OONI Team Meeting: Montreal 2017
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:
OONI Probe desktop apps for Windows and macOS: Developing native desktop apps that are easy to install and run.
OONI Probe mobile apps: Implementing new features based on community feedback, integrating more OONI Probe tests, and improving the design and usability.
Data analysis pipeline: Improving our data analysis techniques to detect censorship events faster and more accurately.
Censorship Alert System: Disseminating timely alerts of emergent censorship events worldwide based on OONI data.
Research reports: Publishing research reports based on the analysis of OONI data.
Partnerships: Expanding our collaboration with more groups and organizations around the world on the study of internet censorship.
Probe Orchestration: Creating a feedback loop for follow-up measurements and for collecting data when emergent censorship events occur.
OONI Explorer: Revamping OONI Explorer to make it more usable by community members, data scientists, researchers, journalists, and policy makers.
Internet blackouts: Creating the first methodology for the automatic detection and examination of internet blackouts around the world.
New censorship detection tests: Designing and developing methodologies for deeper and more diverse testing of internet censorship.
Below we outline the sessions we held during our Montreal team meeting and link to relevant notes.
Day 1 - 7th October 2017
- Session 1: Time for reflection!
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.
- Session 2: Building native OONI Probe apps for Windows and macOS
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.
- Session 3: OONI Probe mobile apps: Feature requests & next steps
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.
- Session 4: OONI Partnership Program: Next 2 years
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
- Session 1: 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.
- Session 2: Revamping the OONI website
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.
- Session 3: Community needs and priorities
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.
- Session 4: OONI data processing pipeline knowledge share
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.
- Session 5: Creating a 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.
- Session 6: Detecting internet blackouts: Next steps
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
- Session 1: Probe Orchestration: Next steps
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.
- Session 2: Measurement-kit: Next steps
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.
- Session 3: The future of Lepidopter
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.
- Session 4: Roles and responsibilities
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.
- Session 5: Avoiding Single Points of Failure (SPOF)
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.
- Session 6: What do you want OONI to be?
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
- Session 1: Grants
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).
- Session 2: New censorship testing methodologies
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.
- Session 3: Collaboration with Censored Planet
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.
- Session 4: Roadmap!
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.
- Session 5: Code hackathon: Deep-dive!
We spent the rest of the day hacking on OONI Probe and Measurement-Kit.
There were several concrete outcomes from the gathering, including:
Detailed roadmap for the next 12 months.
Assigning point people responsible for each track of deliverables, and mapping dependencies to make sure we understand the critical path for meeting goals.
15 bugs for “new testing methodologies” added to Measurement-Kit as opportunities for collaboration, pulling in new contributors, and development.
Progress on Node bindings for Measurement-Kit as a basis for desktop applications.
Scoping our planned Censorship Alert System, and clarifying our first step of extracting potential events as a new pipeline table for interactive analysis to better understand what events should be broadcast.
Scoping the next steps for the development of the desktop apps, improvements to the mobile apps, and the methodology for examining internet blackouts.