[tor-project] Anti-censorship team monthly report: October 2020

[Cecylia Bocovich]

Hi everyone,

Here is a report of what we've been up to in October 2020:

== Snowflake ==

* Made it possible for the bridge to count user bandwidth more
accurately. The bridge bandwidth graphs are now closer to reality.
Released in version 0.4.2 of the Snowflake WebExtension.

https://bugs.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake/33157
* Made Snowflake proxies self-test their NAT characteristics more
robustly, by making a test connection. Released in version 0.5.0 of the
Snowflake WebExtension.

https://bugs.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake/40013
* Increased the frequency of Snowflake proxy polling, to keep pace with
the increasing number of clients.

https://bugs.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake-webext/17
* Updated bug-reporting instructions at snowflake.torproject.org.

https://bugs.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake-webext/15
* Fixed a crash in the Snowflake client caused by a race condition.

https://bugs.torproject.org/tpo/anti-censorship/pluggable-transports/snowflake/40017

== Rdsys ==

* We keep refactoring the code base to avoid technical debt.
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/20
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/19
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/18

* We made progress on the Salmon distributor. We're now working on an
experimental improvement that makes Salmon store client state in Tor
Browser instead of on the server.
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/21
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/15

* Experimented with metrics collection for rdsys and bridgestrap. The
combination of StatsD and Graphite is promising -- at least for internal
metrics.
https://gitlab.torproject.org/tpo/anti-censorship/rdsys/-/issues/14

== Bridgestrap ==

* Re-implemented the way bridgestrap tests bridges. Instead of spawning
a Tor process each time bridgestrap tests a bridge, it now re-uses its
Tor process and relies on the SETCONF controller command to fetch a
bridge's descriptor.
https://gitlab.torproject.org/tpo/anti-censorship/bridgestrap/-/issues/3

== Miscellaneous ==

* Studied the potential Tor interference that took place during the
election in Tanzania. An unusually high number of connections to
directory authorities and default bridges timed out but users were
generally still able to use the Tor network: we saw a sharp increase in
both directly connecting and bridge users:
https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-10-30&country=tz&events=off
https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-10-30&country=tz

* Added "First Contribution" label to several issues. If you're
interested in contributing, take a look at the following issues:
https://gitlab.torproject.org/groups/tpo/anti-censorship/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=First%20Contribution