OONI Team Meeting: Montreal 2017

Will Scott, Maria Xynou 2017-10-20

OONI team meeting in Montreal

Last week, right before the Tor meeting, the OONI team gathered in Montreal for a 4-day meeting to reflect, regroup, hack, and plan.

This post shares information from our meeting and future plans with the broader community. All session notes are available on GitHub.

Summary: What OONI is working on

We have our toes in many metaphorical pools. The major development efforts underway include:

Sessions

Below we outline the sessions we held during our Montreal team meeting and link to relevant notes.

Day 1 - 7th October 2017

A stronger team makes a stronger project. This session included a discussion of what worked well over the last year, and what didn’t. Based on this discussion, we developed strategies for improving the (horizontal) management of our project, intra-team collaboration, and for re-distributing work among team members.

Over the next year, we will be developing native OONI Probe apps for Windows and macOS. This session involved a discussion of the design and development choices and next steps towards implementation. The session notes are available here.

Navigation flow

Most of the global OONI community runs OONI Probe through our mobile apps. This session involved a discussion of some of the next steps in terms of feature requests and their implementation. The session notes are available here.

Over the next 2 years, we aim to expand OONI’s network of partners to work with many more organizations around the world on the study of internet censorship. This means that we need to ensure that we are better equipped to meet expanding community needs in terms of data analysis, censorship measurement testing coordination, and report writing, among other things. The session notes are available here.

Day 2 - 8th October 2017

Revamping OONI Explorer

No doubt, OONI Explorer needs to be revamped to be more useful to community members. As part of this session, we discussed what the “ideal” OONI Explorer would look like based on all the feedback that we have collected from community members over the last year and we discussed what’s required in terms of development and design. The session notes are available here.

Our current website can be better presented to better engage journalists, advocacy groups, and other new members of our community. As part of this session, we brainstormed on the new information architecture, design, and next steps towards implementation. The session notes are available here.

Through our partnership program, the OONI Partner Gathering, and other community engagement activities, we have been receiving feedback from community members on an ongoing basis. As part of this session, we discussed some of the top, recurring community needs and priorities that we have identified. We also discussed strategies and development tasks for meeting those needs. The session notes are available here.

Pipeline architecture

Over the last year, OONI’s Leonid has been re-engineering our data processing pipeline to analyze data faster and more accurately. As part of this session, he shared his knowledge with the rest of the team. Documentation from the knowledge share is available here.

Censorship Alert System

Over the next year we aim to create the first Censorship Alert System which will disseminate timely alerts of emergent censorship events based on OONI data. As part of this session, we discussed some of the next steps (and the questions to consider) towards creating the prototype. The session notes are available here.

We also aim to create the first methodology for the automatic examination and detection of internet blackouts around the world. As part of this session, we discussed some of the next steps towards designing and implementing this methodology, as well as foreseeable challenges and ways to mitigate them. The session notes are available here.

Day 3 - 9th October 2017

Over the last months we have been working on “Probe Orchestration”, the instruction of OONI Probe tests (for cases where OONI Probe users have opted-in). This can be particularly useful for collecting network measurements when emergent censorship events occur around the world. While Probe Orchestration has already been designed and developed, we’re still addressing security questions, which we discussed as part of this session. The session notes are available here.

Measurement-kit (MK) is the network measurement library that powers OONI Probe mobile apps. We are in the process of integrating more OONI Probe tests into MK and it will provide support for OONI’s upcoming desktop apps. As part of this session, we discussed all the new development tasks for MK over the next year. The session notes are available here.

Lepidopter is an OONI Probe distribution for Raspberry Pis that has been used by many of our partners over the last year for the stable collection of measurements. However, we have learned that shipping hardware can be time-consuming, challenging, and expensive, without always leading to the desired outcomes. As part of this session, we discussed questions around continuing to maintain this distribution, as well as potential next steps. The session notes are available here.

As part of this session, we discussed what each team member is most passionate about, and what they found less enjoyable over the last year. We subsequently mapped out and assigned roles and responsibilities among team members for all the various moving components of the OONI project. The aim of this session was to reflect on team members’ performance and to ensure that responsibilities are reasonably distributed among team members.

As a follow up to the previous session, we mapped out and identified many of the SPOFs in terms of human resources, infrastructure, code, etc. The aim of this session was to ensure that we have backup strategies for most (if not all) components of our project.

We are often overly consumed by that latest bug and striving to meet tight deliverable deadlines. As part of this session, we stepped back and tried to think about the big picture, together as a team. More specifically, we discussed OONI’s mission, what we want to achieve in the long-run, and what we should be doing to achieve those goals. The session notes are available here.

Day 4 - 10th October 2017

As part of this session, we discussed current grants that support OONI’s work, as well as the deliverables assigned to them. We also brainstormed on all the other things we want to be working on, but don’t currently have funding for (and mapped out relevant potential funders).

We discussed the limitations to current OONI Probe tests and all the other questions that we would like to answer. Based on this, we brainstormed on new censorship testing methodologies and filed 15 tickets.

Will Scott, an Oonitarian and researcher with Censored Planet, joined our meeting and provided us an overview of the project’s next steps. Based on this, we identified areas for collaboration on the study of internet censorship. The session notes are available here.

All of the previous sessions enabled us to gain a better understanding of our goals and priorities over the next year, based on which we had a two-hour road-mapping session for the next 12 months. We mapped out our activities for each month and assigned team members.

We spent the rest of the day hacking on OONI Probe and Measurement-Kit.

Meeting outcomes

There were several concrete outcomes from the gathering, including: