated box we had 17:56:03 and a new one I acquired 17:56:28 I will also submit soon the proposal for mlab 17:56:33 hellais: sorry for overlapping 17:57:31 hellais: how much storage do we need ? 17:58:04 I'm not sure if we have enough hard disks that are usable 17:58:33 well since we are going for the dedicated way we should fill it up with as many disks we can fit 17:59:06 how many drive bays and what connectors are there in those servers? 17:59:36 sbs: more bandwidth :P 18:00:56 anadahz: anyways if you give me the specs I'll ship you 4x4TB disks 18:01:04 OK 18:01:39 i 'll hand complete specs off meeting 18:01:48 hellais: ^ 18:02:22 anadahz: excellent 18:02:27 ok anything else left? 18:04:44 well in that case I guess we can close this 18:04:49 thanks all for attending! 18:05:10 #endmeeting