TRVE Data update — November 2016
Hello TRVE Data supporters,
Here's an update on what we've been doing in November.
- Stephan has made great progress on understanding Tor network traffic. He is now able to trace almost all data transfer through the various layers — application data, Tor circuits, calls to the OpenSSL library, down to actual network packets — and thus analyse each packet to determine why it was sent or received. This is especially important when using Tor on mobile devices, since every network packet uses battery energy and costs money (depending on your data plan). By labelling network packets, we can compare protocol changes to examine how they will affect use of battery and cellular data allowance. Stephan and Alastair are planning to submit this work to the IMPS workshop.
- Diana is about to go on maternity leave with her second child. We are excited for her and wish her the very best!
- Martin presented a talk "Conflict resolution for eventual consistency" at the Code Mesh conference in London, and the GOTO conference in Berlin. The talk introduces the problems that occur when several people concurrently edit a document, and informally explains the key ideas behind our JSON CRDT algorithm.
- We have also received helpful feedback from reviewers about the JSON CRDT paper, which we will be addressing in the coming weeks. We have also continued working on the formalisation of the CRDT using Isabelle, as mentioned in last month's update. We have got a basic commutativity proof working, but there are still open challenges with regard to modelling network communication.
- Martin has finished writing his book Designing Data-Intensive Applications, and has submitted the manuscript to O'Reilly. It is about to go into production, which means some more review and editing work ahead, but the end is in sight. Having finished the book will free up more time to work on TRVE!
- In the UK, the Investigatory Powers Bill has been passed by parliament and is becoming law. Among other problems, this law threatens to weaken end-to-end encryption systems by allowing the secretary of state to mandate communications service providers to maintain the "technical capability" to recover the plaintext of encrypted communications, which has widely been interpreted as government-mandated backdoors and/or key escrow. We actively opposed this bill and provided evidence to parliament during the committee stage on why technical capability notices are dangerous, but unfortunately our concerns were ignored. It is not yet clear how this law will be interpreted and applied in practice, but we firmly believe that robust end-to-end encryption without backdoors is more important than ever, and we will continue working towards making it ubiquitous.