Version 1 (modified by isabela, 5 years ago) (diff)




OONI is a measurement platform for detecting symptoms of internet censorship.
It has also been used for measurements beyond that, for example using the data to look at how different users were treated based on their tor usage.
Other characteristics are the presence of software products / proxies present on different networks.
It's mostly a desktop client now, but there is work on a mobile client.

Moving Parts
There are tests and a test suite
there are also test lists with domains of interest. There's collaboration with citizen lab
Work on strengthening the data processing pipeline. can batch & analyize historic data and present it to users.
There are heuristics to determine places censorship is happening.
It currently doesn't handle blocking that isn't with a block-page.

There's discussion on figuring out blocking at granularities other than countries.
Dicussion on throttling. maybe can measure speed.
There aren't well developed tests for that.

How do we get testing through africa or more generally around the world?
There's work on developing a framework with 10 countries using partners in local organizations. The goal is to have the capacity to handle the legal and social implications of conducting the tests.
There isn't comfort yet about informed consent / safety around the measurement. It's one thing for an organization. It's another for individual users.

A mobile implementation of the probe for IOS/Android. In addition to the ooni tests, would also like to have the NDT tests.
it's a C++ library on mobile.
Some development tasks are shared to keep report format in sync

OONI Hopes & Dreams
Can we measure net neutrality violations?
OONI could do things like show that specific protocols are blocked / throttled.
Are there entities / networks who can help understand network health?
Would also like to work with partners who want to know about network health
Better detection of network level censorship that isn't obvious from e.g. blockpages, which are already (relatively) easy to confirm and report on.
Have the process of translating ooni data into advocacy. THe thought is to publish reports that ooni analyzes but allow partners to take up the advocacy.
is it possible for political groups to easily pick up the data?

Providing a central way for OONI to ask probes to run tasks.
Currently there's a slow reaction to real-world events.
There's no way to ask probes to confirm blocking when it's heard about in the real world. Instead, users need to get the new test and run it.
Organizations can already run their own deployments.

There's a partnership program to work with the in-country groups
getting tool more useful to non-technical users to help them understand the nature of the blocking. It'll support running the tool, telling you the reason, and telling you how you can get around that problem.
Developing an informed-consent procedure with lawyers.
Strengthening data processing to be able to do more than just block pages.
Building out raspberry pi based deployment.
There's also ongoing mobile library deployment.
Recommendation for looking for large political events that will occur and try to have measurements built out in those areas where there is already journalistic focus.

One possible strategy for deployment is to work with groups doing monitoring in countries (around elections or peace processes). Those entities often have diplomatic protection, and may be able to argue that network measurement is within their scope.

There's also censorship in the opposite direction where external users can't access services within a country.

The ooni model is in measuring censorship from your own location.
there are also instances where services will block access in specific countries due to their interpretation of sanctions.